diff --git "a/ipynb/llm/2. Hyper_Fine_tune_Llama_2_with_MCQ.ipynb" "b/ipynb/llm/2. Hyper_Fine_tune_Llama_2_with_MCQ.ipynb" --- "a/ipynb/llm/2. Hyper_Fine_tune_Llama_2_with_MCQ.ipynb" +++ "b/ipynb/llm/2. Hyper_Fine_tune_Llama_2_with_MCQ.ipynb" @@ -1 +1 @@ -{"cells":[{"cell_type":"markdown","source":["---"],"metadata":{"id":"LlkcVvTDs96X"}},{"cell_type":"markdown","source":["# MCQ Custom Dataset Creation\n","\n","## 1. Introduction\n","\n","High-quality data is fundamental for producing a good model; the higher the quality of the data, the better the resulting model. The following steps outline the process of creating a dataset specifically for fine-tuning our Llama2 model.\n","\n","\n","\n","![](https://i.imgur.com/IDNhAWH.png)\n","\n","\n","There are several types of datasets that can be used to fine-tune Large Language Models (LLMs):\n","\n","1. **Instruction Datasets:** These datasets contain direct instructions or prompts followed by the correct or expected outputs.\n","\n","2. **Raw Completion:** This involves providing a prompt to the model and letting it generate a response without any predefined correct answer.\n","\n","3. **Preference Datasets:** These datasets include human feedback in the form of preferences, where annotators compare pairs of model outputs to determine which is better.\n","\n","4. **Human Feedback Data:** This is specific to Reinforcement Learning from Human Feedback (RLHF) and involves direct feedback on the model's outputs from human annotators.\n","\n","5. **Demonstration Data:** Also used in RLHF, these datasets consist of examples showing ideal model outputs or actions, typically created by humans.\n","\n","6. **Reward Modeling Data:** Used to train a reward model in RLHF, this dataset predicts human feedback on model outputs based on actual feedback data.\n","\n","7. **Dialogue Data:** Particularly relevant for conversational AI, this includes annotated conversations that indicate the quality of responses or provide corrections.\n","\n","\n","---\n","\n","\n","\n","* Typically, an instruction dataset is utilized for fine-tuning the Llama 2 Model. Since we are focusing on Supervised Fine Tuning, the instruction dataset becomes our primary choice.\n","\n","Therefore, we have 2 options:\n","\n","1. Create our own Instruction Dataset.\n","2. Modify an existing instruction dataset, which involves filtering, modifying, and enriching it.\n","\n","We have decided to proceed with the 1st option: creating our own Instruction Dataset.\n","\n","* This will involve prompt engineering and incorporating sanity checks to ensure quality and relevance."],"metadata":{"id":"wAQMA1-DKZZ5"}},{"cell_type":"markdown","source":["## 2. Load and analyze the dataset"],"metadata":{"id":"hU_mUK-nol-t"}},{"cell_type":"code","execution_count":null,"metadata":{"id":"8P7g6eHuxxKe"},"outputs":[],"source":["# Install libraries\n","!pip install -q datasets transformers sentence_transformers faiss-gpu huggingface_hub"]},{"cell_type":"code","source":["# Import the required libraries\n","import json\n","import sys\n","import pandas as pd\n","from datasets import Dataset, DatasetDict, load_dataset\n","\n","from transformers import AutoTokenizer\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","\n","from sentence_transformers import SentenceTransformer\n","import faiss\n","from tqdm.autonotebook import tqdm\n","import numpy as np"],"metadata":{"id":"KKb-ikj4J-in"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Load JSON data from a file\n","with open(\"mcq_data.json\", \"r\") as f:\n"," data = json.load(f)\n","\n","# Create a Pandas DataFrame from the list of dictionaries\n","df = pd.DataFrame(data)\n","\n","# Calculate the number of rows for each dataset split\n","num_rows = len(df)\n","train_end = int(num_rows * 0.8) # 80% for training\n","test_end = train_end + int(num_rows * 0.1) # 10% for testing\n","\n","# Split the DataFrame into training, testing, and validation sets\n","df_train = df[:train_end]\n","df_test = df[train_end:test_end]\n","df_val = df[test_end:] # Ensures the remainder is used for validation\n","\n","# Create Datasets from the DataFrames\n","dataset_train = Dataset.from_pandas(df_train)\n","dataset_test = Dataset.from_pandas(df_test)\n","dataset_val = Dataset.from_pandas(df_val)\n","\n","# Create a DatasetDict containing the split datasets\n","dataset = DatasetDict({\n"," 'train': dataset_train,\n"," 'test': dataset_test,\n"," 'val': dataset_val\n","})\n","\n","# Print the structure of the created DatasetDict\n","print(dataset)"],"metadata":{"id":"bGi9FdmhdBDg","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1708322802114,"user_tz":-480,"elapsed":19,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"e9369555-5be7-4b43-a6c4-3defec1485d6"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["DatasetDict({\n"," train: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 334\n"," })\n"," test: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 41\n"," })\n"," val: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 43\n"," })\n","})\n"]}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['train'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":719},"id":"-MOvcr5mD8li","executionInfo":{"status":"ok","timestamp":1708322802114,"user_tz":-480,"elapsed":18,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"5b69a1a1-5307-4ec8-9b04-2c5035176d98"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the structure of artificial n... \n","1 Create an MCQ on the training process of artif... \n","2 Create an MCQ on the role of artificial neuron... \n","3 Create an MCQ on the purpose of hidden layers ... \n","4 Create an MCQ on the basics of deep learning \n",".. ... \n","329 Create an MCQ on the hyperparameter 'Kernel' i... \n","330 Create an MCQ on the hyperparameter 'Gamma' in... \n","331 Create an MCQ on the hyperparameter 'learning_... \n","332 Create an MCQ on the hyperparameter 'n_estimat... \n","333 Create an MCQ on the application of deep learn... \n","\n"," Question \\\n","0 What is the structure of an artificial neural ... \n","1 What is the purpose of the training process in... \n","2 What is the role of artificial neurons in neur... \n","3 What is the purpose of hidden layers in artifi... \n","4 What is deep learning? \n",".. ... \n","329 What does the hyperparameter 'Kernel' define i... \n","330 What does the hyperparameter 'Gamma' control i... \n","331 What does the hyperparameter 'learning_rate' d... \n","332 What does the hyperparameter 'n_estimators' de... \n","333 Which of the following is an application of de... \n","\n"," A \\\n","0 It consists of input layers and hidden layers ... \n","1 To adjust the weights of the connections betwe... \n","2 To receive input from external sources \n","3 To receive input from external sources \n","4 A branch of machine learning based on artifici... \n",".. ... \n","329 The step size taken by the optimizer during ea... \n","330 The step size taken by the optimizer during ea... \n","331 The step size taken by the optimizer during ea... \n","332 The step size taken by the optimizer during ea... \n","333 Analyzing sensor data in autonomous vehicles \n","\n"," B \\\n","0 It consists of input layers, hidden layers, an... \n","1 To propagate input data forward through the la... \n","2 To compute the weighted total of inputs \n","3 To compute the weighted total of inputs \n","4 A programming technique to explicitly define c... \n",".. ... \n","329 The trade-off between the margin and the numbe... \n","330 The trade-off between the margin and the numbe... \n","331 The trade-off between the margin and the numbe... \n","332 The trade-off between the margin and the numbe... \n","333 Recognizing objects and scenes in images \n","\n"," C \\\n","0 It consists of input layers, hidden layers, ou... \n","1 To calculate the error between the output and ... \n","2 To transfer information to the next layer \n","3 To transfer information to the next layer \n","4 A method to process large datasets using deep ... \n",".. ... \n","329 The similarity between data points \n","330 The similarity between data points \n","331 The similarity between data points \n","332 The number of boosting trees to be trained \n","333 Transcribing spoken words into text \n","\n"," D Correct Answer \\\n","0 It consists of input layers, hidden layers, ou... C \n","1 To achieve the desired level of performance A \n","2 All of the above D \n","3 To process and transform the input data D \n","4 A type of data structure inspired by the human... A \n",".. ... ... \n","329 The maximum depth of each tree in the ensemble C \n","330 The influence of support vectors on the decisi... D \n","331 The maximum depth of each tree in the ensemble A \n","332 The maximum depth of each tree in the ensemble C \n","333 Making personalized recommendations based on u... B \n","\n"," Explanation \n","0 An artificial neural network consists of input... \n","1 The purpose of the training process in artific... \n","2 The role of artificial neurons in neural netwo... \n","3 The purpose of hidden layers in artificial neu... \n","4 Deep learning is a branch of machine learning ... \n",".. ... \n","329 The hyperparameter 'Kernel' in Support Vector ... \n","330 The hyperparameter 'Gamma' in Support Vector M... \n","331 The hyperparameter 'learning_rate' in XGBoost ... \n","332 The hyperparameter 'n_estimators' in XGBoost d... \n","333 Deep learning algorithms are used in image and... \n","\n","[334 rows x 8 columns]"],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the structure of artificial n...What is the structure of an artificial neural ...It consists of input layers and hidden layers ...It consists of input layers, hidden layers, an...It consists of input layers, hidden layers, ou...It consists of input layers, hidden layers, ou...CAn artificial neural network consists of input...
1Create an MCQ on the training process of artif...What is the purpose of the training process in...To adjust the weights of the connections betwe...To propagate input data forward through the la...To calculate the error between the output and ...To achieve the desired level of performanceAThe purpose of the training process in artific...
2Create an MCQ on the role of artificial neuron...What is the role of artificial neurons in neur...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerAll of the aboveDThe role of artificial neurons in neural netwo...
3Create an MCQ on the purpose of hidden layers ...What is the purpose of hidden layers in artifi...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerTo process and transform the input dataDThe purpose of hidden layers in artificial neu...
4Create an MCQ on the basics of deep learningWhat is deep learning?A branch of machine learning based on artifici...A programming technique to explicitly define c...A method to process large datasets using deep ...A type of data structure inspired by the human...ADeep learning is a branch of machine learning ...
...........................
329Create an MCQ on the hyperparameter 'Kernel' i...What does the hyperparameter 'Kernel' define i...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe maximum depth of each tree in the ensembleCThe hyperparameter 'Kernel' in Support Vector ...
330Create an MCQ on the hyperparameter 'Gamma' in...What does the hyperparameter 'Gamma' control i...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe influence of support vectors on the decisi...DThe hyperparameter 'Gamma' in Support Vector M...
331Create an MCQ on the hyperparameter 'learning_...What does the hyperparameter 'learning_rate' d...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe maximum depth of each tree in the ensembleAThe hyperparameter 'learning_rate' in XGBoost ...
332Create an MCQ on the hyperparameter 'n_estimat...What does the hyperparameter 'n_estimators' de...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleCThe hyperparameter 'n_estimators' in XGBoost d...
333Create an MCQ on the application of deep learn...Which of the following is an application of de...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...BDeep learning algorithms are used in image and...
\n","

334 rows × 8 columns

\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['train']\",\n \"rows\": 334,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"Create an MCQ on the parameter gamma in Support Vector Machines (SVMs)\",\n \"Create an MCQ on the disadvantages of Artificial Neural Networks (ANNs)\",\n \"Create an MCQ on the role of machine learning in recommendation systems\"\n ],\n \"num_unique_values\": 165,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"What does the hyperparameter 'Kernel' determine in Support Vector Machines (SVMs)?\",\n \"Which of the following are types of deep learning architectures?\",\n \"Which of the following is NOT an application of deep learning in reinforcement learning?\"\n ],\n \"num_unique_values\": 221,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"Requires large amounts of labeled data\",\n \"AI is the broader family consisting of ML and DL as its components\",\n \"Increased computational cost\"\n ],\n \"num_unique_values\": 162,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Data clustering, dimensionality reduction, and anomaly detection\",\n \"Analyzing medical images to assist doctors in making diagnoses\",\n \"The reliance on manual feature engineering\"\n ],\n \"num_unique_values\": 193,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"It may result in overfitting\",\n \"Reduced overfitting and underfitting\",\n \"To automatically learn features from visual data\"\n ],\n \"num_unique_values\": 211,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Overfitting\",\n \"Evaluating all possible combinations of hyperparameter values\",\n \"A branch of machine learning that uses linear regression\"\n ],\n \"num_unique_values\": 211,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"A\",\n \"B\",\n \"C\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"A key difference between machine learning and deep learning is the type of algorithms used. Machine learning applies statistical algorithms, while deep learning utilizes artificial neural network architecture to learn patterns and relationships.\",\n \"Hyperparameter tuning helps reduce overfitting and underfitting, leading to improved model performance and generalizability.\",\n \"Artificial Intelligence consists of the components: Artificial Intelligence, Machine Learning, and Deep Learning.\"\n ],\n \"num_unique_values\": 302,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":19}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['test'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"7WkFWJSQhKUV","executionInfo":{"status":"ok","timestamp":1708322802115,"user_tz":-480,"elapsed":17,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"8fd77b53-8311-42f7-a84d-00e396647a72"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the use of deep learning in g... \n","1 Create an MCQ on the use of deep learning in a... \n","2 Create an MCQ on the application of deep learn... \n","3 Create an MCQ on the use of deep learning in r... \n","4 Create an MCQ on the structure of artificial n... \n","5 Create an MCQ on the purpose of adjusting weig... \n","6 Create an MCQ on the role of artificial neuron... \n","7 Create an MCQ on the complexities of neural ne... \n","8 Create an MCQ on the difference between machin... \n","9 Create an MCQ on the definition of deep learning \n","10 Create an MCQ on the key characteristic of dee... \n","11 Create an MCQ on the applications of deep lear... \n","12 Create an MCQ on the training requirements of ... \n","13 Create an MCQ on the types of machine learning... \n","14 Create an MCQ on the types of neural networks ... \n","15 Create an MCQ on the applications of deep lear... \n","16 Create an MCQ on the applications of deep lear... \n","17 Create an MCQ on the applications of deep lear... \n","18 Create an MCQ on the main types of neural netw... \n","19 Create an MCQ on the definition of Artificial ... \n","20 Create an MCQ on the difference between Machin... \n","21 Create an MCQ on the components of Artificial ... \n","22 Create an MCQ on the aim of Machine Learning \n","23 Create an MCQ on the aim of Deep Learning \n","24 Create an MCQ on the difference between AI, Ma... \n","25 Create an MCQ on the application of AI in spee... \n","26 Create an MCQ on the application of AI in pers... \n","27 Create an MCQ on the application of AI in pred... \n","28 Create an MCQ on the application of AI in medi... \n","29 Create an MCQ on the difference between AI, ML... \n","30 Create an MCQ on the responsibilities of an AI... \n","31 Create an MCQ on the skills required for a Mac... \n","32 Create an MCQ on the tasks of a Deep Learning ... \n","33 Create an MCQ on the difference between ML and DL \n","34 Create an MCQ on the advantages of Artificial ... \n","35 Create an MCQ on the disadvantages of Artifici... \n","36 Create an MCQ on the advantages of Biological ... \n","37 Create an MCQ on the disadvantages of Biologic... \n","38 Create an MCQ on the differences between Artif... \n","39 Create an MCQ on hyperparameter tuning in mach... \n","40 Create an MCQ on the types of hyperparameters ... \n","\n"," Question \\\n","0 What is the role of deep learning in generativ... \n","1 How is deep learning used in autonomous vehicles? \n","2 What is the role of deep learning in speech re... \n","3 What is the application of deep learning in re... \n","4 What is the structure of an artificial neural ... \n","5 What is the purpose of adjusting weights in ar... \n","6 What is the role of artificial neurons in neur... \n","7 What determines the complexities of neural net... \n","8 What is a key difference between machine learn... \n","9 What is the definition of deep learning? \n","10 What is the key characteristic of deep learning? \n","11 Which of the following are applications of dee... \n","12 What are the training requirements for deep ne... \n","13 Which types of machine learning tasks can be p... \n","14 Which type of neural network is specifically d... \n","15 Which application of deep learning in computer... \n","16 Which application of deep learning in NLP invo... \n","17 Which application of deep learning in reinforc... \n","18 Which of the following are the main types of n... \n","19 Which of the following best defines Artificial... \n","20 What is the main difference between Machine Le... \n","21 Which of the following components are part of ... \n","22 What is the aim of Machine Learning? \n","23 What is the aim of Deep Learning? \n","24 Which of the following best describes the diff... \n","25 Which of the following is an example of AI app... \n","26 Which of the following is an example of AI app... \n","27 Which of the following is an example of AI app... \n","28 Which of the following is an example of AI app... \n","29 Which of the following statements accurately d... \n","30 Which of the following is a key responsibility... \n","31 Which of the following skills is essential for... \n","32 Which of the following is a key task of a Deep... \n","33 What distinguishes Deep Learning (DL) from Mac... \n","34 Which of the following is an advantage of Arti... \n","35 Which of the following is a disadvantage of Ar... \n","36 Which of the following is an advantage of Biol... \n","37 Which of the following is a disadvantage of Bi... \n","38 Which of the following is a difference between... \n","39 What is the purpose of hyperparameter tuning i... \n","40 Which of the following is a type of hyperparam... \n","\n"," A \\\n","0 Analyzing sensor data in autonomous vehicles \n","1 Analyzing sensor data in autonomous vehicles \n","2 Analyzing sensor data in autonomous vehicles \n","3 Analyzing sensor data in autonomous vehicles \n","4 It consists of input layers, hidden layers, an... \n","5 To increase the speed of training models \n","6 To receive input from external sources \n","7 The number of layers in the network \n","8 The type of algorithms used \n","9 A branch of machine learning that uses artific... \n","10 The use of shallow neural networks with a sing... \n","11 Image recognition, natural language processing... \n","12 Small datasets and limited computational resou... \n","13 Supervised machine learning only \n","14 Feedforward Neural Networks (FNNs) \n","15 Object detection and recognition \n","16 Automatic Text Generation \n","17 Game playing \n","18 Feedforward Neural Networks (FNNs) \n","19 The study of training machines to mimic human ... \n","20 Machine Learning uses statistical methods, whi... \n","21 Machine Learning and Deep Learning \n","22 To increase chances of success \n","23 To increase chances of success \n","24 AI is a subset of Machine Learning, which is a... \n","25 Analyzing users' browsing and viewing history ... \n","26 Analyzing users' browsing and viewing history ... \n","27 Analyzing users' browsing and viewing history ... \n","28 Analyzing users' browsing and viewing history ... \n","29 AI, ML, and DL are interchangeable terms that ... \n","30 Design and development of AI algorithms \n","31 Strong background in computer science, mathema... \n","32 Design and development of DL algorithms \n","33 DL is a more advanced form of ML that can perf... \n","34 Ability to learn irrespective of the type of data \n","35 Ability to learn irrespective of the type of data \n","36 Ability to learn irrespective of the type of data \n","37 Ability to learn irrespective of the type of data \n","38 Both ANNs and BNNs have complex and diverse ne... \n","39 To adjust the weights and biases of the model \n","40 Weights \n","\n"," B \\\n","0 Creating new content based on existing data \n","1 Recognizing objects and scenes in images \n","2 Recognizing objects and scenes in images \n","3 Recognizing objects and scenes in images \n","4 It consists of input layers and output layers ... \n","5 To prevent overfitting by validating the model... \n","6 To compute the weighted total of inputs \n","7 The number of units in each layer \n","8 The amount of data required \n","9 A type of programming that explicitly defines ... \n","10 The requirement for manual feature engineering \n","11 Data clustering, dimensionality reduction, and... \n","12 Large amounts of data and computational resources \n","13 Unsupervised machine learning only \n","14 Convolutional Neural Networks (CNNs) \n","15 Image classification \n","16 Language translation \n","17 Robotics \n","18 Convolutional Neural Networks (CNNs) \n","19 The study of statistical methods enabling mach... \n","20 Machine Learning focuses on learning from expe... \n","21 Machine Learning and Decision Trees \n","22 To increase accuracy \n","23 To increase accuracy \n","24 Machine Learning is a subset of AI, which is a... \n","25 Analyzing medical images to assist doctors in ... \n","26 Analyzing medical images to assist doctors in ... \n","27 Analyzing medical images to assist doctors in ... \n","28 Analyzing medical images to assist doctors in ... \n","29 AI focuses on creating intelligent machines, M... \n","30 Analysis and interpretation of data \n","31 Experience in developing AI algorithms and sol... \n","32 Analysis and interpretation of data \n","33 DL focuses on developing algorithms that enabl... \n","34 Simple architecture that makes it easy to expl... \n","35 Simple architecture that makes it easy to expl... \n","36 Simple architecture that makes it easy to expl... \n","37 Simple architecture that makes it easy to expl... \n","38 ANNs have fixed connections between neurons, w... \n","39 To select the optimal values for the model's h... \n","40 Biases \n","\n"," C \\\n","0 Transcribing spoken words into text \n","1 Transcribing spoken words into text \n","2 Transcribing spoken words into text \n","3 Making personalized recommendations based on u... \n","4 It consists of input layers, hidden layers, ou... \n","5 To enhance the model's performance on the trai... \n","6 To transfer information to the next layer \n","7 The type of activation function used \n","8 The complexity of the models \n","9 A technique that requires manual feature engin... \n","10 The use of deep neural networks with multiple ... \n","11 Supervised machine learning and unsupervised m... \n","12 Manual feature engineering and domain expertise \n","13 Reinforcement machine learning only \n","14 Recurrent Neural Networks (RNNs) \n","15 Image segmentation \n","16 Sentiment analysis \n","17 Control systems \n","18 Recurrent Neural Networks (RNNs) \n","19 The study that uses neural networks to imitate... \n","20 Machine Learning is a subset of Deep Learning \n","21 Artificial Intelligence and Machine Learning \n","22 To improve system efficiency \n","23 To improve system efficiency \n","24 Deep Learning is a subset of AI, which is a su... \n","25 Recognizing and classifying images and speech \n","26 Recognizing and classifying images and speech \n","27 Analyzing sensor data to predict equipment fai... \n","28 Recognizing and classifying images and speech \n","29 AI is a subset of ML that uses neural networks... \n","30 Training and evaluation of ML models \n","31 Familiarity with programming languages such as... \n","32 Training and evaluation of ML models \n","33 DL is a subset of ML that uses neural networks... \n","34 Dependence on hardware for functioning \n","35 Dependence on hardware for functioning \n","36 No controlling mechanism \n","37 No controlling mechanism \n","38 Both ANNs and BNNs have simple and predetermin... \n","39 To preprocess the input data before training t... \n","40 Learning rate \n","\n"," D Correct Answer \\\n","0 Making personalized recommendations based on u... B \n","1 Making personalized recommendations based on u... A \n","2 Making personalized recommendations based on u... C \n","3 Transcribing spoken words into text C \n","4 It consists of input layers and artificial neu... A \n","5 To reduce the computational cost of training C \n","6 All of the above D \n","7 The size of the dataset B \n","8 The performance on complex tasks B \n","9 A method of machine learning that only works w... A \n","10 The reliance on labeled datasets for training C \n","11 Data visualization and exploratory data analysis A \n","12 Pre-trained models and transfer learning B \n","13 Supervised, unsupervised, and reinforcement ma... D \n","14 None of the above B \n","15 None of the above A \n","16 Speech recognition C \n","17 None of the above B \n","18 All of the above D \n","19 The study of incorporating human intelligence ... D \n","20 Machine Learning requires human intervention, ... A \n","21 Artificial Intelligence and Deep Learning C \n","22 To analyze data and provide output B \n","23 To analyze data and provide output A \n","24 AI, Machine Learning, and Deep Learning are co... B \n","25 Analyzing sensor data to make decisions about ... C \n","26 Analyzing sensor data to make decisions about ... A \n","27 Recognizing and classifying images and speech C \n","28 Analyzing sensor data to make decisions about ... B \n","29 AI focuses on developing algorithms that enabl... B \n","30 Deployment and maintenance of DL models A \n","31 All of the above D \n","32 Deployment and maintenance of AI models A \n","33 ML is a more advanced form of DL that can perf... C \n","34 High speed of processing A \n","35 The simplest architecture makes it difficult t... D \n","36 Ability to process highly complex parallel inputs D \n","37 Speed of processing is slow D \n","38 ANNs and BNNs have the same processing speed B \n","39 To evaluate the performance of the model on a ... B \n","40 Activation function C \n","\n"," Explanation \n","0 Deep learning algorithms are used in generativ... \n","1 Deep learning algorithms are used in autonomou... \n","2 Deep learning algorithms are used in speech re... \n","3 Deep learning algorithms are used in recommend... \n","4 An artificial neural network consists of input... \n","5 The purpose of adjusting weights in artificial... \n","6 The role of artificial neurons in neural netwo... \n","7 The complexities of neural networks are determ... \n","8 A key difference between machine learning and ... \n","9 Deep learning is a branch of machine learning ... \n","10 The key characteristic of deep learning is the... \n","11 Deep learning has achieved significant success... \n","12 Training deep neural networks typically requir... \n","13 Deep learning can be used for supervised, unsu... \n","14 Convolutional Neural Networks (CNNs) are speci... \n","15 Object detection and recognition is the applic... \n","16 Sentiment analysis is the application of deep ... \n","17 Robotics is the application of deep learning i... \n","18 The main types of neural networks used in deep... \n","19 Artificial Intelligence is the mechanism to in... \n","20 The main difference between Machine Learning a... \n","21 Artificial Intelligence is the broader family ... \n","22 The aim of Machine Learning is to increase acc... \n","23 The aim of Deep Learning is to increase chance... \n","24 AI is the broader concept that encompasses the... \n","25 Speech recognition is an example of AI applica... \n","26 Personalized recommendations, as an AI applica... \n","27 AI-powered predictive maintenance systems anal... \n","28 AI-powered medical diagnosis systems analyze m... \n","29 AI, ML, and DL are related but distinct concep... \n","30 One of the key responsibilities of an AI Engin... \n","31 A Machine Learning Engineer should have a stro... \n","32 One of the key tasks of a Deep Learning Engine... \n","33 Deep Learning (DL) is a subset of Machine Lear... \n","34 One of the advantages of Artificial Neural Net... \n","35 One of the disadvantages of Artificial Neural ... \n","36 One of the advantages of Biological Neural Net... \n","37 One of the disadvantages of Biological Neural ... \n","38 One of the differences between Artificial Neur... \n","39 Hyperparameter tuning is the process of select... \n","40 In neural networks, the learning rate is a hyp... "],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the use of deep learning in g...What is the role of deep learning in generativ...Analyzing sensor data in autonomous vehiclesCreating new content based on existing dataTranscribing spoken words into textMaking personalized recommendations based on u...BDeep learning algorithms are used in generativ...
1Create an MCQ on the use of deep learning in a...How is deep learning used in autonomous vehicles?Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...ADeep learning algorithms are used in autonomou...
2Create an MCQ on the application of deep learn...What is the role of deep learning in speech re...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...CDeep learning algorithms are used in speech re...
3Create an MCQ on the use of deep learning in r...What is the application of deep learning in re...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesMaking personalized recommendations based on u...Transcribing spoken words into textCDeep learning algorithms are used in recommend...
4Create an MCQ on the structure of artificial n...What is the structure of an artificial neural ...It consists of input layers, hidden layers, an...It consists of input layers and output layers ...It consists of input layers, hidden layers, ou...It consists of input layers and artificial neu...AAn artificial neural network consists of input...
5Create an MCQ on the purpose of adjusting weig...What is the purpose of adjusting weights in ar...To increase the speed of training modelsTo prevent overfitting by validating the model...To enhance the model's performance on the trai...To reduce the computational cost of trainingCThe purpose of adjusting weights in artificial...
6Create an MCQ on the role of artificial neuron...What is the role of artificial neurons in neur...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerAll of the aboveDThe role of artificial neurons in neural netwo...
7Create an MCQ on the complexities of neural ne...What determines the complexities of neural net...The number of layers in the networkThe number of units in each layerThe type of activation function usedThe size of the datasetBThe complexities of neural networks are determ...
8Create an MCQ on the difference between machin...What is a key difference between machine learn...The type of algorithms usedThe amount of data requiredThe complexity of the modelsThe performance on complex tasksBA key difference between machine learning and ...
9Create an MCQ on the definition of deep learningWhat is the definition of deep learning?A branch of machine learning that uses artific...A type of programming that explicitly defines ...A technique that requires manual feature engin...A method of machine learning that only works w...ADeep learning is a branch of machine learning ...
10Create an MCQ on the key characteristic of dee...What is the key characteristic of deep learning?The use of shallow neural networks with a sing...The requirement for manual feature engineeringThe use of deep neural networks with multiple ...The reliance on labeled datasets for trainingCThe key characteristic of deep learning is the...
11Create an MCQ on the applications of deep lear...Which of the following are applications of dee...Image recognition, natural language processing...Data clustering, dimensionality reduction, and...Supervised machine learning and unsupervised m...Data visualization and exploratory data analysisADeep learning has achieved significant success...
12Create an MCQ on the training requirements of ...What are the training requirements for deep ne...Small datasets and limited computational resou...Large amounts of data and computational resourcesManual feature engineering and domain expertisePre-trained models and transfer learningBTraining deep neural networks typically requir...
13Create an MCQ on the types of machine learning...Which types of machine learning tasks can be p...Supervised machine learning onlyUnsupervised machine learning onlyReinforcement machine learning onlySupervised, unsupervised, and reinforcement ma...DDeep learning can be used for supervised, unsu...
14Create an MCQ on the types of neural networks ...Which type of neural network is specifically d...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)None of the aboveBConvolutional Neural Networks (CNNs) are speci...
15Create an MCQ on the applications of deep lear...Which application of deep learning in computer...Object detection and recognitionImage classificationImage segmentationNone of the aboveAObject detection and recognition is the applic...
16Create an MCQ on the applications of deep lear...Which application of deep learning in NLP invo...Automatic Text GenerationLanguage translationSentiment analysisSpeech recognitionCSentiment analysis is the application of deep ...
17Create an MCQ on the applications of deep lear...Which application of deep learning in reinforc...Game playingRoboticsControl systemsNone of the aboveBRobotics is the application of deep learning i...
18Create an MCQ on the main types of neural netw...Which of the following are the main types of n...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)All of the aboveDThe main types of neural networks used in deep...
19Create an MCQ on the definition of Artificial ...Which of the following best defines Artificial...The study of training machines to mimic human ...The study of statistical methods enabling mach...The study that uses neural networks to imitate...The study of incorporating human intelligence ...DArtificial Intelligence is the mechanism to in...
20Create an MCQ on the difference between Machin...What is the main difference between Machine Le...Machine Learning uses statistical methods, whi...Machine Learning focuses on learning from expe...Machine Learning is a subset of Deep LearningMachine Learning requires human intervention, ...AThe main difference between Machine Learning a...
21Create an MCQ on the components of Artificial ...Which of the following components are part of ...Machine Learning and Deep LearningMachine Learning and Decision TreesArtificial Intelligence and Machine LearningArtificial Intelligence and Deep LearningCArtificial Intelligence is the broader family ...
22Create an MCQ on the aim of Machine LearningWhat is the aim of Machine Learning?To increase chances of successTo increase accuracyTo improve system efficiencyTo analyze data and provide outputBThe aim of Machine Learning is to increase acc...
23Create an MCQ on the aim of Deep LearningWhat is the aim of Deep Learning?To increase chances of successTo increase accuracyTo improve system efficiencyTo analyze data and provide outputAThe aim of Deep Learning is to increase chance...
24Create an MCQ on the difference between AI, Ma...Which of the following best describes the diff...AI is a subset of Machine Learning, which is a...Machine Learning is a subset of AI, which is a...Deep Learning is a subset of AI, which is a su...AI, Machine Learning, and Deep Learning are co...BAI is the broader concept that encompasses the...
25Create an MCQ on the application of AI in spee...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...CSpeech recognition is an example of AI applica...
26Create an MCQ on the application of AI in pers...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...APersonalized recommendations, as an AI applica...
27Create an MCQ on the application of AI in pred...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Analyzing sensor data to predict equipment fai...Recognizing and classifying images and speechCAI-powered predictive maintenance systems anal...
28Create an MCQ on the application of AI in medi...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...BAI-powered medical diagnosis systems analyze m...
29Create an MCQ on the difference between AI, ML...Which of the following statements accurately d...AI, ML, and DL are interchangeable terms that ...AI focuses on creating intelligent machines, M...AI is a subset of ML that uses neural networks...AI focuses on developing algorithms that enabl...BAI, ML, and DL are related but distinct concep...
30Create an MCQ on the responsibilities of an AI...Which of the following is a key responsibility...Design and development of AI algorithmsAnalysis and interpretation of dataTraining and evaluation of ML modelsDeployment and maintenance of DL modelsAOne of the key responsibilities of an AI Engin...
31Create an MCQ on the skills required for a Mac...Which of the following skills is essential for...Strong background in computer science, mathema...Experience in developing AI algorithms and sol...Familiarity with programming languages such as...All of the aboveDA Machine Learning Engineer should have a stro...
32Create an MCQ on the tasks of a Deep Learning ...Which of the following is a key task of a Deep...Design and development of DL algorithmsAnalysis and interpretation of dataTraining and evaluation of ML modelsDeployment and maintenance of AI modelsAOne of the key tasks of a Deep Learning Engine...
33Create an MCQ on the difference between ML and DLWhat distinguishes Deep Learning (DL) from Mac...DL is a more advanced form of ML that can perf...DL focuses on developing algorithms that enabl...DL is a subset of ML that uses neural networks...ML is a more advanced form of DL that can perf...CDeep Learning (DL) is a subset of Machine Lear...
34Create an MCQ on the advantages of Artificial ...Which of the following is an advantage of Arti...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...Dependence on hardware for functioningHigh speed of processingAOne of the advantages of Artificial Neural Net...
35Create an MCQ on the disadvantages of Artifici...Which of the following is a disadvantage of Ar...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...Dependence on hardware for functioningThe simplest architecture makes it difficult t...DOne of the disadvantages of Artificial Neural ...
36Create an MCQ on the advantages of Biological ...Which of the following is an advantage of Biol...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...No controlling mechanismAbility to process highly complex parallel inputsDOne of the advantages of Biological Neural Net...
37Create an MCQ on the disadvantages of Biologic...Which of the following is a disadvantage of Bi...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...No controlling mechanismSpeed of processing is slowDOne of the disadvantages of Biological Neural ...
38Create an MCQ on the differences between Artif...Which of the following is a difference between...Both ANNs and BNNs have complex and diverse ne...ANNs have fixed connections between neurons, w...Both ANNs and BNNs have simple and predetermin...ANNs and BNNs have the same processing speedBOne of the differences between Artificial Neur...
39Create an MCQ on hyperparameter tuning in mach...What is the purpose of hyperparameter tuning i...To adjust the weights and biases of the modelTo select the optimal values for the model's h...To preprocess the input data before training t...To evaluate the performance of the model on a ...BHyperparameter tuning is the process of select...
40Create an MCQ on the types of hyperparameters ...Which of the following is a type of hyperparam...WeightsBiasesLearning rateActivation functionCIn neural networks, the learning rate is a hyp...
\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['test']\",\n \"rows\": 41,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Create an MCQ on the difference between AI, Machine Learning, and Deep Learning\",\n \"Create an MCQ on the types of machine learning tasks that can be performed using deep learning\",\n \"Create an MCQ on the difference between machine learning and deep learning\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Which of the following best describes the difference between AI, Machine Learning, and Deep Learning?\",\n \"Which types of machine learning tasks can be performed using deep learning?\",\n \"What is a key difference between machine learning and deep learning?\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Both ANNs and BNNs have complex and diverse neurons\",\n \"The study of training machines to mimic human behavior\",\n \"Strong background in computer science, mathematics, and statistics\"\n ],\n \"num_unique_values\": 30,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"ANNs have fixed connections between neurons, while BNNs have flexible connections\",\n \"Robotics\",\n \"Analysis and interpretation of data\"\n ],\n \"num_unique_values\": 30,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Both ANNs and BNNs have simple and predetermined neural pathways\",\n \"Control systems\",\n \"Training and evaluation of ML models\"\n ],\n \"num_unique_values\": 32,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"ANNs and BNNs have the same processing speed\",\n \"Machine Learning requires human intervention, while Deep Learning does not\",\n \"ML is a more advanced form of DL that can perform complex tasks.\"\n ],\n \"num_unique_values\": 32,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"A\",\n \"D\",\n \"B\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"AI is the broader concept that encompasses the development of computer systems that can perform tasks requiring human intelligence. Machine Learning is a subset of AI, focused on algorithms that can learn from data and make predictions or decisions. Deep Learning is a subset of Machine Learning, specifically using neural networks with multiple layers to learn and represent complex patterns.\",\n \"Deep learning can be used for supervised, unsupervised, as well as reinforcement machine learning tasks. It provides a versatile approach to process and learn from data in various learning scenarios.\",\n \"A key difference between machine learning and deep learning is the amount of data required. Machine learning can work with a smaller amount of data, while deep learning requires a larger volume of data to train the complex neural network architectures effectively.\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":20}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['val'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"03oFFWL4hM4S","executionInfo":{"status":"ok","timestamp":1708322802115,"user_tz":-480,"elapsed":16,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"284e74e0-5907-4982-cd82-31dfc59696e3"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the impact of learning rate o... \n","1 Create an MCQ on the impact of number of epoch... \n","2 Create an MCQ on the impact of architecture on... \n","3 Create an MCQ on the strategies for hyperparam... \n","4 Create an MCQ on the drawback of GridSearchCV \n","5 Create an MCQ on the strategy that selects val... \n","6 Create an MCQ on the advantage of RandomizedSe... \n","7 Create an MCQ on the strategy that treats hype... \n","8 Create an MCQ on the challenges in deep learning \n","9 Create an MCQ on the advantages of deep learning \n","10 Create an MCQ on the disadvantages of deep lea... \n","11 Create an MCQ on the challenges in interpretin... \n","12 Create an MCQ on the risk of overfitting in de... \n","13 Create an MCQ on the application of machine le... \n","14 Create an MCQ on the use of machine learning i... \n","15 Create an MCQ on the role of machine learning ... \n","16 Create an MCQ on the application of machine le... \n","17 Create an MCQ on the use of machine learning i... \n","18 Create an MCQ on the regularization parameter ... \n","19 Create an MCQ on the kernel function in SVMs \n","20 Create an MCQ on the parameter that controls t... \n","21 Create an MCQ on the learning rate hyperparame... \n","22 Create an MCQ on the max_depth hyperparameter ... \n","23 Create an MCQ on the structure of artificial n... \n","24 Create an MCQ on the training process of artif... \n","25 Create an MCQ on the differences between machi... \n","26 Create an MCQ on the complexity of interpretin... \n","27 Create an MCQ on the computing power requireme... \n","28 Create an MCQ on the definition of deep learning \n","29 Create an MCQ on the key characteristic of dee... \n","30 Create an MCQ on the success of deep learning ... \n","31 Create an MCQ on the requirements for training... \n","32 Create an MCQ on the types of machine learning... \n","33 Create an MCQ on the types of neural networks ... \n","34 Create an MCQ on the applications of deep lear... \n","35 Create an MCQ on the applications of deep lear... \n","36 Create an MCQ on the applications of deep lear... \n","37 Create an MCQ on the main purpose of deep lear... \n","38 Create an MCQ on the definition of Artificial ... \n","39 Create an MCQ on the definition of Machine Lea... \n","40 Create an MCQ on the definition of Deep Learning. \n","41 Create an MCQ on the relationship between Arti... \n","42 Create an MCQ on the aim of Deep Learning. \n","\n"," Question \\\n","0 What impact does the learning rate hyperparame... \n","1 How does the number of epochs hyperparameter a... \n","2 How does the architecture of a neural network ... \n","3 Which of the following strategies is considere... \n","4 What is a drawback of using GridSearchCV for h... \n","5 Which hyperparameter tuning strategy selects v... \n","6 What is an advantage of using RandomizedSearch... \n","7 Which hyperparameter tuning strategy treats th... \n","8 What is one of the challenges in deep learning? \n","9 What is one of the advantages of deep learning? \n","10 What is one of the disadvantages of deep learn... \n","11 What is one of the challenges in interpreting ... \n","12 What is the risk associated with overfitting i... \n","13 In which of the following applications are mac... \n","14 Which of the following applications utilize ma... \n","15 Which of the following applications involve th... \n","16 Which of the following applications utilize ma... \n","17 In which of the following applications are mac... \n","18 What is the role of the regularization paramet... \n","19 What is the purpose of the kernel function in ... \n","20 Which parameter controls the influence of supp... \n","21 What does the learning rate hyperparameter det... \n","22 What does the max_depth hyperparameter determi... \n","23 Which layer of an artificial neural network re... \n","24 What is adjusted during the training process o... \n","25 Which of the following requires a larger volum... \n","26 Which of the following is true regarding the i... \n","27 Which of the following requires a high-perform... \n","28 What is deep learning? \n","29 What is the key characteristic of deep learning? \n","30 In which fields has deep learning achieved sig... \n","31 What are the requirements for training deep ne... \n","32 Which types of machine learning are used in de... \n","33 Which type of neural network is specifically d... \n","34 What is one of the main applications of deep l... \n","35 What is one of the main applications of deep l... \n","36 What is one of the main applications of deep l... \n","37 What is the main purpose of deep learning mode... \n","38 What is the definition of Artificial Intellige... \n","39 What is the definition of Machine Learning? \n","40 What is the definition of Deep Learning? \n","41 What is the relationship between Artificial In... \n","42 What is the aim of Deep Learning? \n","\n"," A \\\n","0 It determines the number of epochs needed for ... \n","1 Increasing the number of epochs always improve... \n","2 The architecture determines the learning rate ... \n","3 GridSearchCV \n","4 It is computationally expensive \n","5 GridSearchCV \n","6 It is computationally faster \n","7 GridSearchCV \n","8 Limited computational resources \n","9 Low accuracy \n","10 Low computational requirements \n","11 Easy interpretability \n","12 Improved performance on new data \n","13 Self-driving cars \n","14 Virtual assistants like Siri and Alexa \n","15 Chatbots \n","16 E-commerce sites \n","17 Social media monitoring \n","18 To control the trade-off between the margin an... \n","19 To control the trade-off between the margin an... \n","20 Regularization parameter (C) \n","21 The step size taken by the optimizer during ea... \n","22 The step size taken by the optimizer during ea... \n","23 Output layer \n","24 Weights \n","25 Machine learning \n","26 Machine learning results are easy to interpret \n","27 Machine learning \n","28 A branch of machine learning that uses artific... \n","29 The use of deep neural networks with multiple ... \n","30 Image recognition, natural language processing... \n","31 A large amount of data and computational resou... \n","32 Supervised, unsupervised, and reinforcement le... \n","33 Feedforward Neural Networks (FNNs) \n","34 Speech recognition \n","35 Object detection and recognition \n","36 Sentiment analysis \n","37 To analyze the sentiment of text \n","38 The study of training machines to mimic human ... \n","39 The study of training machines to mimic human ... \n","40 The study of training machines to mimic human ... \n","41 AI is a subset of ML \n","42 To increase chances of success \n","\n"," B \\\n","0 It controls the step size taken by the optimiz... \n","1 Increasing the number of epochs can lead to ov... \n","2 The architecture controls the step size taken ... \n","3 RandomizedSearchCV \n","4 It requires expert knowledge \n","5 RandomizedSearchCV \n","6 It guarantees optimal performance \n","7 RandomizedSearchCV \n","8 Easy interpretability of results \n","9 Manual feature engineering \n","10 Small amount of labeled data \n","11 Clear decision-making process \n","12 No impact on model performance \n","13 Security systems \n","14 Call centers \n","15 Virtual assistants \n","16 Streaming services \n","17 Sentiment analysis systems \n","18 To define the similarity between data points \n","19 To define the similarity between data points \n","20 Kernel function \n","21 The number of boosting trees to be trained \n","22 The number of boosting trees to be trained \n","23 Hidden layer \n","24 Layers \n","25 Deep learning \n","26 Deep learning results are easy to interpret \n","27 Deep learning \n","28 A technique in machine learning that involves ... \n","29 The use of decision trees for modeling \n","30 Clustering, dimensionality reduction, and anom... \n","31 The availability of cloud computing and specia... \n","32 Supervised and unsupervised learning \n","33 Convolutional Neural Networks (CNNs) \n","34 Sentiment analysis \n","35 Image classification \n","36 Image segmentation \n","37 To translate text from one language to another \n","38 The study of improving machines with experienc... \n","39 The study of improving machines with experienc... \n","40 The study of improving machines with experienc... \n","41 ML is a subset of AI \n","42 To increase accuracy \n","\n"," C \\\n","0 It determines the depth of the neural network \n","1 The number of epochs does not have any impact ... \n","2 The architecture determines the depth and widt... \n","3 Bayesian Optimization \n","4 It may result in overfitting \n","5 Bayesian Optimization \n","6 It requires less expertise \n","7 Bayesian Optimization \n","8 Small amount of training data \n","9 Limited scalability \n","10 Easy interpretability \n","11 Limited complexity \n","12 Poor performance on new data \n","13 Medical imaging \n","14 Speech recognition systems \n","15 NLP systems \n","16 Recommendation systems \n","17 Spam filters \n","18 To determine the influence of support vectors ... \n","19 To determine the influence of support vectors ... \n","20 Gamma \n","21 The maximum depth of each tree in the ensemble \n","22 The maximum depth of each tree in the ensemble \n","23 Input layer \n","24 Neurons \n","25 Both require the same volume of dataset \n","26 Both machine learning and deep learning result... \n","27 Both require the same computing power \n","28 A type of unsupervised machine learning that c... \n","29 The use of reinforcement learning algorithms \n","30 Supervised machine learning tasks like image c... \n","31 Manual feature engineering \n","32 Unsupervised and reinforcement learning \n","33 Recurrent Neural Networks (RNNs) \n","34 Image classification \n","35 Language translation \n","36 Game playing \n","37 To identify and understand visual data \n","38 The study that uses neural networks to imitate... \n","39 The study that uses neural networks to imitate... \n","40 The study that uses neural networks to imitate... \n","41 AI and ML are independent of each other \n","42 To achieve high accuracy with a small amount o... \n","\n"," D Correct Answer \\\n","0 It controls the width of the neural network B \n","1 The number of epochs determines the learning r... B \n","2 The architecture affects the activation functi... C \n","3 None of the above A \n","4 It is not effective for high-dimensional hyper... A \n","5 None of the above B \n","6 It is more effective for high-dimensional hype... A \n","7 None of the above C \n","8 No risk of overfitting A \n","9 Continual improvement D \n","10 No risk of overfitting B \n","11 Black box nature D \n","12 Reduced computational requirements C \n","13 All of the above D \n","14 All of the above D \n","15 All of the above D \n","16 All of the above D \n","17 All of the above D \n","18 To determine the maximum depth of each tree in... A \n","19 To determine the maximum depth of each tree in... B \n","20 Learning rate C \n","21 The minimum sum of instance weight needed in a... A \n","22 The minimum sum of instance weight needed in a... C \n","23 Final layer C \n","24 Connections A \n","25 Neither requires a dataset B \n","26 Deep learning results are difficult to interpret D \n","27 Neither requires computing power B \n","28 A method in machine learning that uses decisio... A \n","29 The use of unsupervised learning techniques A \n","30 Reinforcement learning tasks like robotics and... A \n","31 Supervised labeled datasets A \n","32 Supervised and reinforcement learning A \n","33 Artificial Neural Networks (ANNs) B \n","34 Game playing C \n","35 Game playing C \n","36 Speech recognition C \n","37 To control complex systems C \n","38 The study that focuses on learning, reasoning,... A \n","39 The study that focuses on learning from data a... D \n","40 The study that focuses on using neural network... D \n","41 AI and ML are the same thing B \n","42 To achieve high accuracy with a large amount o... D \n","\n"," Explanation \n","0 The learning rate hyperparameter controls the ... \n","1 Increasing the number of epochs can improve th... \n","2 The architecture of a neural network determine... \n","3 GridSearchCV is considered a 'brute force' app... \n","4 GridSearchCV is computationally expensive as i... \n","5 RandomizedSearchCV selects values at random fo... \n","6 RandomizedSearchCV is computationally faster t... \n","7 Bayesian Optimization treats the search for op... \n","8 One of the challenges in deep learning is the ... \n","9 One of the advantages of deep learning is its ... \n","10 One of the disadvantages of deep learning is t... \n","11 One of the challenges in interpreting deep lea... \n","12 The risk associated with overfitting in deep l... \n","13 Machine learning algorithms are used in image ... \n","14 Machine learning algorithms are used in speech... \n","15 Machine learning algorithms are used in NLP sy... \n","16 Machine learning algorithms are used in recomm... \n","17 Machine learning algorithms are used in sentim... \n","18 The regularization parameter (C) in SVMs contr... \n","19 The kernel function in SVMs defines the simila... \n","20 The parameter Gamma controls the influence of ... \n","21 The learning rate hyperparameter in XGBoost de... \n","22 The max_depth hyperparameter in XGBoost determ... \n","23 The input layer of an artificial neural networ... \n","24 During the training process of an artificial n... \n","25 Deep learning requires a larger volume of data... \n","26 Deep learning results are more complex and dif... \n","27 Deep learning requires a high-performance comp... \n","28 Deep learning is a branch of machine learning ... \n","29 The key characteristic of deep learning is the... \n","30 Deep learning has achieved significant success... \n","31 Training deep neural networks typically requir... \n","32 Deep learning can be used for supervised, unsu... \n","33 Convolutional Neural Networks (CNNs) are speci... \n","34 One of the main applications of deep learning ... \n","35 One of the main applications of deep learning ... \n","36 One of the main applications of deep learning ... \n","37 The main purpose of deep learning models in co... \n","38 Artificial Intelligence is the mechanism to in... \n","39 Machine Learning is the study/process which pr... \n","40 Deep Learning is a sub-part of the broader fam... \n","41 Artificial Intelligence is the broader family ... \n","42 The aim of Deep Learning is to achieve high ac... "],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the impact of learning rate o...What impact does the learning rate hyperparame...It determines the number of epochs needed for ...It controls the step size taken by the optimiz...It determines the depth of the neural networkIt controls the width of the neural networkBThe learning rate hyperparameter controls the ...
1Create an MCQ on the impact of number of epoch...How does the number of epochs hyperparameter a...Increasing the number of epochs always improve...Increasing the number of epochs can lead to ov...The number of epochs does not have any impact ...The number of epochs determines the learning r...BIncreasing the number of epochs can improve th...
2Create an MCQ on the impact of architecture on...How does the architecture of a neural network ...The architecture determines the learning rate ...The architecture controls the step size taken ...The architecture determines the depth and widt...The architecture affects the activation functi...CThe architecture of a neural network determine...
3Create an MCQ on the strategies for hyperparam...Which of the following strategies is considere...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveAGridSearchCV is considered a 'brute force' app...
4Create an MCQ on the drawback of GridSearchCVWhat is a drawback of using GridSearchCV for h...It is computationally expensiveIt requires expert knowledgeIt may result in overfittingIt is not effective for high-dimensional hyper...AGridSearchCV is computationally expensive as i...
5Create an MCQ on the strategy that selects val...Which hyperparameter tuning strategy selects v...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveBRandomizedSearchCV selects values at random fo...
6Create an MCQ on the advantage of RandomizedSe...What is an advantage of using RandomizedSearch...It is computationally fasterIt guarantees optimal performanceIt requires less expertiseIt is more effective for high-dimensional hype...ARandomizedSearchCV is computationally faster t...
7Create an MCQ on the strategy that treats hype...Which hyperparameter tuning strategy treats th...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveCBayesian Optimization treats the search for op...
8Create an MCQ on the challenges in deep learningWhat is one of the challenges in deep learning?Limited computational resourcesEasy interpretability of resultsSmall amount of training dataNo risk of overfittingAOne of the challenges in deep learning is the ...
9Create an MCQ on the advantages of deep learningWhat is one of the advantages of deep learning?Low accuracyManual feature engineeringLimited scalabilityContinual improvementDOne of the advantages of deep learning is its ...
10Create an MCQ on the disadvantages of deep lea...What is one of the disadvantages of deep learn...Low computational requirementsSmall amount of labeled dataEasy interpretabilityNo risk of overfittingBOne of the disadvantages of deep learning is t...
11Create an MCQ on the challenges in interpretin...What is one of the challenges in interpreting ...Easy interpretabilityClear decision-making processLimited complexityBlack box natureDOne of the challenges in interpreting deep lea...
12Create an MCQ on the risk of overfitting in de...What is the risk associated with overfitting i...Improved performance on new dataNo impact on model performancePoor performance on new dataReduced computational requirementsCThe risk associated with overfitting in deep l...
13Create an MCQ on the application of machine le...In which of the following applications are mac...Self-driving carsSecurity systemsMedical imagingAll of the aboveDMachine learning algorithms are used in image ...
14Create an MCQ on the use of machine learning i...Which of the following applications utilize ma...Virtual assistants like Siri and AlexaCall centersSpeech recognition systemsAll of the aboveDMachine learning algorithms are used in speech...
15Create an MCQ on the role of machine learning ...Which of the following applications involve th...ChatbotsVirtual assistantsNLP systemsAll of the aboveDMachine learning algorithms are used in NLP sy...
16Create an MCQ on the application of machine le...Which of the following applications utilize ma...E-commerce sitesStreaming servicesRecommendation systemsAll of the aboveDMachine learning algorithms are used in recomm...
17Create an MCQ on the use of machine learning i...In which of the following applications are mac...Social media monitoringSentiment analysis systemsSpam filtersAll of the aboveDMachine learning algorithms are used in sentim...
18Create an MCQ on the regularization parameter ...What is the role of the regularization paramet...To control the trade-off between the margin an...To define the similarity between data pointsTo determine the influence of support vectors ...To determine the maximum depth of each tree in...AThe regularization parameter (C) in SVMs contr...
19Create an MCQ on the kernel function in SVMsWhat is the purpose of the kernel function in ...To control the trade-off between the margin an...To define the similarity between data pointsTo determine the influence of support vectors ...To determine the maximum depth of each tree in...BThe kernel function in SVMs defines the simila...
20Create an MCQ on the parameter that controls t...Which parameter controls the influence of supp...Regularization parameter (C)Kernel functionGammaLearning rateCThe parameter Gamma controls the influence of ...
21Create an MCQ on the learning rate hyperparame...What does the learning rate hyperparameter det...The step size taken by the optimizer during ea...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleThe minimum sum of instance weight needed in a...AThe learning rate hyperparameter in XGBoost de...
22Create an MCQ on the max_depth hyperparameter ...What does the max_depth hyperparameter determi...The step size taken by the optimizer during ea...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleThe minimum sum of instance weight needed in a...CThe max_depth hyperparameter in XGBoost determ...
23Create an MCQ on the structure of artificial n...Which layer of an artificial neural network re...Output layerHidden layerInput layerFinal layerCThe input layer of an artificial neural networ...
24Create an MCQ on the training process of artif...What is adjusted during the training process o...WeightsLayersNeuronsConnectionsADuring the training process of an artificial n...
25Create an MCQ on the differences between machi...Which of the following requires a larger volum...Machine learningDeep learningBoth require the same volume of datasetNeither requires a datasetBDeep learning requires a larger volume of data...
26Create an MCQ on the complexity of interpretin...Which of the following is true regarding the i...Machine learning results are easy to interpretDeep learning results are easy to interpretBoth machine learning and deep learning result...Deep learning results are difficult to interpretDDeep learning results are more complex and dif...
27Create an MCQ on the computing power requireme...Which of the following requires a high-perform...Machine learningDeep learningBoth require the same computing powerNeither requires computing powerBDeep learning requires a high-performance comp...
28Create an MCQ on the definition of deep learningWhat is deep learning?A branch of machine learning that uses artific...A technique in machine learning that involves ...A type of unsupervised machine learning that c...A method in machine learning that uses decisio...ADeep learning is a branch of machine learning ...
29Create an MCQ on the key characteristic of dee...What is the key characteristic of deep learning?The use of deep neural networks with multiple ...The use of decision trees for modelingThe use of reinforcement learning algorithmsThe use of unsupervised learning techniquesAThe key characteristic of deep learning is the...
30Create an MCQ on the success of deep learning ...In which fields has deep learning achieved sig...Image recognition, natural language processing...Clustering, dimensionality reduction, and anom...Supervised machine learning tasks like image c...Reinforcement learning tasks like robotics and...ADeep learning has achieved significant success...
31Create an MCQ on the requirements for training...What are the requirements for training deep ne...A large amount of data and computational resou...The availability of cloud computing and specia...Manual feature engineeringSupervised labeled datasetsATraining deep neural networks typically requir...
32Create an MCQ on the types of machine learning...Which types of machine learning are used in de...Supervised, unsupervised, and reinforcement le...Supervised and unsupervised learningUnsupervised and reinforcement learningSupervised and reinforcement learningADeep learning can be used for supervised, unsu...
33Create an MCQ on the types of neural networks ...Which type of neural network is specifically d...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)Artificial Neural Networks (ANNs)BConvolutional Neural Networks (CNNs) are speci...
34Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Speech recognitionSentiment analysisImage classificationGame playingCOne of the main applications of deep learning ...
35Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Object detection and recognitionImage classificationLanguage translationGame playingCOne of the main applications of deep learning ...
36Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Sentiment analysisImage segmentationGame playingSpeech recognitionCOne of the main applications of deep learning ...
37Create an MCQ on the main purpose of deep lear...What is the main purpose of deep learning mode...To analyze the sentiment of textTo translate text from one language to anotherTo identify and understand visual dataTo control complex systemsCThe main purpose of deep learning models in co...
38Create an MCQ on the definition of Artificial ...What is the definition of Artificial Intellige...The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on learning, reasoning,...AArtificial Intelligence is the mechanism to in...
39Create an MCQ on the definition of Machine Lea...What is the definition of Machine Learning?The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on learning from data a...DMachine Learning is the study/process which pr...
40Create an MCQ on the definition of Deep Learning.What is the definition of Deep Learning?The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on using neural network...DDeep Learning is a sub-part of the broader fam...
41Create an MCQ on the relationship between Arti...What is the relationship between Artificial In...AI is a subset of MLML is a subset of AIAI and ML are independent of each otherAI and ML are the same thingBArtificial Intelligence is the broader family ...
42Create an MCQ on the aim of Deep Learning.What is the aim of Deep Learning?To increase chances of successTo increase accuracyTo achieve high accuracy with a small amount o...To achieve high accuracy with a large amount o...DThe aim of Deep Learning is to achieve high ac...
\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['val']\",\n \"rows\": 43,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Create an MCQ on the main purpose of deep learning models in computer vision.\",\n \"Create an MCQ on the training process of artificial neural networks\",\n \"Create an MCQ on the differences between machine learning and deep learning\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"What is the main purpose of deep learning models in computer vision?\",\n \"What is adjusted during the training process of an artificial neural network?\",\n \"Which of the following requires a larger volume of dataset compared to machine learning?\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"To increase chances of success\",\n \"Chatbots\",\n \"A large amount of data and computational resources\"\n ],\n \"num_unique_values\": 36,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"To increase accuracy\",\n \"Virtual assistants\",\n \"The availability of cloud computing and specialized hardware\"\n ],\n \"num_unique_values\": 36,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Gamma\",\n \"NLP systems\",\n \"It may result in overfitting\"\n ],\n \"num_unique_values\": 37,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"AI and ML are the same thing\",\n \"Connections\",\n \"Speech recognition\"\n ],\n \"num_unique_values\": 33,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"C\",\n \"D\",\n \"B\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"The main purpose of deep learning models in computer vision is to identify and understand visual data. Deep learning models can be used to perform tasks such as object detection and recognition, image classification, and image segmentation.\",\n \"During the training process of an artificial neural network, the weights of the connections between neurons are adjusted to enhance the performance of the model.\",\n \"Deep learning requires a larger volume of dataset compared to machine learning.\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":21}]},{"cell_type":"code","source":["from huggingface_hub import notebook_login\n","notebook_login()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":145,"referenced_widgets":["ac3383a4853c484ca009cddb6e305853","095e9de8c9c24583baa8c7d074c5eacc","7a6eab206dcb4cd3b0cd4681a38f192f","e40bf3020d024dc0bc72a6f23b471494","8fc0cb635ac64733b957061f58ed16b6","9dcbf15879f94eddaf99a1603d1ca4c8","b0cf0967450c415f92e8540b30944ccf","e588245da98b412d849fbe2f94fb4b79","58409bb2b4c845f589fce0e2c2078a8c","17a1cd2fe40e4dde8b6b196e29abbec1","10bef54de0f64e86a3c9b1c039885ee9","40c244e062aa47a7846cace18c952cad","0d5efa7a7e7a4fe2ac2a789fe79ea94a","51e3dfdb61374128a65a208209c93060","283deaf2c9544e80a9433ae9148492d5","b9eb5c03173143aeb2793da889a2428b","8a3773ae30cf482a80edc5af98dc9cd8","0e34328104b64acb9a63e76013ccfc0d","e4fb41fce6b44fffbb1e0a6ab9d477ab","b5b3620259184090b78f008fc8a789db","bdec133061344a838b792668b98a5daa","f26867efbb384d1791ccc7e6ff2f7b2b","e35c557212ea4317ba624707f8d42dde","bfee25db776447eb9126cdbdbdfe6a8c","82764a4acf0e49e58547e9f2e52d5533","ab32080c5ec54c8abb90f0d85855adf5","5eacd5686b6d4486b772e957a80d894d","6c6af529b3dd449cb2826706c220d0cc","079da2bd7cfd4ed3b3324a2e54a8ae87","570cfe79345a43b5a9b34f05e72cc6c9","b21ff34c2fd844f2baee7dd45eda5700","8f7bf4ff652b4ae2b137b0518a4d6a00"]},"id":"U2YZULfoCura","executionInfo":{"status":"ok","timestamp":1708322802639,"user_tz":-480,"elapsed":9,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"31a4d102-5e31-4500-c9bc-34d0274857cf"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["VBox(children=(HTML(value='
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":["## 3. Filter out rows with more than 2048 tokens\n","\n","We will remove samples with more than 2048 tokens (max context size of Llama 2 by default = 4096)."],"metadata":{"id":"_RXe958fNLwH"}},{"cell_type":"code","source":["def filter_by_token_count(dataset_split, combined_token_counts, max_tokens=2048):\n"," # Filter out rows with more than 'max_tokens' tokens\n"," filtered_dataset = [example for example, count in zip(dataset_split, combined_token_counts) if count <= max_tokens]\n"," return filtered_dataset\n","\n","\n","# Assuming 'dataset' contains your data splits\n","fig, axs = plt.subplots(3, 5, figsize=(25, 15)) # Adjust figure size as necessary\n","\n","split_names = ['train', 'test', 'val']\n","for row, split_name in enumerate(split_names):\n"," # Tokenize and count\n"," instruction_counts, explanation_counts, question_counts, options_counts, combined_counts = tokenize_and_count(dataset[split_name])\n","\n"," # Filter dataset based on combined token count\n"," filtered_dataset = filter_by_token_count(dataset[split_name], combined_counts)\n","\n"," # Re-tokenize and count for the filtered dataset\n"," filtered_instruction_counts, filtered_explanation_counts, filtered_question_counts, filtered_options_counts, filtered_combined_counts = tokenize_and_count(filtered_dataset)\n","\n"," # Plotting the distributions for the filtered datasets, organizing by row based on the split\n"," plot_distribution(filtered_instruction_counts, f\"{split_name} (filtered): Instruction\", axs[row, 0])\n"," plot_distribution(filtered_explanation_counts, f\"{split_name} (filtered): Explanation\", axs[row, 1])\n"," plot_distribution(filtered_question_counts, f\"{split_name} (filtered): Question\", axs[row, 2])\n"," plot_distribution(filtered_options_counts, f\"{split_name} (filtered): Options\", axs[row, 3])\n"," plot_distribution(filtered_combined_counts, f\"{split_name} (filtered): Combined\", axs[row, 4])\n","\n","# Adjust layout to prevent overlap and ensure clarity\n","plt.tight_layout(pad=3.0)\n","plt.show()\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":898},"id":"WYzaKhSvz9Yk","executionInfo":{"status":"ok","timestamp":1708322857920,"user_tz":-480,"elapsed":8192,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"ee8da4e6-3afb-4534-885b-9780571cde25"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Function to filter dataset and plot new distribution\n","def filter_and_plot(dataset_split_name, dataset_split_data, combined_token_counts, axs, position):\n"," # Filter out rows with more than 2048 tokens\n"," valid_indices = [i for i, count in enumerate(combined_token_counts) if count <= 2048]\n"," print(f\"Number of valid rows in {dataset_split_name}: {len(valid_indices)}\")\n"," print(f\"Removing {len(dataset_split_data) - len(valid_indices)} rows from {dataset_split_name}...\")\n","\n"," # Extract valid rows based on indices\n"," valid_dataset = [dataset_split_data[i] for i in valid_indices]\n","\n"," # Re-calculate token counts for the valid dataset if necessary\n"," # This step is assumed necessary only if the token counts need to be recalculated for the filtered dataset\n"," # Otherwise, valid_token_counts = [combined_token_counts[i] for i in valid_indices] would suffice\n"," _, _, _, _, valid_combined_counts = tokenize_and_count(valid_dataset)\n","\n"," # Plot the new distribution for valid rows\n"," plot_distribution(valid_combined_counts, f\"New distribution after filtering {dataset_split_name}\", axs[position])\n","\n","# Create a figure with subplots\n","fig, axs = plt.subplots(3, 1, figsize=(6, 9)) # Adjust figsize as necessary\n","\n","# Assuming the 'dataset' variable is a dictionary containing data splits 'train', 'test', and 'val'\n","for i, split_name in enumerate(['train', 'test', 'val']):\n"," # Tokenize and count for the specific dataset split\n"," _, _, _, _, combined_counts = tokenize_and_count(dataset[split_name])\n","\n"," # Filter datasets based on token count and plot the new distribution\n"," filter_and_plot(split_name, dataset[split_name], combined_counts, axs, i)\n","\n","plt.tight_layout()\n","plt.show()\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"LTw4ITHDl92J","executionInfo":{"status":"ok","timestamp":1708322864043,"user_tz":-480,"elapsed":2649,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"9f5de7d2-6e9a-4e2e-d3ee-73df3a6ad59f"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Number of valid rows in train: 334\n","Removing 0 rows from train...\n","Number of valid rows in test: 41\n","Removing 0 rows from test...\n","Number of valid rows in val: 43\n","Removing 0 rows from val...\n"]},{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Initialize a flag to indicate whether any entries were removed in any split\n","entries_removed = False\n","\n","# Iterate over each split in the dataset\n","for split_name in ['train', 'test', 'val']:\n"," # Get the original length of the split\n"," original_length = len(dataset[split_name])\n"," # Tokenize and count tokens in the split\n"," _, _, _, _, combined_counts = tokenize_and_count(dataset[split_name])\n"," # Determine valid indices (entries with <= 2048 tokens)\n"," valid_indices = [i for i, count in enumerate(combined_counts) if count <= 2048]\n"," # Check if any entries were removed\n"," if len(valid_indices) < original_length:\n"," entries_removed = True\n"," # Update the dataset split with filtered entries\n"," dataset[split_name] = dataset[split_name].select(valid_indices)\n","\n","# Flag to control execution of subsequent code\n","continue_execution = True\n","\n","if not entries_removed:\n"," print(\"No entries removed due to token count. Skipping saving.\")\n"," continue_execution = False\n","\n","# Proceed with further steps only if entries were removed\n","if continue_execution:\n"," # Save the filtered dataset to disk\n"," dataset.save_to_disk('new_mcq_data')\n"," print(\"Filtered dataset saved successfully.\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2N8lcoTS5jkm","executionInfo":{"status":"ok","timestamp":1708322873795,"user_tz":-480,"elapsed":922,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"d1028a6f-0388-47ea-8486-e715bcbc52fd"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["No entries removed due to token count. Skipping saving.\n"]}]},{"cell_type":"markdown","source":["---\n","\n","## 4. Near-deduplication Using Embeddings\n","\n","* Near-deduplication with embeddings is a technique that employs vector representations to effectively identify and manage nearly identical data entries.\n","\n","* By transforming data into these vectors (embeddings), we can quantitatively measure how similar different pieces of data are. This transformation significantly improves our ability to manage large datasets, where sorting through and removing near-duplicates manually would be impractical.\n","\n","* Widely used in fields like database management, information retrieval, and machine learning, this approach is crucial for efficient data handling and analysis.\n","\n","---\n","\n","### We Will Not Perform Deduplication on Our MCQ Dataset.\n","\n","* **Intentional Repetition for Emphasis**: In educational contexts, certain concepts may be intentionally repeated to underscore their significance. Deduplication could diminish the dataset's educational effectiveness by removing these purposeful repetitions.\n","\n","* **Variations of Similar Questions**: MCQ datasets often feature questions that, while seemingly similar, include minor variations in wording, options, or context. Inadequately designed deduplication algorithms risk eliminating these nuances, thereby losing valuable elements of the dataset.\n","\n","* **Difficulty in Defining \"Duplicates\"**: Identifying duplicates within MCQs poses a significant challenge, as questions that appear identical might differ in subtle yet crucial ways. These distinctions often represent unique learning opportunities that would be lost through deduplication.\n","\n","---\n"],"metadata":{"id":"kXzTKu99w3g4"}},{"cell_type":"markdown","source":["## 5. Top-k sampling\n","\n","Only keep the top k samples with the most tokens.\n","\n","---\n","\n","### Decision on \"Top-k Sampling\" for Our MCQ Dataset\n","\n","\n","We have decided against employing \"Top-k sampling\" to select only the top k samples with the most tokens in our MCQ dataset. This approach does not align with the core objectives of MCQ dataset development for several critical reasons:\n","\n","\n","**Practical Considerations**\n","\n","* **Conciseness and Effectiveness**: The hallmark of high-quality MCQs lies in their conciseness and meaningfulness. Favoring question length over substance could detract from the dataset's quality, as longer questions do not necessarily equate to higher educational value. Succinct yet profound questions are typically the most beneficial and stimulating for learners.\n","\n","\n","* Given these considerations, we conclude that \"Top-k sampling,\" which prioritizes token count, falls short of fulfilling the requirements of our MCQ dataset. The true merit of a valuable MCQ dataset resides in its diverse and balanced assortment of topics and difficulty levels, not merely in question length. This philosophy ensures our dataset remains versatile and effective across various educational and machine learning applications.\n","---"],"metadata":{"id":"TOCspcgXNOav"}},{"cell_type":"code","source":["# @title\n","# Push to Hugging Face Hub\n","dataset.push_to_hub(\"ssoh/mcq_dataset_2\")"],"metadata":{"id":"pj1b5S_68KB0","colab":{"base_uri":"https://localhost:8080/","height":244,"referenced_widgets":["d245dbe5cf8946e28f08287c15926a6d","45f5b19c94a544d79b3d65e13e67a0f1","ecb708463202466b96b244822fddd7c3","9cc0c38f4fb4452e8a14f1b208aaaefd","9ec8035efb334973bf54923d9b25b491","60fc22bbcb204b008cf416f44835ab84","686e9cf6109241b28c44aba78667aaa9","bf5c401f943e4040893e01cade21b632","5bb2a8fc3af2421c9b1f92039ea61ba2","5050f71288f2478c8d5bdeff47fcdca5","551a53a84dd8459b8cdd3e3e09ae687c","5af01ff36b2d4edebc842f0aea68bbdd","243e232f096047ca84cafe43fdd02abf","aee15bf2685349e684a7c1e0847fe830","324b3953949e4416b3614572969d1124","20471d4effb54abfabce62c072f90ebe","34914f01bdad45aca133fda48cb7b77d","015ca0b73d984087b070630305d794af","df0642f7d06d4a1eaf10ddf27460f3f1","9398e6c79bd947ae8e7090363bd669a4","814c3545d598458182c9d638cd7d78c0","46cbf4315b0e4a00b41157bc4d03f0fd","00efd4b16cd24e7c92d9dd2daa03b10f","9e4171a5c49a494a9ae72aa72a67862c","7569df6a04a14194866afbb4e78f6468","39ff79fe9a214782894e983ce4d45acd","91b5f1a2be93471eb880235771722cd9","57c0c6ab34974fc89e2036bd1074a098","bb364cb3c7134948b24ba1b3a0c2186d","4108281cd53941288d99b60e994770f1","1e11face32ee40f890cb97751209d6ac","45a866686d954d33ba1b3034397661ea","f5014c5a31d342e7952bcebf024250a2","8288c75935d74e85bd62e1682f8f04fe","1981153d1d2943b1936946a362048f3f","d5aec20dc1034e57980a8e9121fdd8e0","9795e18cccce455485accc5309d8d9de","f0bea944fba44d508575eb73b8684596","f9b825f130304fccadb8d660daa0b3ad","4d18d9a7763749faacffaf660cb6c100","4ba50000d3b14f74a288aa8249723fd6","152308af6482481aa88914fb1f37eb13","2e3600f228c446349e2fe7126fb95255","57849a6643f64b58802bb57ab8e18412","426bad77baaa4db49164e3cf7b3505b9","2f6008ded0274199ba675ec6f0a7469d","7c290b50d93a4f0cb209a5e444934596","bc438acf4f2047e6b7e49829c8b078ee","9862cd6a17af4bfd80a02f9b68b80629","c949f9dd143c4ccda644e3a601a46302","b603160aed09476c937847a2c1fa6236","e3e9b4527bd84cbe9f83479d090416c0","3db20951b33c4ce3b14acd7fe1a438ce","96497d48b0bd436b964826dcbaa92a76","2fb73ee317e7482e94e7c696c6a0a2de","48cd0eb61edc469482a93d6a5cf8dabd","1ee577704a9048e39d73ad278b587215","5cd9bf0a588947a8895e187842dfca0c","cfcd77f72de043b3ad28dbaa216909d3","621a509dcda84c90b6f1e76b39b9b97a","e7ea517d5821476fa388b2d67c098675","14348abed9184177a87d6a8c0f558097","dbc3be056d4e4e7d87beb01252eecb37","0acc8312636b4361b915e9bd39a02057","f0f3a78cb34f41caba416d266c270952","f118a0a1a7b444508e6c02aa0acc83d7"]},"executionInfo":{"status":"ok","timestamp":1708322881797,"user_tz":-480,"elapsed":3008,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"266d2243-1d5a-41a4-b0e2-25a573b7f602"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["Uploading the dataset shards: 0%| | 0/1 [00:00 "]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_142852-r9yt9wy0"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run vermilion-springroll-11 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/r9yt9wy0"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"4e64b2267c9c4213a5914056f6811d1b","version_major":2,"version_minor":0},"text/plain":["Map: 0%| | 0/334 [00:00\n"," \n"," \n"," [66/66 10:49, Epoch 6/6]\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.0054002.068154
201.4362001.290664
300.7781001.072349
400.6132000.931616
500.4969000.921341
600.4680000.888872

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:04]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.006 MB uploaded\\r'), FloatProgress(value=1.0, max=1.0)))"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▁▁▁▁
eval/runtime▁▅▇▇▇█▇
eval/samples_per_second█▄▂▂▂▁▂
eval/steps_per_second█▄▂▂▁▁▂
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▄▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.88887
eval/runtime5.8859
eval/samples_per_second7.306
eval/steps_per_second1.019
eval_loss0.88887
train/epoch6.0
train/global_step66
train/learning_rate2e-05
train/loss0.468
train/total_flos2090258212601856.0
train/train_loss1.06766
train/train_runtime658.5513
train/train_samples_per_second3.043
train/train_steps_per_second0.1

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run vermilion-springroll-11 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/r9yt9wy0
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_142852-r9yt9wy0/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 14:40:09,515] Trial 0 finished with value: 0.8888720870018005 and parameters: {'learning_rate': 0.00022063199006940203, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 0 with value: 0.8888720870018005.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"3de6384b3841485d8068f7265fbda30e","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113223888807826, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_144009-p10q3kv9"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run lunar-ox-12 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/p10q3kv9"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"0a6c66cadea549f78fe5a183e841734e","version_major":2,"version_minor":0},"text/plain":["Map: 0%| | 0/334 [00:00\n"," \n"," \n"," [66/66 10:55, Epoch 6/6]\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8819001.890017
201.0036001.057150
300.6157000.908779
400.5004000.848109
500.4300000.856009
600.4028000.839791

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.023 MB uploaded\\r'), FloatProgress(value=0.2557933392427504, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁
eval/runtime▁▄▆█▆▆▅
eval/samples_per_second█▅▃▁▃▃▃
eval/steps_per_second█▅▃▁▄▃▄
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83979
eval/runtime5.9095
eval/samples_per_second7.276
eval/steps_per_second1.015
eval_loss0.83979
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.4028
train/total_flos2090258212601856.0
train/train_loss0.91779
train/train_runtime664.2846
train/train_samples_per_second3.017
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run lunar-ox-12 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/p10q3kv9
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_144009-p10q3kv9/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 14:51:33,062] Trial 1 finished with value: 0.8397907614707947 and parameters: {'learning_rate': 0.000388078354781562, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 5}. Best is trial 1 with value: 0.8397907614707947.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"9a11d1cc83a44c18a071c37ee29cb191","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113095899862755, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_145133-zcwhia3h"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run lunar-envelope-13 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/zcwhia3h"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [120/168 11:45 < 04:46, 0.17 it/s, Epoch 5/8]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8054001.864337
201.0785001.049373
300.6376000.906046
400.5493000.883616
500.4913000.882419
600.4567000.849847
700.4312000.859990
800.4107000.856455
900.3987000.832274
1000.3759000.863837
1100.3876000.843072
1200.3749000.846136

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17044687077892842, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▂▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▃▄▅▃▄▃▃▅█▅▅▄
eval/samples_per_second█▆▅▄▆▅▆▆▄▁▄▄▅
eval/steps_per_second█▆▅▄▅▅▆▆▄▁▄▄▄
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83227
eval/runtime5.8969
eval/samples_per_second7.292
eval/steps_per_second1.017
eval_loss0.83227
train/epoch5.71
train/global_step120
train/learning_rate0.00011
train/loss0.3749
train/total_flos1821144316674048.0
train/train_loss0.69982
train/train_runtime710.1571
train/train_samples_per_second3.763
train/train_steps_per_second0.237

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run lunar-envelope-13 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/zcwhia3h
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_145133-zcwhia3h/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:03:40,491] Trial 2 finished with value: 0.8322736024856567 and parameters: {'learning_rate': 0.00038013816677024434, 'num_train_epochs': 8, 'per_device_train_batch_size': 16, 'warmup_steps': 4}. Best is trial 2 with value: 0.8322736024856567.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"192adca007164c1188758c608dfd03f7","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.01111326985539765, max=1.0)…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_150340-0rux0mbt"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run dazzling-orchid-14 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/0rux0mbt"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [84/84 08:09, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.1171002.101251
201.5435001.223637
300.7658001.053006
400.6335000.956405
500.5513000.914150
600.4887000.883170
700.4532000.863325
800.4317000.861434

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.022 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.6507556781402156, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▂▁▁▁▁▁
eval/runtime▁▄▃▅█▆▅▇▅
eval/samples_per_second█▅▆▄▁▃▄▂▄
eval/steps_per_second█▅▆▅▁▃▅▂▄
eval_loss▁
train/epoch▁▁▂▂▃▃▄▄▅▅▆▆▇▇████
train/global_step▁▁▂▂▃▃▄▄▅▅▆▆▇▇█████
train/learning_rate█▇▆▅▄▃▂▁
train/loss█▄▂▂▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.86143
eval/runtime5.9003
eval/samples_per_second7.288
eval/steps_per_second1.017
eval_loss0.86143
train/epoch4.0
train/global_step84
train/learning_rate1e-05
train/loss0.4317
train/total_flos1274022822739968.0
train/train_loss0.96978
train/train_runtime494.7488
train/train_samples_per_second2.7
train/train_steps_per_second0.17

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run dazzling-orchid-14 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/0rux0mbt
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_150340-0rux0mbt/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:12:11,748] Trial 3 finished with value: 0.8614340424537659 and parameters: {'learning_rate': 0.00023956952379873406, 'num_train_epochs': 4, 'per_device_train_batch_size': 16, 'warmup_steps': 5}. Best is trial 2 with value: 0.8322736024856567.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"f7b8197a4e4c4242a7965774efc249e1","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113144411097488, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_151211-6nazgql4"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run prosperous-dragon-15 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/6nazgql4"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [66/66 10:56, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.6801001.782772
200.8941001.025657
300.5874000.897337
400.4763000.830805
500.4204000.864126
600.3949000.841267

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.1705031517334534, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁
eval/runtime▁▇▆▆█▇▅
eval/samples_per_second█▂▄▄▁▂▄
eval/steps_per_second█▁▃▃▁▁▃
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83081
eval/runtime5.9104
eval/samples_per_second7.275
eval/steps_per_second1.015
eval_loss0.83081
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.3949
train/total_flos2090258212601856.0
train/train_loss0.85955
train/train_runtime665.5241
train/train_samples_per_second3.011
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run prosperous-dragon-15 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/6nazgql4
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_151211-6nazgql4/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:23:34,430] Trial 4 finished with value: 0.8308054208755493 and parameters: {'learning_rate': 0.00041915607985727055, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d8dc9ac41d544e3fb4094e8c4b80cecc","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113028755709011, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_152334-gqh7mnqx"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run crimson-dragon-16 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/gqh7mnqx"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [44/44 07:13, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.6064001.642761
200.8001001.017194
300.5857000.903642
400.4734000.855594

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17057383277516674, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁
eval/runtime▁▅▅█▁
eval/samples_per_second█▄▅▁█
eval/steps_per_second█▅▅▁█
eval_loss▁
train/epoch▁▁▃▃▅▅▇▇██
train/global_step▁▁▃▃▅▅▇▇███
train/learning_rate█▆▃▁
train/loss█▂▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.85559
eval/runtime5.9033
eval/samples_per_second7.284
eval/steps_per_second1.016
eval_loss0.85559
train/epoch4.0
train/global_step44
train/learning_rate5e-05
train/loss0.4734
train/total_flos1395517145776128.0
train/train_loss1.0546
train/train_runtime442.5635
train/train_samples_per_second3.019
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run crimson-dragon-16 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/gqh7mnqx
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_152334-gqh7mnqx/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:31:13,768] Trial 5 finished with value: 0.855594277381897 and parameters: {'learning_rate': 0.0004882684074952214, 'num_train_epochs': 4, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"b17199cf7d1c41ee969d27814d9efe8b","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113176900026802, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_153113-emmid59b"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run beaming-paper-17 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/emmid59b"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [66/66 10:55, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8145001.876597
201.0222001.064405
300.6246000.902199
400.5075000.854691
500.4362000.856049
600.4078000.836567

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17026547676022635, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▁▁▁▁▁
eval/runtime▁▁▄▃█▅▃
eval/samples_per_second██▅▅▁▄▆
eval/steps_per_second██▄▅▁▄▅
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83657
eval/runtime5.9028
eval/samples_per_second7.285
eval/steps_per_second1.016
eval_loss0.83657
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.4078
train/total_flos2090258212601856.0
train/train_loss0.91464
train/train_runtime664.3977
train/train_samples_per_second3.016
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run beaming-paper-17 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/emmid59b
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_153113-emmid59b/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:42:35,259] Trial 6 finished with value: 0.8365665674209595 and parameters: {'learning_rate': 0.000371977101120841, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 4}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"3718f807c93646f8ba07bbc2d3594547","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113231277947003, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_154235-x78afq0n"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run glittering-monkey-18 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/x78afq0n"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [44/44 07:14, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.1951002.187073
201.5745001.620554
300.9015001.120180
400.6909001.036435

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17055463319920083, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▅▂▁▁
eval/runtime▁▆▄█▅
eval/samples_per_second█▃▅▁▄
eval/steps_per_second█▃▅▁▄
eval_loss▁
train/epoch▁▁▃▃▅▅▇▇██
train/global_step▁▁▃▃▅▅▇▇███
train/learning_rate█▆▃▁
train/loss█▃▂▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss1.03644
eval/runtime5.9099
eval/samples_per_second7.276
eval/steps_per_second1.015
eval_loss1.03644
train/epoch4.0
train/global_step44
train/learning_rate2e-05
train/loss0.6909
train/total_flos1395517145776128.0
train/train_loss1.50181
train/train_runtime442.6998
train/train_samples_per_second3.018
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run glittering-monkey-18 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/x78afq0n
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_154235-x78afq0n/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:50:14,768] Trial 7 finished with value: 1.0364350080490112 and parameters: {'learning_rate': 0.00021352963324526537, 'num_train_epochs': 4, 'per_device_train_batch_size': 32, 'warmup_steps': 5}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d31e8ddbc07347adb1c8a3798ee36db8","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113219644499218, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_155014-8e8ip35f"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run beaming-fuse-19 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/8e8ip35f"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [120/126 11:45 < 00:35, 0.17 it/s, Epoch 5/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7196001.833074
201.0752001.046847
300.6402000.922950
400.5504000.871228
500.4986000.876081
600.4539000.857398
700.4283000.860341
800.4096000.856458
900.3941000.830379
1000.3715000.847909
1100.3818000.842489
1200.3595000.851102

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:04]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.012 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.35110723430597374, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▂▂▃▄▆█▄█▇▅█▁
eval/samples_per_second█▇▇▆▅▃▁▅▁▂▄▁█
eval/steps_per_second▇▇▇▆▅▃▁▅▁▁▅▁█
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83038
eval/runtime5.8734
eval/samples_per_second7.321
eval/steps_per_second1.022
eval_loss0.83038
train/epoch5.71
train/global_step120
train/learning_rate2e-05
train/loss0.3595
train/total_flos1821144316674048.0
train/train_loss0.69023
train/train_runtime710.9352
train/train_samples_per_second2.819
train/train_steps_per_second0.177

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run beaming-fuse-19 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/8e8ip35f
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_155014-8e8ip35f/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 16:02:22,467] Trial 8 finished with value: 0.8303791284561157 and parameters: {'learning_rate': 0.0003782307395143863, 'num_train_epochs': 6, 'per_device_train_batch_size': 16, 'warmup_steps': 3}. Best is trial 8 with value: 0.8303791284561157.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d8cdb5224da6481691ab8959404e3d24","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113188377607407, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_160222-bob3m06p"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run dazzling-ox-20 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/bob3m06p"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [140/168 13:43 < 02:47, 0.17 it/s, Epoch 6/8]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7859001.793209
201.0033001.026370
300.6111000.900247
400.5459000.883518
500.4864000.883806
600.4442000.881959
700.4328000.877395
800.4108000.867789
900.4000000.852276
1000.3731000.867085
1100.3862000.850435
1200.3738000.855305
1300.3920000.853369
1400.3604000.869668

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.1705731731337404, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▄▅▆▆▆▆▇▇█▆▆▆▇▆
eval/samples_per_second█▅▄▃▃▃▃▂▁▁▃▃▃▂▃
eval/steps_per_second█▅▄▂▃▄▃▂▂▁▂▃▄▂▃
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▃▃▄▄▄▄▅▅▅▅▆▆▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▃▃▄▄▄▄▅▅▅▅▆▆▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▆▅▅▄▄▃▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.85044
eval/runtime5.8867
eval/samples_per_second7.305
eval/steps_per_second1.019
eval_loss0.85044
train/epoch6.67
train/global_step140
train/learning_rate8e-05
train/loss0.3604
train/total_flos2129326975303680.0
train/train_loss0.64328
train/train_runtime828.8076
train/train_samples_per_second3.224
train/train_steps_per_second0.203

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run dazzling-ox-20 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/bob3m06p
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_160222-bob3m06p/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 16:16:28,206] Trial 9 finished with value: 0.8504351377487183 and parameters: {'learning_rate': 0.000457264058410859, 'num_train_epochs': 8, 'per_device_train_batch_size': 16, 'warmup_steps': 5}. Best is trial 8 with value: 0.8303791284561157.\n"]}],"source":["def objective(trial):\n","\n"," # Define hyperparameters outside the wandb.init to use them later in the code\n"," learning_rate = trial.suggest_float('learning_rate', 2e-4, 5e-4, log=True)\n"," num_train_epochs = trial.suggest_categorical('num_train_epochs', [4, 6, 8])\n"," per_device_train_batch_size = trial.suggest_categorical('per_device_train_batch_size', [16, 32])\n"," warmup_steps = trial.suggest_int('warmup_steps', 3, 5)\n","\n"," wandb.init(\n"," project=\"Education-Chatbot-Optimization\",\n"," entity=\"szehanz\",\n"," group=\"optuna-optimization\",\n"," job_type=\"hyperparameter_search\",\n"," reinit=True,\n"," config={\n"," \"learning_rate\": learning_rate,\n"," \"num_train_epochs\": num_train_epochs,\n"," \"per_device_train_batch_size\": per_device_train_batch_size,\n"," \"warmup_steps\": warmup_steps\n"," }\n"," )\n","\n"," # Format the current date and time\n"," current_time = datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n"," output_dir = f\"train_out_dir_{current_time}\" # Append the current date and time to the directory name\n","\n"," # Create the output directory\n"," os.makedirs(output_dir, exist_ok=True) # Using exist_ok=True to avoid error if the directory already exists\n","\n","\n"," # Define TrainingArguments with the suggested hyperparameters\n"," training_args = TrainingArguments(\n"," output_dir=output_dir, # Directory for saving output models and checkpoints.\n"," save_strategy=\"steps\", # Save model checkpoints at regular step intervals.\n"," save_steps=10, # Save model checkpoints every 10 steps.\n"," learning_rate=learning_rate, # Initial learning rate for the optimizer.\n"," per_device_train_batch_size=per_device_train_batch_size, # Batch size per device during training.\n"," per_device_eval_batch_size=8, # Batch size per device during evaluation.\n"," num_train_epochs=num_train_epochs, # Total number of training epochs.\n"," warmup_steps=warmup_steps, # Number of warmup steps for the learning rate scheduler.\n"," evaluation_strategy='steps', # Perform evaluation at regular step intervals.\n"," eval_steps=10, # Perform evaluation every 10 steps.\n"," logging_steps=10,\n"," optim='paged_adamw_8bit', # Specifies the optimizer to use.\n"," lr_scheduler_type='linear', # Type of learning rate scheduler.\n"," gradient_accumulation_steps=1, # Number of steps to accumulate gradients before performing an update.\n"," load_best_model_at_end=True, # Load the best model based on evaluation metric at the end of training.\n"," report_to='wandb', # Disable automatic integrations with external reporting tools.\n"," )\n","\n","\n"," # Initialize the Trainer with early stopping callback inside the objective function\n"," trainer = SFTTrainer(\n"," model=model, # Ensure a function or a mechanism to initialize your model\n"," train_dataset=train_dataset,\n"," eval_dataset=val_dataset,\n"," peft_config=peft_config,\n"," dataset_text_field=\"Instruction\",\n"," tokenizer=tokenizer,\n"," args=training_args,\n"," max_seq_length=4096,\n"," callbacks=[EarlyStoppingCallback(early_stopping_patience=3)],\n"," )\n","\n"," # Train the model and evaluate within the objective function\n"," trainer.train()\n"," eval_result = trainer.evaluate()\n","\n"," # Log the primary metric to WandB\n"," wandb.log({\"eval_loss\": eval_result[\"eval_loss\"]})\n","\n"," # Finish the WandB run for this trial\n"," wandb.finish()\n","\n"," # Return the metric to be optimized\n"," return eval_result[\"eval_loss\"]\n","\n","\n","# Run the optimization\n","study = optuna.create_study(direction='minimize')\n","study.optimize(objective, n_trials=10)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"fmdlQTVSHT8e","outputId":"a2935a56-5cad-4dbc-c55c-53b3b5ad1368"},"outputs":[{"name":"stdout","output_type":"stream","text":["Best trial:\n"," Value: 0.8303791284561157\n"," Params: \n"," learning_rate: 0.0003782307395143863\n"," num_train_epochs: 6\n"," per_device_train_batch_size: 16\n"," warmup_steps: 3\n"]}],"source":["# Best trial results\n","print(\"Best trial:\")\n","print(f\" Value: {study.best_trial.value}\")\n","print(\" Params: \")\n","for key, value in study.best_trial.params.items():\n"," print(f\" {key}: {value}\")"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"mKlA_ahVHT8e","outputId":"6365a674-b011-48bb-94ea-7aa9d657d323","colab":{"referenced_widgets":[""]}},"outputs":[{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_161628-5gyifk7s"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run floating-fish-2 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/huggingface"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/huggingface/runs/5gyifk7s"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [126/126 12:16, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7158001.857712
201.0773001.051454
300.6475000.913019
400.5472000.881412
500.4899000.886365
600.4574000.855178
700.4284000.860198
800.4072000.863780
900.3950000.834071
1000.3723000.848378
1100.3795000.848452
1200.3588000.857301

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17130191715842674, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▂▁▁▁▁▁▁▁▁▁
eval/runtime▁▄▄▆▆▆▅█▇▇▆█
eval/samples_per_second█▅▅▃▃▃▄▁▁▂▃▁
eval/steps_per_second█▅▅▃▃▃▄▂▂▂▃▁
train/epoch▁▁▂▂▂▂▃▃▃▃▄▄▅▅▅▅▆▆▆▆▇▇███
train/global_step▁▁▂▂▂▂▃▃▃▃▄▄▅▅▅▅▆▆▆▆▇▇███
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.8573
eval/runtime5.9051
eval/samples_per_second7.282
eval/steps_per_second1.016
train/epoch6.0
train/global_step126
train/learning_rate2e-05
train/loss0.3588
train/total_flos1913972332118016.0
train/train_loss0.67577
train/train_runtime746.485
train/train_samples_per_second2.685
train/train_steps_per_second0.169

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run floating-fish-2 at: https://wandb.ai/szehanz/huggingface/runs/5gyifk7s
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_161628-5gyifk7s/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"}],"source":["# Use best hyperparameters from the study\n","best_trial = study.best_trial\n","\n","best_learning_rate = best_trial.params['learning_rate']\n","best_num_train_epochs = best_trial.params['num_train_epochs']\n","best_per_device_train_batch_size = best_trial.params['per_device_train_batch_size']\n","best_warmup_steps = best_trial.params['warmup_steps']\n","\n","\n","# Define TrainingArguments with the best hyperparameters for retraining\n","best_training_args = TrainingArguments(\n"," output_dir=\"best_train_out_dir\",\n"," save_strategy=\"steps\",\n"," save_steps=10,\n"," learning_rate=best_learning_rate,\n"," per_device_train_batch_size=best_per_device_train_batch_size,\n"," per_device_eval_batch_size=8,\n"," num_train_epochs=best_num_train_epochs,\n"," warmup_steps=best_warmup_steps,\n"," evaluation_strategy='steps',\n"," eval_steps=10,\n"," logging_steps=10,\n"," optim='paged_adamw_8bit',\n"," lr_scheduler_type='linear',\n"," gradient_accumulation_steps=1,\n"," load_best_model_at_end=True,\n"," report_to='wandb',\n",")\n","\n","# Reinitialize the Trainer with the best hyperparameters\n","best_trainer = SFTTrainer(\n"," model=model,\n"," train_dataset=train_dataset,\n"," eval_dataset=val_dataset,\n"," peft_config=peft_config,\n"," dataset_text_field=\"Instruction\",\n"," tokenizer=tokenizer,\n"," args=best_training_args,\n"," max_seq_length=4096,\n",")\n","\n","# Retrain the model with the best hyperparameters\n","best_trainer.train()\n","\n","\n","# Save trained model\n","best_trainer.model.save_pretrained(new_model)\n","\n","# Finish the WandB run for this trial\n","wandb.finish()"]},{"cell_type":"markdown","metadata":{"id":"_g0fB7P9s0ol"},"source":["Merging the base model with the trained adapter."]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"referenced_widgets":["aafec7a64d034e05b1aaf17bb153136b","1191c9b140394f1aa3952c1cecda8fed","68107c402ec343ffa40e22171e9fe3e9"]},"id":"QQn30cRtAZ-P","outputId":"6508be7b-0a96-494e-bd33-d35c5c331f52"},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"68107c402ec343ffa40e22171e9fe3e9","version_major":2,"version_minor":0},"text/plain":["Loading checkpoint shards: 0%| | 0/2 [00:00
Copy a token from your Hugging Face\ntokens page and paste it below.
Immediately click login after copying\nyour token or it might be stored in plain text in this notebook file.
"}},"14d0ff6e07a148c0b65224af86af6b95":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"187745df1f154e7da09a0c646a22cebb":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_638517f3c94445c7b610b84744859ef8","placeholder":"​","style":"IPY_MODEL_5206f9fc25234586b735fc9c65e6b19a","value":"generation_config.json: 100%"}},"19b8b7692cfd46eca5abb50ac2854262":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"19d3adedf1a245f69a2498b1838f6415":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_31ffba7793b944cd80bec0fb6ac2fbbe","IPY_MODEL_1ae848be349a41239a7ba6efadc82edf","IPY_MODEL_9a33d92ad0604eff9ac56944586facbb"],"layout":"IPY_MODEL_293d425206684064b0572e990a815cee"}},"1ae848be349a41239a7ba6efadc82edf":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_8ae8a9d03adc476a8dd64f09e194907d","max":2,"min":0,"orientation":"horizontal","style":"IPY_MODEL_aeb94dbf3f8a4b43bb07c833af3b98c9","value":2}},"1c6e5dd4ca3446ec8cf8a03b85ad57dc":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_0786fc506a7640a287016991f57d3768","IPY_MODEL_3af05a188f40469dbd52fda55c7a9e22","IPY_MODEL_0af30e36f2524c1084c1237948f3b18e"],"layout":"IPY_MODEL_dfe1c2648a564676bfe1e09bda61d439"}},"1ea12323b1244c768a17a6ed5420f854":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"1f9e8ddb08814b2db7761478eb9069f5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_7982d1d315964d138f31643db445b48a","max":614,"min":0,"orientation":"horizontal","style":"IPY_MODEL_67346bc716384552a69d1afdc844db00","value":614}},"20cc750b1bde4efdb3c3a5b5321222b4":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_a199f20edde447829b17804b17af2fe5","max":188,"min":0,"orientation":"horizontal","style":"IPY_MODEL_447f608108364159ab7ed546ad02559a","value":188}},"21556be54ed34b15b909bf8e7b8fd93a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b8113970ea7245e9890221d4e4cf5e8e","placeholder":"​","style":"IPY_MODEL_d3224d16458249a3bfd29253c2d6a86f","value":"\nPro Tip: If you don't already have one, you can create a dedicated\n'notebooks' token with 'write' access, that you can then easily reuse for all\nnotebooks. "}},"2254fe28315f43a99c3579195c0c0008":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"22e8bae9ff6744bca3e90d46d220106f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"23176fa0130348c5bc1b0deb591709f3":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"236b3417f0e94039a1a449e7ab96738b":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"25543b118c8e423aa7fd8c898d5e1256":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"2924e96aa10346efb39684e5369e2170":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_bce8d1501218410ba8b042aeb3f0fc26","placeholder":"​","style":"IPY_MODEL_405603de026d484ab283f053f4b17c6d","value":"Your token has been saved to /root/.cache/huggingface/token"}},"293d425206684064b0572e990a815cee":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"29f303aa6ac8464aa91124c3fe659379":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2cb9cfbde1e0483c97a2c531e0034adf":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2dce1978d19e4de3a6a1b1cef6ed518f":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"3076e4abb7fe427fa4fccb43e9f3371e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"30d47a9da70a4cf5a7e597f13d28d526":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"31ffba7793b944cd80bec0fb6ac2fbbe":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_7ee17b5879ec4909b7b616b79d5eab36","placeholder":"​","style":"IPY_MODEL_d8b847adb46d4049a4b43bba2a0e06da","value":"Downloading shards: 100%"}},"32a57f9df45b41da91e1d638f531505d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"3af05a188f40469dbd52fda55c7a9e22":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_82829039182948efb800703e3e21d96d","max":26788,"min":0,"orientation":"horizontal","style":"IPY_MODEL_32a57f9df45b41da91e1d638f531505d","value":26788}},"3e4a06b9b13444e3b82e0c3c26e17b8f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"CheckboxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"CheckboxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"CheckboxView","description":"Add token as git credential?","description_tooltip":null,"disabled":false,"indent":true,"layout":"IPY_MODEL_f91ebc43c1344e8688e2eeb2771c7b65","style":"IPY_MODEL_ea00aa1eb73949fc94083f1d31372915","value":true}},"405603de026d484ab283f053f4b17c6d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"42262413163d49fd8902c275c13da1e6":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_cf370d3015b04b6caa9cbfb8f21dd5e7","placeholder":"​","style":"IPY_MODEL_1ea12323b1244c768a17a6ed5420f854","value":" 9.98G/9.98G [01:29<00:00, 126MB/s]"}},"441c37014e6b4c8995e7a5305f96fa38":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"447f608108364159ab7ed546ad02559a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"46e41557d0bc4cb4a03c6894fcc57d0e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"4cee3e23bcbd4643ba542d94b82dbfd2":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"5206f9fc25234586b735fc9c65e6b19a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"52f2b6adb9904cbeb014387429c0d1ad":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_b61ac64b2a0c4beb8ea209782b9dd09b","IPY_MODEL_9e9c2d712ae84d3aa87fbe433ae0701b","IPY_MODEL_fe4d028a78e64a4d8b47f8bc8b8ec15d"],"layout":"IPY_MODEL_88a7aeaf3b2547ea91cf20ed2f1a3dc0"}},"537d173a4313491f8f89f799f1dc7b2a":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"541ef20ab6f34337a2d6d20098f6fef5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b0f0ac261e364edd99d7b75e747e2c47","placeholder":"​","style":"IPY_MODEL_ff9f726db3434e3184e723d5da884d0a","value":"Token is valid (permission: write)."}},"55e506bd03d24dd7b24129dc64da9c01":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"5e5b95c9801443cdbce9c8e629c33589":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"PasswordModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"PasswordModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"PasswordView","continuous_update":true,"description":"Token:","description_tooltip":null,"disabled":false,"layout":"IPY_MODEL_29f303aa6ac8464aa91124c3fe659379","placeholder":"​","style":"IPY_MODEL_3076e4abb7fe427fa4fccb43e9f3371e","value":""}},"62f1cf19fe204aa4a424248e807ce061":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"638517f3c94445c7b610b84744859ef8":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"6430379d01874ec3a7cf9fea59c42914":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ButtonModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ButtonModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ButtonView","button_style":"","description":"Login","disabled":false,"icon":"","layout":"IPY_MODEL_2dce1978d19e4de3a6a1b1cef6ed518f","style":"IPY_MODEL_e81c501824f94e7d839684fafbc65b31","tooltip":""}},"67346bc716384552a69d1afdc844db00":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"6edf40f558f54d8b82d949f83557d609":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_08400a144d3c497a94ae4d84e72a1067","placeholder":"​","style":"IPY_MODEL_976a3440d2c3423c8be835b0d6f56492","value":"Connecting..."}},"77c0d63f6cd6494daf7e94cfe0397e45":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_4cee3e23bcbd4643ba542d94b82dbfd2","placeholder":"​","style":"IPY_MODEL_25543b118c8e423aa7fd8c898d5e1256","value":" 188/188 [00:00<00:00, 12.3kB/s]"}},"77daad9def9f4254933afdeb7202e5d2":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"7982d1d315964d138f31643db445b48a":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"7c701d53772344f6a057cd43742d47f5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b9e4e59acd4d468393f42857d5635e25","placeholder":"​","style":"IPY_MODEL_30d47a9da70a4cf5a7e597f13d28d526","value":"Loading checkpoint shards: 100%"}},"7ee17b5879ec4909b7b616b79d5eab36":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"8080eb1edf8f44f69b5f218046715f36":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_e2df35aaf7e0420b9524d533cb122521","max":2,"min":0,"orientation":"horizontal","style":"IPY_MODEL_d90c3a424bac4bb09e99339c28df2eda","value":2}},"82829039182948efb800703e3e21d96d":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"83127281047a429a846b9bb3b3209760":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"86014f569ee34895badef4acfe958051":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"86616a0d2f9d4da1a75e00d092ca9d4d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_a7a17a5883784ed9961437d0102fd9a4","IPY_MODEL_9934540648a444a09f0783843ab2d901","IPY_MODEL_42262413163d49fd8902c275c13da1e6"],"layout":"IPY_MODEL_83127281047a429a846b9bb3b3209760"}},"88a7aeaf3b2547ea91cf20ed2f1a3dc0":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"8ae8a9d03adc476a8dd64f09e194907d":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"963973c1e970410cb2ba97e9fb4ef511":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"976a3440d2c3423c8be835b0d6f56492":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"9934540648a444a09f0783843ab2d901":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_f702ae2d46834e9b934a40abca49a990","max":9976576152,"min":0,"orientation":"horizontal","style":"IPY_MODEL_14d0ff6e07a148c0b65224af86af6b95","value":9976576152}},"9a33d92ad0604eff9ac56944586facbb":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_537d173a4313491f8f89f799f1dc7b2a","placeholder":"​","style":"IPY_MODEL_55e506bd03d24dd7b24129dc64da9c01","value":" 2/2 [01:56<00:00, 52.47s/it]"}},"9c7550ec29c34ef2b674a3667ebd740e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_236b3417f0e94039a1a449e7ab96738b","placeholder":"​","style":"IPY_MODEL_77daad9def9f4254933afdeb7202e5d2","value":"config.json: 100%"}},"9e4d26b87e1243f3ae744ddbbe2c5b51":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"9e9c2d712ae84d3aa87fbe433ae0701b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_86014f569ee34895badef4acfe958051","max":3500296424,"min":0,"orientation":"horizontal","style":"IPY_MODEL_22e8bae9ff6744bca3e90d46d220106f","value":3500296424}},"a066a28914bc46cc85ebcb7d90d27fbe":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a199f20edde447829b17804b17af2fe5":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a39b2a49ff814181aa0df84c57e9245b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"a7a17a5883784ed9961437d0102fd9a4":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_ae173a50f4374c7fa55d8e360d30d4d1","placeholder":"​","style":"IPY_MODEL_b93a21bcaf6344c49a0ab8e2a887922c","value":"model-00001-of-00002.safetensors: 100%"}},"a7ef8ff133144d4b9817800e5b4739a4":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a9628200c0a34cccacdb0ea66e200b87":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_0cda6495ef4b4c9ab3eeff2392d89163","placeholder":"​","style":"IPY_MODEL_9e4d26b87e1243f3ae744ddbbe2c5b51","value":" 2/2 [00:58<00:00, 26.87s/it]"}},"a96de5c1f80f43d8b0b689d5db0fa248":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"ae173a50f4374c7fa55d8e360d30d4d1":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"aeb94dbf3f8a4b43bb07c833af3b98c9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"b0f0ac261e364edd99d7b75e747e2c47":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b4ed4d5e266a4cd5b9011436b4e92951":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_7c701d53772344f6a057cd43742d47f5","IPY_MODEL_8080eb1edf8f44f69b5f218046715f36","IPY_MODEL_a9628200c0a34cccacdb0ea66e200b87"],"layout":"IPY_MODEL_2254fe28315f43a99c3579195c0c0008"}},"b577aeffe358446ea41e4b3c58b80641":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_441c37014e6b4c8995e7a5305f96fa38","placeholder":"​","style":"IPY_MODEL_a39b2a49ff814181aa0df84c57e9245b","value":" 614/614 [00:00<00:00, 15.7kB/s]"}},"b61ac64b2a0c4beb8ea209782b9dd09b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_a066a28914bc46cc85ebcb7d90d27fbe","placeholder":"​","style":"IPY_MODEL_a96de5c1f80f43d8b0b689d5db0fa248","value":"model-00002-of-00002.safetensors: 100%"}},"b634ae14caf84bc99dcafd3a70b7a383":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b64f26ac024c46eabfc4728586369130":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_a7ef8ff133144d4b9817800e5b4739a4","placeholder":"​","style":"IPY_MODEL_ba4cf32b2f71428282721e7818b34a5a","value":"Login successful"}},"b8113970ea7245e9890221d4e4cf5e8e":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b93a21bcaf6344c49a0ab8e2a887922c":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"b9e4e59acd4d468393f42857d5635e25":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ba4cf32b2f71428282721e7818b34a5a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"bce8d1501218410ba8b042aeb3f0fc26":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"c3d08bc595a74c3180a7a83afc569584":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"VBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"VBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"VBoxView","box_style":"","children":["IPY_MODEL_541ef20ab6f34337a2d6d20098f6fef5","IPY_MODEL_fe46e1cf697f4b1fab764104be32da95","IPY_MODEL_2924e96aa10346efb39684e5369e2170","IPY_MODEL_b64f26ac024c46eabfc4728586369130"],"layout":"IPY_MODEL_cad60b6f14f249c187d573dd3a4428e0"}},"ca8d07b9b454471e883d43e03b475a9f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"cad60b6f14f249c187d573dd3a4428e0":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":"center","align_self":null,"border":null,"bottom":null,"display":"flex","flex":null,"flex_flow":"column","grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":"50%"}},"cf370d3015b04b6caa9cbfb8f21dd5e7":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"d3224d16458249a3bfd29253c2d6a86f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d8b847adb46d4049a4b43bba2a0e06da":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d90c3a424bac4bb09e99339c28df2eda":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"de756c426cf0492bb122a45b94d4bbe7":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"df0d8231953b44f1bf8e6d2cd39b2f66":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_9c7550ec29c34ef2b674a3667ebd740e","IPY_MODEL_1f9e8ddb08814b2db7761478eb9069f5","IPY_MODEL_b577aeffe358446ea41e4b3c58b80641"],"layout":"IPY_MODEL_19b8b7692cfd46eca5abb50ac2854262"}},"dfe1c2648a564676bfe1e09bda61d439":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"e2df35aaf7e0420b9524d533cb122521":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"e81c501824f94e7d839684fafbc65b31":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ButtonStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ButtonStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","button_color":null,"font_weight":""}},"ea00aa1eb73949fc94083f1d31372915":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"f702ae2d46834e9b934a40abca49a990":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"f81ada25e7ff4f5da6b3f6c6e73590e4":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"f91ebc43c1344e8688e2eeb2771c7b65":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"fe46e1cf697f4b1fab764104be32da95":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_f81ada25e7ff4f5da6b3f6c6e73590e4","placeholder":"​","style":"IPY_MODEL_62f1cf19fe204aa4a424248e807ce061","value":"Your token has been saved in your configured git credential helpers (store)."}},"fe4d028a78e64a4d8b47f8bc8b8ec15d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b634ae14caf84bc99dcafd3a70b7a383","placeholder":"​","style":"IPY_MODEL_46e41557d0bc4cb4a03c6894fcc57d0e","value":" 3.50G/3.50G [00:26<00:00, 176MB/s]"}},"ff9f726db3434e3184e723d5da884d0a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}}}}},"nbformat":4,"nbformat_minor":0} \ No newline at end of file +{"cells":[{"cell_type":"markdown","source":["## Creation of the MCQ-tuned model by fine-tuning the Base model with Custom MCQ dataset"],"metadata":{"id":"dcQxOEQQ6btB"}},{"cell_type":"markdown","source":["---"],"metadata":{"id":"LlkcVvTDs96X"}},{"cell_type":"markdown","source":["# MCQ Custom Dataset Creation\n","\n","## 1. Introduction\n","\n","High-quality data is fundamental for producing a good model; the higher the quality of the data, the better the resulting model. The following steps outline the process of creating a dataset specifically for fine-tuning our Llama2 model.\n","\n","\n","\n","![](https://i.imgur.com/IDNhAWH.png)\n","\n","\n","There are several types of datasets that can be used to fine-tune Large Language Models (LLMs):\n","\n","1. **Instruction Datasets:** These datasets contain direct instructions or prompts followed by the correct or expected outputs.\n","\n","2. **Raw Completion:** This involves providing a prompt to the model and letting it generate a response without any predefined correct answer.\n","\n","3. **Preference Datasets:** These datasets include human feedback in the form of preferences, where annotators compare pairs of model outputs to determine which is better.\n","\n","4. **Human Feedback Data:** This is specific to Reinforcement Learning from Human Feedback (RLHF) and involves direct feedback on the model's outputs from human annotators.\n","\n","5. **Demonstration Data:** Also used in RLHF, these datasets consist of examples showing ideal model outputs or actions, typically created by humans.\n","\n","6. **Reward Modeling Data:** Used to train a reward model in RLHF, this dataset predicts human feedback on model outputs based on actual feedback data.\n","\n","7. **Dialogue Data:** Particularly relevant for conversational AI, this includes annotated conversations that indicate the quality of responses or provide corrections.\n","\n","\n","---\n","\n","\n","\n","* Typically, an instruction dataset is utilized for fine-tuning the Llama 2 Model. Since we are focusing on Supervised Fine Tuning, the instruction dataset becomes our primary choice.\n","\n","Therefore, we have 2 options:\n","\n","1. Create our own Instruction Dataset.\n","2. Modify an existing instruction dataset, which involves filtering, modifying, and enriching it.\n","\n","We have decided to proceed with the 1st option: creating our own Instruction Dataset.\n","\n","* This will involve prompt engineering and incorporating sanity checks to ensure quality and relevance."],"metadata":{"id":"wAQMA1-DKZZ5"}},{"cell_type":"markdown","source":["## 2. Load and analyze the dataset"],"metadata":{"id":"hU_mUK-nol-t"}},{"cell_type":"code","execution_count":null,"metadata":{"id":"8P7g6eHuxxKe"},"outputs":[],"source":["# Install libraries\n","!pip install -q datasets transformers sentence_transformers faiss-gpu huggingface_hub"]},{"cell_type":"code","source":["# Import the required libraries\n","import json\n","import sys\n","import pandas as pd\n","from datasets import Dataset, DatasetDict, load_dataset\n","\n","from transformers import AutoTokenizer\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","\n","from sentence_transformers import SentenceTransformer\n","import faiss\n","from tqdm.autonotebook import tqdm\n","import numpy as np"],"metadata":{"id":"KKb-ikj4J-in"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Load JSON data from a file\n","with open(\"mcq_data.json\", \"r\") as f:\n"," data = json.load(f)\n","\n","# Create a Pandas DataFrame from the list of dictionaries\n","df = pd.DataFrame(data)\n","\n","# Calculate the number of rows for each dataset split\n","num_rows = len(df)\n","train_end = int(num_rows * 0.8) # 80% for training\n","test_end = train_end + int(num_rows * 0.1) # 10% for testing\n","\n","# Split the DataFrame into training, testing, and validation sets\n","df_train = df[:train_end]\n","df_test = df[train_end:test_end]\n","df_val = df[test_end:] # Ensures the remainder is used for validation\n","\n","# Create Datasets from the DataFrames\n","dataset_train = Dataset.from_pandas(df_train)\n","dataset_test = Dataset.from_pandas(df_test)\n","dataset_val = Dataset.from_pandas(df_val)\n","\n","# Create a DatasetDict containing the split datasets\n","dataset = DatasetDict({\n"," 'train': dataset_train,\n"," 'test': dataset_test,\n"," 'val': dataset_val\n","})\n","\n","# Print the structure of the created DatasetDict\n","print(dataset)"],"metadata":{"id":"bGi9FdmhdBDg","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1708322802114,"user_tz":-480,"elapsed":19,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"e9369555-5be7-4b43-a6c4-3defec1485d6"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["DatasetDict({\n"," train: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 334\n"," })\n"," test: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 41\n"," })\n"," val: Dataset({\n"," features: ['Instruction', 'Question', 'A', 'B', 'C', 'D', 'Correct Answer', 'Explanation'],\n"," num_rows: 43\n"," })\n","})\n"]}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['train'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":719},"id":"-MOvcr5mD8li","executionInfo":{"status":"ok","timestamp":1708322802114,"user_tz":-480,"elapsed":18,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"5b69a1a1-5307-4ec8-9b04-2c5035176d98"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the structure of artificial n... \n","1 Create an MCQ on the training process of artif... \n","2 Create an MCQ on the role of artificial neuron... \n","3 Create an MCQ on the purpose of hidden layers ... \n","4 Create an MCQ on the basics of deep learning \n",".. ... \n","329 Create an MCQ on the hyperparameter 'Kernel' i... \n","330 Create an MCQ on the hyperparameter 'Gamma' in... \n","331 Create an MCQ on the hyperparameter 'learning_... \n","332 Create an MCQ on the hyperparameter 'n_estimat... \n","333 Create an MCQ on the application of deep learn... \n","\n"," Question \\\n","0 What is the structure of an artificial neural ... \n","1 What is the purpose of the training process in... \n","2 What is the role of artificial neurons in neur... \n","3 What is the purpose of hidden layers in artifi... \n","4 What is deep learning? \n",".. ... \n","329 What does the hyperparameter 'Kernel' define i... \n","330 What does the hyperparameter 'Gamma' control i... \n","331 What does the hyperparameter 'learning_rate' d... \n","332 What does the hyperparameter 'n_estimators' de... \n","333 Which of the following is an application of de... \n","\n"," A \\\n","0 It consists of input layers and hidden layers ... \n","1 To adjust the weights of the connections betwe... \n","2 To receive input from external sources \n","3 To receive input from external sources \n","4 A branch of machine learning based on artifici... \n",".. ... \n","329 The step size taken by the optimizer during ea... \n","330 The step size taken by the optimizer during ea... \n","331 The step size taken by the optimizer during ea... \n","332 The step size taken by the optimizer during ea... \n","333 Analyzing sensor data in autonomous vehicles \n","\n"," B \\\n","0 It consists of input layers, hidden layers, an... \n","1 To propagate input data forward through the la... \n","2 To compute the weighted total of inputs \n","3 To compute the weighted total of inputs \n","4 A programming technique to explicitly define c... \n",".. ... \n","329 The trade-off between the margin and the numbe... \n","330 The trade-off between the margin and the numbe... \n","331 The trade-off between the margin and the numbe... \n","332 The trade-off between the margin and the numbe... \n","333 Recognizing objects and scenes in images \n","\n"," C \\\n","0 It consists of input layers, hidden layers, ou... \n","1 To calculate the error between the output and ... \n","2 To transfer information to the next layer \n","3 To transfer information to the next layer \n","4 A method to process large datasets using deep ... \n",".. ... \n","329 The similarity between data points \n","330 The similarity between data points \n","331 The similarity between data points \n","332 The number of boosting trees to be trained \n","333 Transcribing spoken words into text \n","\n"," D Correct Answer \\\n","0 It consists of input layers, hidden layers, ou... C \n","1 To achieve the desired level of performance A \n","2 All of the above D \n","3 To process and transform the input data D \n","4 A type of data structure inspired by the human... A \n",".. ... ... \n","329 The maximum depth of each tree in the ensemble C \n","330 The influence of support vectors on the decisi... D \n","331 The maximum depth of each tree in the ensemble A \n","332 The maximum depth of each tree in the ensemble C \n","333 Making personalized recommendations based on u... B \n","\n"," Explanation \n","0 An artificial neural network consists of input... \n","1 The purpose of the training process in artific... \n","2 The role of artificial neurons in neural netwo... \n","3 The purpose of hidden layers in artificial neu... \n","4 Deep learning is a branch of machine learning ... \n",".. ... \n","329 The hyperparameter 'Kernel' in Support Vector ... \n","330 The hyperparameter 'Gamma' in Support Vector M... \n","331 The hyperparameter 'learning_rate' in XGBoost ... \n","332 The hyperparameter 'n_estimators' in XGBoost d... \n","333 Deep learning algorithms are used in image and... \n","\n","[334 rows x 8 columns]"],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the structure of artificial n...What is the structure of an artificial neural ...It consists of input layers and hidden layers ...It consists of input layers, hidden layers, an...It consists of input layers, hidden layers, ou...It consists of input layers, hidden layers, ou...CAn artificial neural network consists of input...
1Create an MCQ on the training process of artif...What is the purpose of the training process in...To adjust the weights of the connections betwe...To propagate input data forward through the la...To calculate the error between the output and ...To achieve the desired level of performanceAThe purpose of the training process in artific...
2Create an MCQ on the role of artificial neuron...What is the role of artificial neurons in neur...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerAll of the aboveDThe role of artificial neurons in neural netwo...
3Create an MCQ on the purpose of hidden layers ...What is the purpose of hidden layers in artifi...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerTo process and transform the input dataDThe purpose of hidden layers in artificial neu...
4Create an MCQ on the basics of deep learningWhat is deep learning?A branch of machine learning based on artifici...A programming technique to explicitly define c...A method to process large datasets using deep ...A type of data structure inspired by the human...ADeep learning is a branch of machine learning ...
...........................
329Create an MCQ on the hyperparameter 'Kernel' i...What does the hyperparameter 'Kernel' define i...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe maximum depth of each tree in the ensembleCThe hyperparameter 'Kernel' in Support Vector ...
330Create an MCQ on the hyperparameter 'Gamma' in...What does the hyperparameter 'Gamma' control i...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe influence of support vectors on the decisi...DThe hyperparameter 'Gamma' in Support Vector M...
331Create an MCQ on the hyperparameter 'learning_...What does the hyperparameter 'learning_rate' d...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The similarity between data pointsThe maximum depth of each tree in the ensembleAThe hyperparameter 'learning_rate' in XGBoost ...
332Create an MCQ on the hyperparameter 'n_estimat...What does the hyperparameter 'n_estimators' de...The step size taken by the optimizer during ea...The trade-off between the margin and the numbe...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleCThe hyperparameter 'n_estimators' in XGBoost d...
333Create an MCQ on the application of deep learn...Which of the following is an application of de...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...BDeep learning algorithms are used in image and...
\n","

334 rows × 8 columns

\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['train']\",\n \"rows\": 334,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"Create an MCQ on the parameter gamma in Support Vector Machines (SVMs)\",\n \"Create an MCQ on the disadvantages of Artificial Neural Networks (ANNs)\",\n \"Create an MCQ on the role of machine learning in recommendation systems\"\n ],\n \"num_unique_values\": 165,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"What does the hyperparameter 'Kernel' determine in Support Vector Machines (SVMs)?\",\n \"Which of the following are types of deep learning architectures?\",\n \"Which of the following is NOT an application of deep learning in reinforcement learning?\"\n ],\n \"num_unique_values\": 221,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"Requires large amounts of labeled data\",\n \"AI is the broader family consisting of ML and DL as its components\",\n \"Increased computational cost\"\n ],\n \"num_unique_values\": 162,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Data clustering, dimensionality reduction, and anomaly detection\",\n \"Analyzing medical images to assist doctors in making diagnoses\",\n \"The reliance on manual feature engineering\"\n ],\n \"num_unique_values\": 193,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"It may result in overfitting\",\n \"Reduced overfitting and underfitting\",\n \"To automatically learn features from visual data\"\n ],\n \"num_unique_values\": 211,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Overfitting\",\n \"Evaluating all possible combinations of hyperparameter values\",\n \"A branch of machine learning that uses linear regression\"\n ],\n \"num_unique_values\": 211,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"A\",\n \"B\",\n \"C\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"A key difference between machine learning and deep learning is the type of algorithms used. Machine learning applies statistical algorithms, while deep learning utilizes artificial neural network architecture to learn patterns and relationships.\",\n \"Hyperparameter tuning helps reduce overfitting and underfitting, leading to improved model performance and generalizability.\",\n \"Artificial Intelligence consists of the components: Artificial Intelligence, Machine Learning, and Deep Learning.\"\n ],\n \"num_unique_values\": 302,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":19}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['test'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"7WkFWJSQhKUV","executionInfo":{"status":"ok","timestamp":1708322802115,"user_tz":-480,"elapsed":17,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"8fd77b53-8311-42f7-a84d-00e396647a72"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the use of deep learning in g... \n","1 Create an MCQ on the use of deep learning in a... \n","2 Create an MCQ on the application of deep learn... \n","3 Create an MCQ on the use of deep learning in r... \n","4 Create an MCQ on the structure of artificial n... \n","5 Create an MCQ on the purpose of adjusting weig... \n","6 Create an MCQ on the role of artificial neuron... \n","7 Create an MCQ on the complexities of neural ne... \n","8 Create an MCQ on the difference between machin... \n","9 Create an MCQ on the definition of deep learning \n","10 Create an MCQ on the key characteristic of dee... \n","11 Create an MCQ on the applications of deep lear... \n","12 Create an MCQ on the training requirements of ... \n","13 Create an MCQ on the types of machine learning... \n","14 Create an MCQ on the types of neural networks ... \n","15 Create an MCQ on the applications of deep lear... \n","16 Create an MCQ on the applications of deep lear... \n","17 Create an MCQ on the applications of deep lear... \n","18 Create an MCQ on the main types of neural netw... \n","19 Create an MCQ on the definition of Artificial ... \n","20 Create an MCQ on the difference between Machin... \n","21 Create an MCQ on the components of Artificial ... \n","22 Create an MCQ on the aim of Machine Learning \n","23 Create an MCQ on the aim of Deep Learning \n","24 Create an MCQ on the difference between AI, Ma... \n","25 Create an MCQ on the application of AI in spee... \n","26 Create an MCQ on the application of AI in pers... \n","27 Create an MCQ on the application of AI in pred... \n","28 Create an MCQ on the application of AI in medi... \n","29 Create an MCQ on the difference between AI, ML... \n","30 Create an MCQ on the responsibilities of an AI... \n","31 Create an MCQ on the skills required for a Mac... \n","32 Create an MCQ on the tasks of a Deep Learning ... \n","33 Create an MCQ on the difference between ML and DL \n","34 Create an MCQ on the advantages of Artificial ... \n","35 Create an MCQ on the disadvantages of Artifici... \n","36 Create an MCQ on the advantages of Biological ... \n","37 Create an MCQ on the disadvantages of Biologic... \n","38 Create an MCQ on the differences between Artif... \n","39 Create an MCQ on hyperparameter tuning in mach... \n","40 Create an MCQ on the types of hyperparameters ... \n","\n"," Question \\\n","0 What is the role of deep learning in generativ... \n","1 How is deep learning used in autonomous vehicles? \n","2 What is the role of deep learning in speech re... \n","3 What is the application of deep learning in re... \n","4 What is the structure of an artificial neural ... \n","5 What is the purpose of adjusting weights in ar... \n","6 What is the role of artificial neurons in neur... \n","7 What determines the complexities of neural net... \n","8 What is a key difference between machine learn... \n","9 What is the definition of deep learning? \n","10 What is the key characteristic of deep learning? \n","11 Which of the following are applications of dee... \n","12 What are the training requirements for deep ne... \n","13 Which types of machine learning tasks can be p... \n","14 Which type of neural network is specifically d... \n","15 Which application of deep learning in computer... \n","16 Which application of deep learning in NLP invo... \n","17 Which application of deep learning in reinforc... \n","18 Which of the following are the main types of n... \n","19 Which of the following best defines Artificial... \n","20 What is the main difference between Machine Le... \n","21 Which of the following components are part of ... \n","22 What is the aim of Machine Learning? \n","23 What is the aim of Deep Learning? \n","24 Which of the following best describes the diff... \n","25 Which of the following is an example of AI app... \n","26 Which of the following is an example of AI app... \n","27 Which of the following is an example of AI app... \n","28 Which of the following is an example of AI app... \n","29 Which of the following statements accurately d... \n","30 Which of the following is a key responsibility... \n","31 Which of the following skills is essential for... \n","32 Which of the following is a key task of a Deep... \n","33 What distinguishes Deep Learning (DL) from Mac... \n","34 Which of the following is an advantage of Arti... \n","35 Which of the following is a disadvantage of Ar... \n","36 Which of the following is an advantage of Biol... \n","37 Which of the following is a disadvantage of Bi... \n","38 Which of the following is a difference between... \n","39 What is the purpose of hyperparameter tuning i... \n","40 Which of the following is a type of hyperparam... \n","\n"," A \\\n","0 Analyzing sensor data in autonomous vehicles \n","1 Analyzing sensor data in autonomous vehicles \n","2 Analyzing sensor data in autonomous vehicles \n","3 Analyzing sensor data in autonomous vehicles \n","4 It consists of input layers, hidden layers, an... \n","5 To increase the speed of training models \n","6 To receive input from external sources \n","7 The number of layers in the network \n","8 The type of algorithms used \n","9 A branch of machine learning that uses artific... \n","10 The use of shallow neural networks with a sing... \n","11 Image recognition, natural language processing... \n","12 Small datasets and limited computational resou... \n","13 Supervised machine learning only \n","14 Feedforward Neural Networks (FNNs) \n","15 Object detection and recognition \n","16 Automatic Text Generation \n","17 Game playing \n","18 Feedforward Neural Networks (FNNs) \n","19 The study of training machines to mimic human ... \n","20 Machine Learning uses statistical methods, whi... \n","21 Machine Learning and Deep Learning \n","22 To increase chances of success \n","23 To increase chances of success \n","24 AI is a subset of Machine Learning, which is a... \n","25 Analyzing users' browsing and viewing history ... \n","26 Analyzing users' browsing and viewing history ... \n","27 Analyzing users' browsing and viewing history ... \n","28 Analyzing users' browsing and viewing history ... \n","29 AI, ML, and DL are interchangeable terms that ... \n","30 Design and development of AI algorithms \n","31 Strong background in computer science, mathema... \n","32 Design and development of DL algorithms \n","33 DL is a more advanced form of ML that can perf... \n","34 Ability to learn irrespective of the type of data \n","35 Ability to learn irrespective of the type of data \n","36 Ability to learn irrespective of the type of data \n","37 Ability to learn irrespective of the type of data \n","38 Both ANNs and BNNs have complex and diverse ne... \n","39 To adjust the weights and biases of the model \n","40 Weights \n","\n"," B \\\n","0 Creating new content based on existing data \n","1 Recognizing objects and scenes in images \n","2 Recognizing objects and scenes in images \n","3 Recognizing objects and scenes in images \n","4 It consists of input layers and output layers ... \n","5 To prevent overfitting by validating the model... \n","6 To compute the weighted total of inputs \n","7 The number of units in each layer \n","8 The amount of data required \n","9 A type of programming that explicitly defines ... \n","10 The requirement for manual feature engineering \n","11 Data clustering, dimensionality reduction, and... \n","12 Large amounts of data and computational resources \n","13 Unsupervised machine learning only \n","14 Convolutional Neural Networks (CNNs) \n","15 Image classification \n","16 Language translation \n","17 Robotics \n","18 Convolutional Neural Networks (CNNs) \n","19 The study of statistical methods enabling mach... \n","20 Machine Learning focuses on learning from expe... \n","21 Machine Learning and Decision Trees \n","22 To increase accuracy \n","23 To increase accuracy \n","24 Machine Learning is a subset of AI, which is a... \n","25 Analyzing medical images to assist doctors in ... \n","26 Analyzing medical images to assist doctors in ... \n","27 Analyzing medical images to assist doctors in ... \n","28 Analyzing medical images to assist doctors in ... \n","29 AI focuses on creating intelligent machines, M... \n","30 Analysis and interpretation of data \n","31 Experience in developing AI algorithms and sol... \n","32 Analysis and interpretation of data \n","33 DL focuses on developing algorithms that enabl... \n","34 Simple architecture that makes it easy to expl... \n","35 Simple architecture that makes it easy to expl... \n","36 Simple architecture that makes it easy to expl... \n","37 Simple architecture that makes it easy to expl... \n","38 ANNs have fixed connections between neurons, w... \n","39 To select the optimal values for the model's h... \n","40 Biases \n","\n"," C \\\n","0 Transcribing spoken words into text \n","1 Transcribing spoken words into text \n","2 Transcribing spoken words into text \n","3 Making personalized recommendations based on u... \n","4 It consists of input layers, hidden layers, ou... \n","5 To enhance the model's performance on the trai... \n","6 To transfer information to the next layer \n","7 The type of activation function used \n","8 The complexity of the models \n","9 A technique that requires manual feature engin... \n","10 The use of deep neural networks with multiple ... \n","11 Supervised machine learning and unsupervised m... \n","12 Manual feature engineering and domain expertise \n","13 Reinforcement machine learning only \n","14 Recurrent Neural Networks (RNNs) \n","15 Image segmentation \n","16 Sentiment analysis \n","17 Control systems \n","18 Recurrent Neural Networks (RNNs) \n","19 The study that uses neural networks to imitate... \n","20 Machine Learning is a subset of Deep Learning \n","21 Artificial Intelligence and Machine Learning \n","22 To improve system efficiency \n","23 To improve system efficiency \n","24 Deep Learning is a subset of AI, which is a su... \n","25 Recognizing and classifying images and speech \n","26 Recognizing and classifying images and speech \n","27 Analyzing sensor data to predict equipment fai... \n","28 Recognizing and classifying images and speech \n","29 AI is a subset of ML that uses neural networks... \n","30 Training and evaluation of ML models \n","31 Familiarity with programming languages such as... \n","32 Training and evaluation of ML models \n","33 DL is a subset of ML that uses neural networks... \n","34 Dependence on hardware for functioning \n","35 Dependence on hardware for functioning \n","36 No controlling mechanism \n","37 No controlling mechanism \n","38 Both ANNs and BNNs have simple and predetermin... \n","39 To preprocess the input data before training t... \n","40 Learning rate \n","\n"," D Correct Answer \\\n","0 Making personalized recommendations based on u... B \n","1 Making personalized recommendations based on u... A \n","2 Making personalized recommendations based on u... C \n","3 Transcribing spoken words into text C \n","4 It consists of input layers and artificial neu... A \n","5 To reduce the computational cost of training C \n","6 All of the above D \n","7 The size of the dataset B \n","8 The performance on complex tasks B \n","9 A method of machine learning that only works w... A \n","10 The reliance on labeled datasets for training C \n","11 Data visualization and exploratory data analysis A \n","12 Pre-trained models and transfer learning B \n","13 Supervised, unsupervised, and reinforcement ma... D \n","14 None of the above B \n","15 None of the above A \n","16 Speech recognition C \n","17 None of the above B \n","18 All of the above D \n","19 The study of incorporating human intelligence ... D \n","20 Machine Learning requires human intervention, ... A \n","21 Artificial Intelligence and Deep Learning C \n","22 To analyze data and provide output B \n","23 To analyze data and provide output A \n","24 AI, Machine Learning, and Deep Learning are co... B \n","25 Analyzing sensor data to make decisions about ... C \n","26 Analyzing sensor data to make decisions about ... A \n","27 Recognizing and classifying images and speech C \n","28 Analyzing sensor data to make decisions about ... B \n","29 AI focuses on developing algorithms that enabl... B \n","30 Deployment and maintenance of DL models A \n","31 All of the above D \n","32 Deployment and maintenance of AI models A \n","33 ML is a more advanced form of DL that can perf... C \n","34 High speed of processing A \n","35 The simplest architecture makes it difficult t... D \n","36 Ability to process highly complex parallel inputs D \n","37 Speed of processing is slow D \n","38 ANNs and BNNs have the same processing speed B \n","39 To evaluate the performance of the model on a ... B \n","40 Activation function C \n","\n"," Explanation \n","0 Deep learning algorithms are used in generativ... \n","1 Deep learning algorithms are used in autonomou... \n","2 Deep learning algorithms are used in speech re... \n","3 Deep learning algorithms are used in recommend... \n","4 An artificial neural network consists of input... \n","5 The purpose of adjusting weights in artificial... \n","6 The role of artificial neurons in neural netwo... \n","7 The complexities of neural networks are determ... \n","8 A key difference between machine learning and ... \n","9 Deep learning is a branch of machine learning ... \n","10 The key characteristic of deep learning is the... \n","11 Deep learning has achieved significant success... \n","12 Training deep neural networks typically requir... \n","13 Deep learning can be used for supervised, unsu... \n","14 Convolutional Neural Networks (CNNs) are speci... \n","15 Object detection and recognition is the applic... \n","16 Sentiment analysis is the application of deep ... \n","17 Robotics is the application of deep learning i... \n","18 The main types of neural networks used in deep... \n","19 Artificial Intelligence is the mechanism to in... \n","20 The main difference between Machine Learning a... \n","21 Artificial Intelligence is the broader family ... \n","22 The aim of Machine Learning is to increase acc... \n","23 The aim of Deep Learning is to increase chance... \n","24 AI is the broader concept that encompasses the... \n","25 Speech recognition is an example of AI applica... \n","26 Personalized recommendations, as an AI applica... \n","27 AI-powered predictive maintenance systems anal... \n","28 AI-powered medical diagnosis systems analyze m... \n","29 AI, ML, and DL are related but distinct concep... \n","30 One of the key responsibilities of an AI Engin... \n","31 A Machine Learning Engineer should have a stro... \n","32 One of the key tasks of a Deep Learning Engine... \n","33 Deep Learning (DL) is a subset of Machine Lear... \n","34 One of the advantages of Artificial Neural Net... \n","35 One of the disadvantages of Artificial Neural ... \n","36 One of the advantages of Biological Neural Net... \n","37 One of the disadvantages of Biological Neural ... \n","38 One of the differences between Artificial Neur... \n","39 Hyperparameter tuning is the process of select... \n","40 In neural networks, the learning rate is a hyp... "],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the use of deep learning in g...What is the role of deep learning in generativ...Analyzing sensor data in autonomous vehiclesCreating new content based on existing dataTranscribing spoken words into textMaking personalized recommendations based on u...BDeep learning algorithms are used in generativ...
1Create an MCQ on the use of deep learning in a...How is deep learning used in autonomous vehicles?Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...ADeep learning algorithms are used in autonomou...
2Create an MCQ on the application of deep learn...What is the role of deep learning in speech re...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesTranscribing spoken words into textMaking personalized recommendations based on u...CDeep learning algorithms are used in speech re...
3Create an MCQ on the use of deep learning in r...What is the application of deep learning in re...Analyzing sensor data in autonomous vehiclesRecognizing objects and scenes in imagesMaking personalized recommendations based on u...Transcribing spoken words into textCDeep learning algorithms are used in recommend...
4Create an MCQ on the structure of artificial n...What is the structure of an artificial neural ...It consists of input layers, hidden layers, an...It consists of input layers and output layers ...It consists of input layers, hidden layers, ou...It consists of input layers and artificial neu...AAn artificial neural network consists of input...
5Create an MCQ on the purpose of adjusting weig...What is the purpose of adjusting weights in ar...To increase the speed of training modelsTo prevent overfitting by validating the model...To enhance the model's performance on the trai...To reduce the computational cost of trainingCThe purpose of adjusting weights in artificial...
6Create an MCQ on the role of artificial neuron...What is the role of artificial neurons in neur...To receive input from external sourcesTo compute the weighted total of inputsTo transfer information to the next layerAll of the aboveDThe role of artificial neurons in neural netwo...
7Create an MCQ on the complexities of neural ne...What determines the complexities of neural net...The number of layers in the networkThe number of units in each layerThe type of activation function usedThe size of the datasetBThe complexities of neural networks are determ...
8Create an MCQ on the difference between machin...What is a key difference between machine learn...The type of algorithms usedThe amount of data requiredThe complexity of the modelsThe performance on complex tasksBA key difference between machine learning and ...
9Create an MCQ on the definition of deep learningWhat is the definition of deep learning?A branch of machine learning that uses artific...A type of programming that explicitly defines ...A technique that requires manual feature engin...A method of machine learning that only works w...ADeep learning is a branch of machine learning ...
10Create an MCQ on the key characteristic of dee...What is the key characteristic of deep learning?The use of shallow neural networks with a sing...The requirement for manual feature engineeringThe use of deep neural networks with multiple ...The reliance on labeled datasets for trainingCThe key characteristic of deep learning is the...
11Create an MCQ on the applications of deep lear...Which of the following are applications of dee...Image recognition, natural language processing...Data clustering, dimensionality reduction, and...Supervised machine learning and unsupervised m...Data visualization and exploratory data analysisADeep learning has achieved significant success...
12Create an MCQ on the training requirements of ...What are the training requirements for deep ne...Small datasets and limited computational resou...Large amounts of data and computational resourcesManual feature engineering and domain expertisePre-trained models and transfer learningBTraining deep neural networks typically requir...
13Create an MCQ on the types of machine learning...Which types of machine learning tasks can be p...Supervised machine learning onlyUnsupervised machine learning onlyReinforcement machine learning onlySupervised, unsupervised, and reinforcement ma...DDeep learning can be used for supervised, unsu...
14Create an MCQ on the types of neural networks ...Which type of neural network is specifically d...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)None of the aboveBConvolutional Neural Networks (CNNs) are speci...
15Create an MCQ on the applications of deep lear...Which application of deep learning in computer...Object detection and recognitionImage classificationImage segmentationNone of the aboveAObject detection and recognition is the applic...
16Create an MCQ on the applications of deep lear...Which application of deep learning in NLP invo...Automatic Text GenerationLanguage translationSentiment analysisSpeech recognitionCSentiment analysis is the application of deep ...
17Create an MCQ on the applications of deep lear...Which application of deep learning in reinforc...Game playingRoboticsControl systemsNone of the aboveBRobotics is the application of deep learning i...
18Create an MCQ on the main types of neural netw...Which of the following are the main types of n...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)All of the aboveDThe main types of neural networks used in deep...
19Create an MCQ on the definition of Artificial ...Which of the following best defines Artificial...The study of training machines to mimic human ...The study of statistical methods enabling mach...The study that uses neural networks to imitate...The study of incorporating human intelligence ...DArtificial Intelligence is the mechanism to in...
20Create an MCQ on the difference between Machin...What is the main difference between Machine Le...Machine Learning uses statistical methods, whi...Machine Learning focuses on learning from expe...Machine Learning is a subset of Deep LearningMachine Learning requires human intervention, ...AThe main difference between Machine Learning a...
21Create an MCQ on the components of Artificial ...Which of the following components are part of ...Machine Learning and Deep LearningMachine Learning and Decision TreesArtificial Intelligence and Machine LearningArtificial Intelligence and Deep LearningCArtificial Intelligence is the broader family ...
22Create an MCQ on the aim of Machine LearningWhat is the aim of Machine Learning?To increase chances of successTo increase accuracyTo improve system efficiencyTo analyze data and provide outputBThe aim of Machine Learning is to increase acc...
23Create an MCQ on the aim of Deep LearningWhat is the aim of Deep Learning?To increase chances of successTo increase accuracyTo improve system efficiencyTo analyze data and provide outputAThe aim of Deep Learning is to increase chance...
24Create an MCQ on the difference between AI, Ma...Which of the following best describes the diff...AI is a subset of Machine Learning, which is a...Machine Learning is a subset of AI, which is a...Deep Learning is a subset of AI, which is a su...AI, Machine Learning, and Deep Learning are co...BAI is the broader concept that encompasses the...
25Create an MCQ on the application of AI in spee...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...CSpeech recognition is an example of AI applica...
26Create an MCQ on the application of AI in pers...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...APersonalized recommendations, as an AI applica...
27Create an MCQ on the application of AI in pred...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Analyzing sensor data to predict equipment fai...Recognizing and classifying images and speechCAI-powered predictive maintenance systems anal...
28Create an MCQ on the application of AI in medi...Which of the following is an example of AI app...Analyzing users' browsing and viewing history ...Analyzing medical images to assist doctors in ...Recognizing and classifying images and speechAnalyzing sensor data to make decisions about ...BAI-powered medical diagnosis systems analyze m...
29Create an MCQ on the difference between AI, ML...Which of the following statements accurately d...AI, ML, and DL are interchangeable terms that ...AI focuses on creating intelligent machines, M...AI is a subset of ML that uses neural networks...AI focuses on developing algorithms that enabl...BAI, ML, and DL are related but distinct concep...
30Create an MCQ on the responsibilities of an AI...Which of the following is a key responsibility...Design and development of AI algorithmsAnalysis and interpretation of dataTraining and evaluation of ML modelsDeployment and maintenance of DL modelsAOne of the key responsibilities of an AI Engin...
31Create an MCQ on the skills required for a Mac...Which of the following skills is essential for...Strong background in computer science, mathema...Experience in developing AI algorithms and sol...Familiarity with programming languages such as...All of the aboveDA Machine Learning Engineer should have a stro...
32Create an MCQ on the tasks of a Deep Learning ...Which of the following is a key task of a Deep...Design and development of DL algorithmsAnalysis and interpretation of dataTraining and evaluation of ML modelsDeployment and maintenance of AI modelsAOne of the key tasks of a Deep Learning Engine...
33Create an MCQ on the difference between ML and DLWhat distinguishes Deep Learning (DL) from Mac...DL is a more advanced form of ML that can perf...DL focuses on developing algorithms that enabl...DL is a subset of ML that uses neural networks...ML is a more advanced form of DL that can perf...CDeep Learning (DL) is a subset of Machine Lear...
34Create an MCQ on the advantages of Artificial ...Which of the following is an advantage of Arti...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...Dependence on hardware for functioningHigh speed of processingAOne of the advantages of Artificial Neural Net...
35Create an MCQ on the disadvantages of Artifici...Which of the following is a disadvantage of Ar...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...Dependence on hardware for functioningThe simplest architecture makes it difficult t...DOne of the disadvantages of Artificial Neural ...
36Create an MCQ on the advantages of Biological ...Which of the following is an advantage of Biol...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...No controlling mechanismAbility to process highly complex parallel inputsDOne of the advantages of Biological Neural Net...
37Create an MCQ on the disadvantages of Biologic...Which of the following is a disadvantage of Bi...Ability to learn irrespective of the type of dataSimple architecture that makes it easy to expl...No controlling mechanismSpeed of processing is slowDOne of the disadvantages of Biological Neural ...
38Create an MCQ on the differences between Artif...Which of the following is a difference between...Both ANNs and BNNs have complex and diverse ne...ANNs have fixed connections between neurons, w...Both ANNs and BNNs have simple and predetermin...ANNs and BNNs have the same processing speedBOne of the differences between Artificial Neur...
39Create an MCQ on hyperparameter tuning in mach...What is the purpose of hyperparameter tuning i...To adjust the weights and biases of the modelTo select the optimal values for the model's h...To preprocess the input data before training t...To evaluate the performance of the model on a ...BHyperparameter tuning is the process of select...
40Create an MCQ on the types of hyperparameters ...Which of the following is a type of hyperparam...WeightsBiasesLearning rateActivation functionCIn neural networks, the learning rate is a hyp...
\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['test']\",\n \"rows\": 41,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Create an MCQ on the difference between AI, Machine Learning, and Deep Learning\",\n \"Create an MCQ on the types of machine learning tasks that can be performed using deep learning\",\n \"Create an MCQ on the difference between machine learning and deep learning\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Which of the following best describes the difference between AI, Machine Learning, and Deep Learning?\",\n \"Which types of machine learning tasks can be performed using deep learning?\",\n \"What is a key difference between machine learning and deep learning?\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Both ANNs and BNNs have complex and diverse neurons\",\n \"The study of training machines to mimic human behavior\",\n \"Strong background in computer science, mathematics, and statistics\"\n ],\n \"num_unique_values\": 30,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"ANNs have fixed connections between neurons, while BNNs have flexible connections\",\n \"Robotics\",\n \"Analysis and interpretation of data\"\n ],\n \"num_unique_values\": 30,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Both ANNs and BNNs have simple and predetermined neural pathways\",\n \"Control systems\",\n \"Training and evaluation of ML models\"\n ],\n \"num_unique_values\": 32,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"ANNs and BNNs have the same processing speed\",\n \"Machine Learning requires human intervention, while Deep Learning does not\",\n \"ML is a more advanced form of DL that can perform complex tasks.\"\n ],\n \"num_unique_values\": 32,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"A\",\n \"D\",\n \"B\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"AI is the broader concept that encompasses the development of computer systems that can perform tasks requiring human intelligence. Machine Learning is a subset of AI, focused on algorithms that can learn from data and make predictions or decisions. Deep Learning is a subset of Machine Learning, specifically using neural networks with multiple layers to learn and represent complex patterns.\",\n \"Deep learning can be used for supervised, unsupervised, as well as reinforcement machine learning tasks. It provides a versatile approach to process and learn from data in various learning scenarios.\",\n \"A key difference between machine learning and deep learning is the amount of data required. Machine learning can work with a smaller amount of data, while deep learning requires a larger volume of data to train the complex neural network architectures effectively.\"\n ],\n \"num_unique_values\": 41,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":20}]},{"cell_type":"code","source":["# Read as pandas DataFrame\n","dataset['val'].to_pandas()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"03oFFWL4hM4S","executionInfo":{"status":"ok","timestamp":1708322802115,"user_tz":-480,"elapsed":16,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"284e74e0-5907-4982-cd82-31dfc59696e3"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" Instruction \\\n","0 Create an MCQ on the impact of learning rate o... \n","1 Create an MCQ on the impact of number of epoch... \n","2 Create an MCQ on the impact of architecture on... \n","3 Create an MCQ on the strategies for hyperparam... \n","4 Create an MCQ on the drawback of GridSearchCV \n","5 Create an MCQ on the strategy that selects val... \n","6 Create an MCQ on the advantage of RandomizedSe... \n","7 Create an MCQ on the strategy that treats hype... \n","8 Create an MCQ on the challenges in deep learning \n","9 Create an MCQ on the advantages of deep learning \n","10 Create an MCQ on the disadvantages of deep lea... \n","11 Create an MCQ on the challenges in interpretin... \n","12 Create an MCQ on the risk of overfitting in de... \n","13 Create an MCQ on the application of machine le... \n","14 Create an MCQ on the use of machine learning i... \n","15 Create an MCQ on the role of machine learning ... \n","16 Create an MCQ on the application of machine le... \n","17 Create an MCQ on the use of machine learning i... \n","18 Create an MCQ on the regularization parameter ... \n","19 Create an MCQ on the kernel function in SVMs \n","20 Create an MCQ on the parameter that controls t... \n","21 Create an MCQ on the learning rate hyperparame... \n","22 Create an MCQ on the max_depth hyperparameter ... \n","23 Create an MCQ on the structure of artificial n... \n","24 Create an MCQ on the training process of artif... \n","25 Create an MCQ on the differences between machi... \n","26 Create an MCQ on the complexity of interpretin... \n","27 Create an MCQ on the computing power requireme... \n","28 Create an MCQ on the definition of deep learning \n","29 Create an MCQ on the key characteristic of dee... \n","30 Create an MCQ on the success of deep learning ... \n","31 Create an MCQ on the requirements for training... \n","32 Create an MCQ on the types of machine learning... \n","33 Create an MCQ on the types of neural networks ... \n","34 Create an MCQ on the applications of deep lear... \n","35 Create an MCQ on the applications of deep lear... \n","36 Create an MCQ on the applications of deep lear... \n","37 Create an MCQ on the main purpose of deep lear... \n","38 Create an MCQ on the definition of Artificial ... \n","39 Create an MCQ on the definition of Machine Lea... \n","40 Create an MCQ on the definition of Deep Learning. \n","41 Create an MCQ on the relationship between Arti... \n","42 Create an MCQ on the aim of Deep Learning. \n","\n"," Question \\\n","0 What impact does the learning rate hyperparame... \n","1 How does the number of epochs hyperparameter a... \n","2 How does the architecture of a neural network ... \n","3 Which of the following strategies is considere... \n","4 What is a drawback of using GridSearchCV for h... \n","5 Which hyperparameter tuning strategy selects v... \n","6 What is an advantage of using RandomizedSearch... \n","7 Which hyperparameter tuning strategy treats th... \n","8 What is one of the challenges in deep learning? \n","9 What is one of the advantages of deep learning? \n","10 What is one of the disadvantages of deep learn... \n","11 What is one of the challenges in interpreting ... \n","12 What is the risk associated with overfitting i... \n","13 In which of the following applications are mac... \n","14 Which of the following applications utilize ma... \n","15 Which of the following applications involve th... \n","16 Which of the following applications utilize ma... \n","17 In which of the following applications are mac... \n","18 What is the role of the regularization paramet... \n","19 What is the purpose of the kernel function in ... \n","20 Which parameter controls the influence of supp... \n","21 What does the learning rate hyperparameter det... \n","22 What does the max_depth hyperparameter determi... \n","23 Which layer of an artificial neural network re... \n","24 What is adjusted during the training process o... \n","25 Which of the following requires a larger volum... \n","26 Which of the following is true regarding the i... \n","27 Which of the following requires a high-perform... \n","28 What is deep learning? \n","29 What is the key characteristic of deep learning? \n","30 In which fields has deep learning achieved sig... \n","31 What are the requirements for training deep ne... \n","32 Which types of machine learning are used in de... \n","33 Which type of neural network is specifically d... \n","34 What is one of the main applications of deep l... \n","35 What is one of the main applications of deep l... \n","36 What is one of the main applications of deep l... \n","37 What is the main purpose of deep learning mode... \n","38 What is the definition of Artificial Intellige... \n","39 What is the definition of Machine Learning? \n","40 What is the definition of Deep Learning? \n","41 What is the relationship between Artificial In... \n","42 What is the aim of Deep Learning? \n","\n"," A \\\n","0 It determines the number of epochs needed for ... \n","1 Increasing the number of epochs always improve... \n","2 The architecture determines the learning rate ... \n","3 GridSearchCV \n","4 It is computationally expensive \n","5 GridSearchCV \n","6 It is computationally faster \n","7 GridSearchCV \n","8 Limited computational resources \n","9 Low accuracy \n","10 Low computational requirements \n","11 Easy interpretability \n","12 Improved performance on new data \n","13 Self-driving cars \n","14 Virtual assistants like Siri and Alexa \n","15 Chatbots \n","16 E-commerce sites \n","17 Social media monitoring \n","18 To control the trade-off between the margin an... \n","19 To control the trade-off between the margin an... \n","20 Regularization parameter (C) \n","21 The step size taken by the optimizer during ea... \n","22 The step size taken by the optimizer during ea... \n","23 Output layer \n","24 Weights \n","25 Machine learning \n","26 Machine learning results are easy to interpret \n","27 Machine learning \n","28 A branch of machine learning that uses artific... \n","29 The use of deep neural networks with multiple ... \n","30 Image recognition, natural language processing... \n","31 A large amount of data and computational resou... \n","32 Supervised, unsupervised, and reinforcement le... \n","33 Feedforward Neural Networks (FNNs) \n","34 Speech recognition \n","35 Object detection and recognition \n","36 Sentiment analysis \n","37 To analyze the sentiment of text \n","38 The study of training machines to mimic human ... \n","39 The study of training machines to mimic human ... \n","40 The study of training machines to mimic human ... \n","41 AI is a subset of ML \n","42 To increase chances of success \n","\n"," B \\\n","0 It controls the step size taken by the optimiz... \n","1 Increasing the number of epochs can lead to ov... \n","2 The architecture controls the step size taken ... \n","3 RandomizedSearchCV \n","4 It requires expert knowledge \n","5 RandomizedSearchCV \n","6 It guarantees optimal performance \n","7 RandomizedSearchCV \n","8 Easy interpretability of results \n","9 Manual feature engineering \n","10 Small amount of labeled data \n","11 Clear decision-making process \n","12 No impact on model performance \n","13 Security systems \n","14 Call centers \n","15 Virtual assistants \n","16 Streaming services \n","17 Sentiment analysis systems \n","18 To define the similarity between data points \n","19 To define the similarity between data points \n","20 Kernel function \n","21 The number of boosting trees to be trained \n","22 The number of boosting trees to be trained \n","23 Hidden layer \n","24 Layers \n","25 Deep learning \n","26 Deep learning results are easy to interpret \n","27 Deep learning \n","28 A technique in machine learning that involves ... \n","29 The use of decision trees for modeling \n","30 Clustering, dimensionality reduction, and anom... \n","31 The availability of cloud computing and specia... \n","32 Supervised and unsupervised learning \n","33 Convolutional Neural Networks (CNNs) \n","34 Sentiment analysis \n","35 Image classification \n","36 Image segmentation \n","37 To translate text from one language to another \n","38 The study of improving machines with experienc... \n","39 The study of improving machines with experienc... \n","40 The study of improving machines with experienc... \n","41 ML is a subset of AI \n","42 To increase accuracy \n","\n"," C \\\n","0 It determines the depth of the neural network \n","1 The number of epochs does not have any impact ... \n","2 The architecture determines the depth and widt... \n","3 Bayesian Optimization \n","4 It may result in overfitting \n","5 Bayesian Optimization \n","6 It requires less expertise \n","7 Bayesian Optimization \n","8 Small amount of training data \n","9 Limited scalability \n","10 Easy interpretability \n","11 Limited complexity \n","12 Poor performance on new data \n","13 Medical imaging \n","14 Speech recognition systems \n","15 NLP systems \n","16 Recommendation systems \n","17 Spam filters \n","18 To determine the influence of support vectors ... \n","19 To determine the influence of support vectors ... \n","20 Gamma \n","21 The maximum depth of each tree in the ensemble \n","22 The maximum depth of each tree in the ensemble \n","23 Input layer \n","24 Neurons \n","25 Both require the same volume of dataset \n","26 Both machine learning and deep learning result... \n","27 Both require the same computing power \n","28 A type of unsupervised machine learning that c... \n","29 The use of reinforcement learning algorithms \n","30 Supervised machine learning tasks like image c... \n","31 Manual feature engineering \n","32 Unsupervised and reinforcement learning \n","33 Recurrent Neural Networks (RNNs) \n","34 Image classification \n","35 Language translation \n","36 Game playing \n","37 To identify and understand visual data \n","38 The study that uses neural networks to imitate... \n","39 The study that uses neural networks to imitate... \n","40 The study that uses neural networks to imitate... \n","41 AI and ML are independent of each other \n","42 To achieve high accuracy with a small amount o... \n","\n"," D Correct Answer \\\n","0 It controls the width of the neural network B \n","1 The number of epochs determines the learning r... B \n","2 The architecture affects the activation functi... C \n","3 None of the above A \n","4 It is not effective for high-dimensional hyper... A \n","5 None of the above B \n","6 It is more effective for high-dimensional hype... A \n","7 None of the above C \n","8 No risk of overfitting A \n","9 Continual improvement D \n","10 No risk of overfitting B \n","11 Black box nature D \n","12 Reduced computational requirements C \n","13 All of the above D \n","14 All of the above D \n","15 All of the above D \n","16 All of the above D \n","17 All of the above D \n","18 To determine the maximum depth of each tree in... A \n","19 To determine the maximum depth of each tree in... B \n","20 Learning rate C \n","21 The minimum sum of instance weight needed in a... A \n","22 The minimum sum of instance weight needed in a... C \n","23 Final layer C \n","24 Connections A \n","25 Neither requires a dataset B \n","26 Deep learning results are difficult to interpret D \n","27 Neither requires computing power B \n","28 A method in machine learning that uses decisio... A \n","29 The use of unsupervised learning techniques A \n","30 Reinforcement learning tasks like robotics and... A \n","31 Supervised labeled datasets A \n","32 Supervised and reinforcement learning A \n","33 Artificial Neural Networks (ANNs) B \n","34 Game playing C \n","35 Game playing C \n","36 Speech recognition C \n","37 To control complex systems C \n","38 The study that focuses on learning, reasoning,... A \n","39 The study that focuses on learning from data a... D \n","40 The study that focuses on using neural network... D \n","41 AI and ML are the same thing B \n","42 To achieve high accuracy with a large amount o... D \n","\n"," Explanation \n","0 The learning rate hyperparameter controls the ... \n","1 Increasing the number of epochs can improve th... \n","2 The architecture of a neural network determine... \n","3 GridSearchCV is considered a 'brute force' app... \n","4 GridSearchCV is computationally expensive as i... \n","5 RandomizedSearchCV selects values at random fo... \n","6 RandomizedSearchCV is computationally faster t... \n","7 Bayesian Optimization treats the search for op... \n","8 One of the challenges in deep learning is the ... \n","9 One of the advantages of deep learning is its ... \n","10 One of the disadvantages of deep learning is t... \n","11 One of the challenges in interpreting deep lea... \n","12 The risk associated with overfitting in deep l... \n","13 Machine learning algorithms are used in image ... \n","14 Machine learning algorithms are used in speech... \n","15 Machine learning algorithms are used in NLP sy... \n","16 Machine learning algorithms are used in recomm... \n","17 Machine learning algorithms are used in sentim... \n","18 The regularization parameter (C) in SVMs contr... \n","19 The kernel function in SVMs defines the simila... \n","20 The parameter Gamma controls the influence of ... \n","21 The learning rate hyperparameter in XGBoost de... \n","22 The max_depth hyperparameter in XGBoost determ... \n","23 The input layer of an artificial neural networ... \n","24 During the training process of an artificial n... \n","25 Deep learning requires a larger volume of data... \n","26 Deep learning results are more complex and dif... \n","27 Deep learning requires a high-performance comp... \n","28 Deep learning is a branch of machine learning ... \n","29 The key characteristic of deep learning is the... \n","30 Deep learning has achieved significant success... \n","31 Training deep neural networks typically requir... \n","32 Deep learning can be used for supervised, unsu... \n","33 Convolutional Neural Networks (CNNs) are speci... \n","34 One of the main applications of deep learning ... \n","35 One of the main applications of deep learning ... \n","36 One of the main applications of deep learning ... \n","37 The main purpose of deep learning models in co... \n","38 Artificial Intelligence is the mechanism to in... \n","39 Machine Learning is the study/process which pr... \n","40 Deep Learning is a sub-part of the broader fam... \n","41 Artificial Intelligence is the broader family ... \n","42 The aim of Deep Learning is to achieve high ac... "],"text/html":["\n","
\n","
\n","\n","\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
InstructionQuestionABCDCorrect AnswerExplanation
0Create an MCQ on the impact of learning rate o...What impact does the learning rate hyperparame...It determines the number of epochs needed for ...It controls the step size taken by the optimiz...It determines the depth of the neural networkIt controls the width of the neural networkBThe learning rate hyperparameter controls the ...
1Create an MCQ on the impact of number of epoch...How does the number of epochs hyperparameter a...Increasing the number of epochs always improve...Increasing the number of epochs can lead to ov...The number of epochs does not have any impact ...The number of epochs determines the learning r...BIncreasing the number of epochs can improve th...
2Create an MCQ on the impact of architecture on...How does the architecture of a neural network ...The architecture determines the learning rate ...The architecture controls the step size taken ...The architecture determines the depth and widt...The architecture affects the activation functi...CThe architecture of a neural network determine...
3Create an MCQ on the strategies for hyperparam...Which of the following strategies is considere...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveAGridSearchCV is considered a 'brute force' app...
4Create an MCQ on the drawback of GridSearchCVWhat is a drawback of using GridSearchCV for h...It is computationally expensiveIt requires expert knowledgeIt may result in overfittingIt is not effective for high-dimensional hyper...AGridSearchCV is computationally expensive as i...
5Create an MCQ on the strategy that selects val...Which hyperparameter tuning strategy selects v...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveBRandomizedSearchCV selects values at random fo...
6Create an MCQ on the advantage of RandomizedSe...What is an advantage of using RandomizedSearch...It is computationally fasterIt guarantees optimal performanceIt requires less expertiseIt is more effective for high-dimensional hype...ARandomizedSearchCV is computationally faster t...
7Create an MCQ on the strategy that treats hype...Which hyperparameter tuning strategy treats th...GridSearchCVRandomizedSearchCVBayesian OptimizationNone of the aboveCBayesian Optimization treats the search for op...
8Create an MCQ on the challenges in deep learningWhat is one of the challenges in deep learning?Limited computational resourcesEasy interpretability of resultsSmall amount of training dataNo risk of overfittingAOne of the challenges in deep learning is the ...
9Create an MCQ on the advantages of deep learningWhat is one of the advantages of deep learning?Low accuracyManual feature engineeringLimited scalabilityContinual improvementDOne of the advantages of deep learning is its ...
10Create an MCQ on the disadvantages of deep lea...What is one of the disadvantages of deep learn...Low computational requirementsSmall amount of labeled dataEasy interpretabilityNo risk of overfittingBOne of the disadvantages of deep learning is t...
11Create an MCQ on the challenges in interpretin...What is one of the challenges in interpreting ...Easy interpretabilityClear decision-making processLimited complexityBlack box natureDOne of the challenges in interpreting deep lea...
12Create an MCQ on the risk of overfitting in de...What is the risk associated with overfitting i...Improved performance on new dataNo impact on model performancePoor performance on new dataReduced computational requirementsCThe risk associated with overfitting in deep l...
13Create an MCQ on the application of machine le...In which of the following applications are mac...Self-driving carsSecurity systemsMedical imagingAll of the aboveDMachine learning algorithms are used in image ...
14Create an MCQ on the use of machine learning i...Which of the following applications utilize ma...Virtual assistants like Siri and AlexaCall centersSpeech recognition systemsAll of the aboveDMachine learning algorithms are used in speech...
15Create an MCQ on the role of machine learning ...Which of the following applications involve th...ChatbotsVirtual assistantsNLP systemsAll of the aboveDMachine learning algorithms are used in NLP sy...
16Create an MCQ on the application of machine le...Which of the following applications utilize ma...E-commerce sitesStreaming servicesRecommendation systemsAll of the aboveDMachine learning algorithms are used in recomm...
17Create an MCQ on the use of machine learning i...In which of the following applications are mac...Social media monitoringSentiment analysis systemsSpam filtersAll of the aboveDMachine learning algorithms are used in sentim...
18Create an MCQ on the regularization parameter ...What is the role of the regularization paramet...To control the trade-off between the margin an...To define the similarity between data pointsTo determine the influence of support vectors ...To determine the maximum depth of each tree in...AThe regularization parameter (C) in SVMs contr...
19Create an MCQ on the kernel function in SVMsWhat is the purpose of the kernel function in ...To control the trade-off between the margin an...To define the similarity between data pointsTo determine the influence of support vectors ...To determine the maximum depth of each tree in...BThe kernel function in SVMs defines the simila...
20Create an MCQ on the parameter that controls t...Which parameter controls the influence of supp...Regularization parameter (C)Kernel functionGammaLearning rateCThe parameter Gamma controls the influence of ...
21Create an MCQ on the learning rate hyperparame...What does the learning rate hyperparameter det...The step size taken by the optimizer during ea...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleThe minimum sum of instance weight needed in a...AThe learning rate hyperparameter in XGBoost de...
22Create an MCQ on the max_depth hyperparameter ...What does the max_depth hyperparameter determi...The step size taken by the optimizer during ea...The number of boosting trees to be trainedThe maximum depth of each tree in the ensembleThe minimum sum of instance weight needed in a...CThe max_depth hyperparameter in XGBoost determ...
23Create an MCQ on the structure of artificial n...Which layer of an artificial neural network re...Output layerHidden layerInput layerFinal layerCThe input layer of an artificial neural networ...
24Create an MCQ on the training process of artif...What is adjusted during the training process o...WeightsLayersNeuronsConnectionsADuring the training process of an artificial n...
25Create an MCQ on the differences between machi...Which of the following requires a larger volum...Machine learningDeep learningBoth require the same volume of datasetNeither requires a datasetBDeep learning requires a larger volume of data...
26Create an MCQ on the complexity of interpretin...Which of the following is true regarding the i...Machine learning results are easy to interpretDeep learning results are easy to interpretBoth machine learning and deep learning result...Deep learning results are difficult to interpretDDeep learning results are more complex and dif...
27Create an MCQ on the computing power requireme...Which of the following requires a high-perform...Machine learningDeep learningBoth require the same computing powerNeither requires computing powerBDeep learning requires a high-performance comp...
28Create an MCQ on the definition of deep learningWhat is deep learning?A branch of machine learning that uses artific...A technique in machine learning that involves ...A type of unsupervised machine learning that c...A method in machine learning that uses decisio...ADeep learning is a branch of machine learning ...
29Create an MCQ on the key characteristic of dee...What is the key characteristic of deep learning?The use of deep neural networks with multiple ...The use of decision trees for modelingThe use of reinforcement learning algorithmsThe use of unsupervised learning techniquesAThe key characteristic of deep learning is the...
30Create an MCQ on the success of deep learning ...In which fields has deep learning achieved sig...Image recognition, natural language processing...Clustering, dimensionality reduction, and anom...Supervised machine learning tasks like image c...Reinforcement learning tasks like robotics and...ADeep learning has achieved significant success...
31Create an MCQ on the requirements for training...What are the requirements for training deep ne...A large amount of data and computational resou...The availability of cloud computing and specia...Manual feature engineeringSupervised labeled datasetsATraining deep neural networks typically requir...
32Create an MCQ on the types of machine learning...Which types of machine learning are used in de...Supervised, unsupervised, and reinforcement le...Supervised and unsupervised learningUnsupervised and reinforcement learningSupervised and reinforcement learningADeep learning can be used for supervised, unsu...
33Create an MCQ on the types of neural networks ...Which type of neural network is specifically d...Feedforward Neural Networks (FNNs)Convolutional Neural Networks (CNNs)Recurrent Neural Networks (RNNs)Artificial Neural Networks (ANNs)BConvolutional Neural Networks (CNNs) are speci...
34Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Speech recognitionSentiment analysisImage classificationGame playingCOne of the main applications of deep learning ...
35Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Object detection and recognitionImage classificationLanguage translationGame playingCOne of the main applications of deep learning ...
36Create an MCQ on the applications of deep lear...What is one of the main applications of deep l...Sentiment analysisImage segmentationGame playingSpeech recognitionCOne of the main applications of deep learning ...
37Create an MCQ on the main purpose of deep lear...What is the main purpose of deep learning mode...To analyze the sentiment of textTo translate text from one language to anotherTo identify and understand visual dataTo control complex systemsCThe main purpose of deep learning models in co...
38Create an MCQ on the definition of Artificial ...What is the definition of Artificial Intellige...The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on learning, reasoning,...AArtificial Intelligence is the mechanism to in...
39Create an MCQ on the definition of Machine Lea...What is the definition of Machine Learning?The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on learning from data a...DMachine Learning is the study/process which pr...
40Create an MCQ on the definition of Deep Learning.What is the definition of Deep Learning?The study of training machines to mimic human ...The study of improving machines with experienc...The study that uses neural networks to imitate...The study that focuses on using neural network...DDeep Learning is a sub-part of the broader fam...
41Create an MCQ on the relationship between Arti...What is the relationship between Artificial In...AI is a subset of MLML is a subset of AIAI and ML are independent of each otherAI and ML are the same thingBArtificial Intelligence is the broader family ...
42Create an MCQ on the aim of Deep Learning.What is the aim of Deep Learning?To increase chances of successTo increase accuracyTo achieve high accuracy with a small amount o...To achieve high accuracy with a large amount o...DThe aim of Deep Learning is to achieve high ac...
\n","
\n","
\n","\n","
\n"," \n","\n"," \n","\n"," \n","
\n","\n","\n","
\n"," \n","\n","\n","\n"," \n","
\n","\n","
\n","
\n"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"dataframe","summary":"{\n \"name\": \"dataset['val']\",\n \"rows\": 43,\n \"fields\": [\n {\n \"column\": \"Instruction\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Create an MCQ on the main purpose of deep learning models in computer vision.\",\n \"Create an MCQ on the training process of artificial neural networks\",\n \"Create an MCQ on the differences between machine learning and deep learning\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Question\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"What is the main purpose of deep learning models in computer vision?\",\n \"What is adjusted during the training process of an artificial neural network?\",\n \"Which of the following requires a larger volume of dataset compared to machine learning?\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"A\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"To increase chances of success\",\n \"Chatbots\",\n \"A large amount of data and computational resources\"\n ],\n \"num_unique_values\": 36,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"B\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"To increase accuracy\",\n \"Virtual assistants\",\n \"The availability of cloud computing and specialized hardware\"\n ],\n \"num_unique_values\": 36,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"C\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"Gamma\",\n \"NLP systems\",\n \"It may result in overfitting\"\n ],\n \"num_unique_values\": 37,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"D\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"AI and ML are the same thing\",\n \"Connections\",\n \"Speech recognition\"\n ],\n \"num_unique_values\": 33,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Correct Answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"samples\": [\n \"C\",\n \"D\",\n \"B\"\n ],\n \"num_unique_values\": 4,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Explanation\",\n \"properties\": {\n \"dtype\": \"string\",\n \"samples\": [\n \"The main purpose of deep learning models in computer vision is to identify and understand visual data. Deep learning models can be used to perform tasks such as object detection and recognition, image classification, and image segmentation.\",\n \"During the training process of an artificial neural network, the weights of the connections between neurons are adjusted to enhance the performance of the model.\",\n \"Deep learning requires a larger volume of dataset compared to machine learning.\"\n ],\n \"num_unique_values\": 43,\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"}},"metadata":{},"execution_count":21}]},{"cell_type":"code","source":["from huggingface_hub import notebook_login\n","notebook_login()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":145,"referenced_widgets":["ac3383a4853c484ca009cddb6e305853","095e9de8c9c24583baa8c7d074c5eacc","7a6eab206dcb4cd3b0cd4681a38f192f","e40bf3020d024dc0bc72a6f23b471494","8fc0cb635ac64733b957061f58ed16b6","9dcbf15879f94eddaf99a1603d1ca4c8","b0cf0967450c415f92e8540b30944ccf","e588245da98b412d849fbe2f94fb4b79","58409bb2b4c845f589fce0e2c2078a8c","17a1cd2fe40e4dde8b6b196e29abbec1","10bef54de0f64e86a3c9b1c039885ee9","40c244e062aa47a7846cace18c952cad","0d5efa7a7e7a4fe2ac2a789fe79ea94a","51e3dfdb61374128a65a208209c93060","283deaf2c9544e80a9433ae9148492d5","b9eb5c03173143aeb2793da889a2428b","8a3773ae30cf482a80edc5af98dc9cd8","0e34328104b64acb9a63e76013ccfc0d","e4fb41fce6b44fffbb1e0a6ab9d477ab","b5b3620259184090b78f008fc8a789db","bdec133061344a838b792668b98a5daa","f26867efbb384d1791ccc7e6ff2f7b2b","e35c557212ea4317ba624707f8d42dde","bfee25db776447eb9126cdbdbdfe6a8c","82764a4acf0e49e58547e9f2e52d5533","ab32080c5ec54c8abb90f0d85855adf5","5eacd5686b6d4486b772e957a80d894d","6c6af529b3dd449cb2826706c220d0cc","079da2bd7cfd4ed3b3324a2e54a8ae87","570cfe79345a43b5a9b34f05e72cc6c9","b21ff34c2fd844f2baee7dd45eda5700","8f7bf4ff652b4ae2b137b0518a4d6a00"]},"id":"U2YZULfoCura","executionInfo":{"status":"ok","timestamp":1708322802639,"user_tz":-480,"elapsed":9,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"31a4d102-5e31-4500-c9bc-34d0274857cf"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["VBox(children=(HTML(value='
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":["## 3. Filter out rows with more than 2048 tokens\n","\n","We will remove samples with more than 2048 tokens (max context size of Llama 2 by default = 4096)."],"metadata":{"id":"_RXe958fNLwH"}},{"cell_type":"code","source":["def filter_by_token_count(dataset_split, combined_token_counts, max_tokens=2048):\n"," # Filter out rows with more than 'max_tokens' tokens\n"," filtered_dataset = [example for example, count in zip(dataset_split, combined_token_counts) if count <= max_tokens]\n"," return filtered_dataset\n","\n","\n","# Assuming 'dataset' contains your data splits\n","fig, axs = plt.subplots(3, 5, figsize=(25, 15)) # Adjust figure size as necessary\n","\n","split_names = ['train', 'test', 'val']\n","for row, split_name in enumerate(split_names):\n"," # Tokenize and count\n"," instruction_counts, explanation_counts, question_counts, options_counts, combined_counts = tokenize_and_count(dataset[split_name])\n","\n"," # Filter dataset based on combined token count\n"," filtered_dataset = filter_by_token_count(dataset[split_name], combined_counts)\n","\n"," # Re-tokenize and count for the filtered dataset\n"," filtered_instruction_counts, filtered_explanation_counts, filtered_question_counts, filtered_options_counts, filtered_combined_counts = tokenize_and_count(filtered_dataset)\n","\n"," # Plotting the distributions for the filtered datasets, organizing by row based on the split\n"," plot_distribution(filtered_instruction_counts, f\"{split_name} (filtered): Instruction\", axs[row, 0])\n"," plot_distribution(filtered_explanation_counts, f\"{split_name} (filtered): Explanation\", axs[row, 1])\n"," plot_distribution(filtered_question_counts, f\"{split_name} (filtered): Question\", axs[row, 2])\n"," plot_distribution(filtered_options_counts, f\"{split_name} (filtered): Options\", axs[row, 3])\n"," plot_distribution(filtered_combined_counts, f\"{split_name} (filtered): Combined\", axs[row, 4])\n","\n","# Adjust layout to prevent overlap and ensure clarity\n","plt.tight_layout(pad=3.0)\n","plt.show()\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":898},"id":"WYzaKhSvz9Yk","executionInfo":{"status":"ok","timestamp":1708322857920,"user_tz":-480,"elapsed":8192,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"ee8da4e6-3afb-4534-885b-9780571cde25"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Function to filter dataset and plot new distribution\n","def filter_and_plot(dataset_split_name, dataset_split_data, combined_token_counts, axs, position):\n"," # Filter out rows with more than 2048 tokens\n"," valid_indices = [i for i, count in enumerate(combined_token_counts) if count <= 2048]\n"," print(f\"Number of valid rows in {dataset_split_name}: {len(valid_indices)}\")\n"," print(f\"Removing {len(dataset_split_data) - len(valid_indices)} rows from {dataset_split_name}...\")\n","\n"," # Extract valid rows based on indices\n"," valid_dataset = [dataset_split_data[i] for i in valid_indices]\n","\n"," # Re-calculate token counts for the valid dataset if necessary\n"," # This step is assumed necessary only if the token counts need to be recalculated for the filtered dataset\n"," # Otherwise, valid_token_counts = [combined_token_counts[i] for i in valid_indices] would suffice\n"," _, _, _, _, valid_combined_counts = tokenize_and_count(valid_dataset)\n","\n"," # Plot the new distribution for valid rows\n"," plot_distribution(valid_combined_counts, f\"New distribution after filtering {dataset_split_name}\", axs[position])\n","\n","# Create a figure with subplots\n","fig, axs = plt.subplots(3, 1, figsize=(6, 9)) # Adjust figsize as necessary\n","\n","# Assuming the 'dataset' variable is a dictionary containing data splits 'train', 'test', and 'val'\n","for i, split_name in enumerate(['train', 'test', 'val']):\n"," # Tokenize and count for the specific dataset split\n"," _, _, _, _, combined_counts = tokenize_and_count(dataset[split_name])\n","\n"," # Filter datasets based on token count and plot the new distribution\n"," filter_and_plot(split_name, dataset[split_name], combined_counts, axs, i)\n","\n","plt.tight_layout()\n","plt.show()\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":1000},"id":"LTw4ITHDl92J","executionInfo":{"status":"ok","timestamp":1708322864043,"user_tz":-480,"elapsed":2649,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"9f5de7d2-6e9a-4e2e-d3ee-73df3a6ad59f"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Number of valid rows in train: 334\n","Removing 0 rows from train...\n","Number of valid rows in test: 41\n","Removing 0 rows from test...\n","Number of valid rows in val: 43\n","Removing 0 rows from val...\n"]},{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Initialize a flag to indicate whether any entries were removed in any split\n","entries_removed = False\n","\n","# Iterate over each split in the dataset\n","for split_name in ['train', 'test', 'val']:\n"," # Get the original length of the split\n"," original_length = len(dataset[split_name])\n"," # Tokenize and count tokens in the split\n"," _, _, _, _, combined_counts = tokenize_and_count(dataset[split_name])\n"," # Determine valid indices (entries with <= 2048 tokens)\n"," valid_indices = [i for i, count in enumerate(combined_counts) if count <= 2048]\n"," # Check if any entries were removed\n"," if len(valid_indices) < original_length:\n"," entries_removed = True\n"," # Update the dataset split with filtered entries\n"," dataset[split_name] = dataset[split_name].select(valid_indices)\n","\n","# Flag to control execution of subsequent code\n","continue_execution = True\n","\n","if not entries_removed:\n"," print(\"No entries removed due to token count. Skipping saving.\")\n"," continue_execution = False\n","\n","# Proceed with further steps only if entries were removed\n","if continue_execution:\n"," # Save the filtered dataset to disk\n"," dataset.save_to_disk('new_mcq_data')\n"," print(\"Filtered dataset saved successfully.\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2N8lcoTS5jkm","executionInfo":{"status":"ok","timestamp":1708322873795,"user_tz":-480,"elapsed":922,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"d1028a6f-0388-47ea-8486-e715bcbc52fd"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["No entries removed due to token count. Skipping saving.\n"]}]},{"cell_type":"markdown","source":["---\n","\n","## 4. Near-deduplication Using Embeddings\n","\n","* Near-deduplication with embeddings is a technique that employs vector representations to effectively identify and manage nearly identical data entries.\n","\n","* By transforming data into these vectors (embeddings), we can quantitatively measure how similar different pieces of data are. This transformation significantly improves our ability to manage large datasets, where sorting through and removing near-duplicates manually would be impractical.\n","\n","* Widely used in fields like database management, information retrieval, and machine learning, this approach is crucial for efficient data handling and analysis.\n","\n","---\n","\n","### We Will Not Perform Deduplication on Our MCQ Dataset.\n","\n","* **Intentional Repetition for Emphasis**: In educational contexts, certain concepts may be intentionally repeated to underscore their significance. Deduplication could diminish the dataset's educational effectiveness by removing these purposeful repetitions.\n","\n","* **Variations of Similar Questions**: MCQ datasets often feature questions that, while seemingly similar, include minor variations in wording, options, or context. Inadequately designed deduplication algorithms risk eliminating these nuances, thereby losing valuable elements of the dataset.\n","\n","* **Difficulty in Defining \"Duplicates\"**: Identifying duplicates within MCQs poses a significant challenge, as questions that appear identical might differ in subtle yet crucial ways. These distinctions often represent unique learning opportunities that would be lost through deduplication.\n","\n","---\n"],"metadata":{"id":"kXzTKu99w3g4"}},{"cell_type":"markdown","source":["## 5. Top-k sampling\n","\n","Only keep the top k samples with the most tokens.\n","\n","---\n","\n","### Decision on \"Top-k Sampling\" for Our MCQ Dataset\n","\n","\n","We have decided against employing \"Top-k sampling\" to select only the top k samples with the most tokens in our MCQ dataset. This approach does not align with the core objectives of MCQ dataset development for several critical reasons:\n","\n","\n","**Practical Considerations**\n","\n","* **Conciseness and Effectiveness**: The hallmark of high-quality MCQs lies in their conciseness and meaningfulness. Favoring question length over substance could detract from the dataset's quality, as longer questions do not necessarily equate to higher educational value. Succinct yet profound questions are typically the most beneficial and stimulating for learners.\n","\n","\n","* Given these considerations, we conclude that \"Top-k sampling,\" which prioritizes token count, falls short of fulfilling the requirements of our MCQ dataset. The true merit of a valuable MCQ dataset resides in its diverse and balanced assortment of topics and difficulty levels, not merely in question length. This philosophy ensures our dataset remains versatile and effective across various educational and machine learning applications.\n","---"],"metadata":{"id":"TOCspcgXNOav"}},{"cell_type":"code","source":["# @title\n","# Push to Hugging Face Hub\n","dataset.push_to_hub(\"ssoh/mcq_dataset_2\")"],"metadata":{"id":"pj1b5S_68KB0","colab":{"base_uri":"https://localhost:8080/","height":244,"referenced_widgets":["d245dbe5cf8946e28f08287c15926a6d","45f5b19c94a544d79b3d65e13e67a0f1","ecb708463202466b96b244822fddd7c3","9cc0c38f4fb4452e8a14f1b208aaaefd","9ec8035efb334973bf54923d9b25b491","60fc22bbcb204b008cf416f44835ab84","686e9cf6109241b28c44aba78667aaa9","bf5c401f943e4040893e01cade21b632","5bb2a8fc3af2421c9b1f92039ea61ba2","5050f71288f2478c8d5bdeff47fcdca5","551a53a84dd8459b8cdd3e3e09ae687c","5af01ff36b2d4edebc842f0aea68bbdd","243e232f096047ca84cafe43fdd02abf","aee15bf2685349e684a7c1e0847fe830","324b3953949e4416b3614572969d1124","20471d4effb54abfabce62c072f90ebe","34914f01bdad45aca133fda48cb7b77d","015ca0b73d984087b070630305d794af","df0642f7d06d4a1eaf10ddf27460f3f1","9398e6c79bd947ae8e7090363bd669a4","814c3545d598458182c9d638cd7d78c0","46cbf4315b0e4a00b41157bc4d03f0fd","00efd4b16cd24e7c92d9dd2daa03b10f","9e4171a5c49a494a9ae72aa72a67862c","7569df6a04a14194866afbb4e78f6468","39ff79fe9a214782894e983ce4d45acd","91b5f1a2be93471eb880235771722cd9","57c0c6ab34974fc89e2036bd1074a098","bb364cb3c7134948b24ba1b3a0c2186d","4108281cd53941288d99b60e994770f1","1e11face32ee40f890cb97751209d6ac","45a866686d954d33ba1b3034397661ea","f5014c5a31d342e7952bcebf024250a2","8288c75935d74e85bd62e1682f8f04fe","1981153d1d2943b1936946a362048f3f","d5aec20dc1034e57980a8e9121fdd8e0","9795e18cccce455485accc5309d8d9de","f0bea944fba44d508575eb73b8684596","f9b825f130304fccadb8d660daa0b3ad","4d18d9a7763749faacffaf660cb6c100","4ba50000d3b14f74a288aa8249723fd6","152308af6482481aa88914fb1f37eb13","2e3600f228c446349e2fe7126fb95255","57849a6643f64b58802bb57ab8e18412","426bad77baaa4db49164e3cf7b3505b9","2f6008ded0274199ba675ec6f0a7469d","7c290b50d93a4f0cb209a5e444934596","bc438acf4f2047e6b7e49829c8b078ee","9862cd6a17af4bfd80a02f9b68b80629","c949f9dd143c4ccda644e3a601a46302","b603160aed09476c937847a2c1fa6236","e3e9b4527bd84cbe9f83479d090416c0","3db20951b33c4ce3b14acd7fe1a438ce","96497d48b0bd436b964826dcbaa92a76","2fb73ee317e7482e94e7c696c6a0a2de","48cd0eb61edc469482a93d6a5cf8dabd","1ee577704a9048e39d73ad278b587215","5cd9bf0a588947a8895e187842dfca0c","cfcd77f72de043b3ad28dbaa216909d3","621a509dcda84c90b6f1e76b39b9b97a","e7ea517d5821476fa388b2d67c098675","14348abed9184177a87d6a8c0f558097","dbc3be056d4e4e7d87beb01252eecb37","0acc8312636b4361b915e9bd39a02057","f0f3a78cb34f41caba416d266c270952","f118a0a1a7b444508e6c02aa0acc83d7"]},"executionInfo":{"status":"ok","timestamp":1708322881797,"user_tz":-480,"elapsed":3008,"user":{"displayName":"szehanz","userId":"16137883221268059572"}},"outputId":"266d2243-1d5a-41a4-b0e2-25a573b7f602"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["Uploading the dataset shards: 0%| | 0/1 [00:00 "]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_142852-r9yt9wy0"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run vermilion-springroll-11 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/r9yt9wy0"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"4e64b2267c9c4213a5914056f6811d1b","version_major":2,"version_minor":0},"text/plain":["Map: 0%| | 0/334 [00:00\n"," \n"," \n"," [66/66 10:49, Epoch 6/6]\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.0054002.068154
201.4362001.290664
300.7781001.072349
400.6132000.931616
500.4969000.921341
600.4680000.888872

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:04]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.006 MB uploaded\\r'), FloatProgress(value=1.0, max=1.0)))"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▁▁▁▁
eval/runtime▁▅▇▇▇█▇
eval/samples_per_second█▄▂▂▂▁▂
eval/steps_per_second█▄▂▂▁▁▂
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▄▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.88887
eval/runtime5.8859
eval/samples_per_second7.306
eval/steps_per_second1.019
eval_loss0.88887
train/epoch6.0
train/global_step66
train/learning_rate2e-05
train/loss0.468
train/total_flos2090258212601856.0
train/train_loss1.06766
train/train_runtime658.5513
train/train_samples_per_second3.043
train/train_steps_per_second0.1

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run vermilion-springroll-11 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/r9yt9wy0
Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_142852-r9yt9wy0/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 14:40:09,515] Trial 0 finished with value: 0.8888720870018005 and parameters: {'learning_rate': 0.00022063199006940203, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 0 with value: 0.8888720870018005.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"3de6384b3841485d8068f7265fbda30e","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113223888807826, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_144009-p10q3kv9"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run lunar-ox-12 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/p10q3kv9"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"0a6c66cadea549f78fe5a183e841734e","version_major":2,"version_minor":0},"text/plain":["Map: 0%| | 0/334 [00:00\n"," \n"," \n"," [66/66 10:55, Epoch 6/6]\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8819001.890017
201.0036001.057150
300.6157000.908779
400.5004000.848109
500.4300000.856009
600.4028000.839791

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.023 MB uploaded\\r'), FloatProgress(value=0.2557933392427504, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁
eval/runtime▁▄▆█▆▆▅
eval/samples_per_second█▅▃▁▃▃▃
eval/steps_per_second█▅▃▁▄▃▄
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83979
eval/runtime5.9095
eval/samples_per_second7.276
eval/steps_per_second1.015
eval_loss0.83979
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.4028
train/total_flos2090258212601856.0
train/train_loss0.91779
train/train_runtime664.2846
train/train_samples_per_second3.017
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run lunar-ox-12 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/p10q3kv9
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_144009-p10q3kv9/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 14:51:33,062] Trial 1 finished with value: 0.8397907614707947 and parameters: {'learning_rate': 0.000388078354781562, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 5}. Best is trial 1 with value: 0.8397907614707947.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"9a11d1cc83a44c18a071c37ee29cb191","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113095899862755, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_145133-zcwhia3h"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run lunar-envelope-13 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/zcwhia3h"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [120/168 11:45 < 04:46, 0.17 it/s, Epoch 5/8]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8054001.864337
201.0785001.049373
300.6376000.906046
400.5493000.883616
500.4913000.882419
600.4567000.849847
700.4312000.859990
800.4107000.856455
900.3987000.832274
1000.3759000.863837
1100.3876000.843072
1200.3749000.846136

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17044687077892842, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▂▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▃▄▅▃▄▃▃▅█▅▅▄
eval/samples_per_second█▆▅▄▆▅▆▆▄▁▄▄▅
eval/steps_per_second█▆▅▄▅▅▆▆▄▁▄▄▄
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83227
eval/runtime5.8969
eval/samples_per_second7.292
eval/steps_per_second1.017
eval_loss0.83227
train/epoch5.71
train/global_step120
train/learning_rate0.00011
train/loss0.3749
train/total_flos1821144316674048.0
train/train_loss0.69982
train/train_runtime710.1571
train/train_samples_per_second3.763
train/train_steps_per_second0.237

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run lunar-envelope-13 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/zcwhia3h
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_145133-zcwhia3h/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:03:40,491] Trial 2 finished with value: 0.8322736024856567 and parameters: {'learning_rate': 0.00038013816677024434, 'num_train_epochs': 8, 'per_device_train_batch_size': 16, 'warmup_steps': 4}. Best is trial 2 with value: 0.8322736024856567.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"192adca007164c1188758c608dfd03f7","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.01111326985539765, max=1.0)…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_150340-0rux0mbt"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run dazzling-orchid-14 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/0rux0mbt"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [84/84 08:09, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.1171002.101251
201.5435001.223637
300.7658001.053006
400.6335000.956405
500.5513000.914150
600.4887000.883170
700.4532000.863325
800.4317000.861434

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.022 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.6507556781402156, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▂▁▁▁▁▁
eval/runtime▁▄▃▅█▆▅▇▅
eval/samples_per_second█▅▆▄▁▃▄▂▄
eval/steps_per_second█▅▆▅▁▃▅▂▄
eval_loss▁
train/epoch▁▁▂▂▃▃▄▄▅▅▆▆▇▇████
train/global_step▁▁▂▂▃▃▄▄▅▅▆▆▇▇█████
train/learning_rate█▇▆▅▄▃▂▁
train/loss█▄▂▂▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.86143
eval/runtime5.9003
eval/samples_per_second7.288
eval/steps_per_second1.017
eval_loss0.86143
train/epoch4.0
train/global_step84
train/learning_rate1e-05
train/loss0.4317
train/total_flos1274022822739968.0
train/train_loss0.96978
train/train_runtime494.7488
train/train_samples_per_second2.7
train/train_steps_per_second0.17

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run dazzling-orchid-14 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/0rux0mbt
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_150340-0rux0mbt/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:12:11,748] Trial 3 finished with value: 0.8614340424537659 and parameters: {'learning_rate': 0.00023956952379873406, 'num_train_epochs': 4, 'per_device_train_batch_size': 16, 'warmup_steps': 5}. Best is trial 2 with value: 0.8322736024856567.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"f7b8197a4e4c4242a7965774efc249e1","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113144411097488, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_151211-6nazgql4"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run prosperous-dragon-15 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/6nazgql4"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [66/66 10:56, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.6801001.782772
200.8941001.025657
300.5874000.897337
400.4763000.830805
500.4204000.864126
600.3949000.841267

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.1705031517334534, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁
eval/runtime▁▇▆▆█▇▅
eval/samples_per_second█▂▄▄▁▂▄
eval/steps_per_second█▁▃▃▁▁▃
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83081
eval/runtime5.9104
eval/samples_per_second7.275
eval/steps_per_second1.015
eval_loss0.83081
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.3949
train/total_flos2090258212601856.0
train/train_loss0.85955
train/train_runtime665.5241
train/train_samples_per_second3.011
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run prosperous-dragon-15 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/6nazgql4
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_151211-6nazgql4/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:23:34,430] Trial 4 finished with value: 0.8308054208755493 and parameters: {'learning_rate': 0.00041915607985727055, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d8dc9ac41d544e3fb4094e8c4b80cecc","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113028755709011, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_152334-gqh7mnqx"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run crimson-dragon-16 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/gqh7mnqx"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [44/44 07:13, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.6064001.642761
200.8001001.017194
300.5857000.903642
400.4734000.855594

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17057383277516674, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁
eval/runtime▁▅▅█▁
eval/samples_per_second█▄▅▁█
eval/steps_per_second█▅▅▁█
eval_loss▁
train/epoch▁▁▃▃▅▅▇▇██
train/global_step▁▁▃▃▅▅▇▇███
train/learning_rate█▆▃▁
train/loss█▂▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.85559
eval/runtime5.9033
eval/samples_per_second7.284
eval/steps_per_second1.016
eval_loss0.85559
train/epoch4.0
train/global_step44
train/learning_rate5e-05
train/loss0.4734
train/total_flos1395517145776128.0
train/train_loss1.0546
train/train_runtime442.5635
train/train_samples_per_second3.019
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run crimson-dragon-16 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/gqh7mnqx
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_152334-gqh7mnqx/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:31:13,768] Trial 5 finished with value: 0.855594277381897 and parameters: {'learning_rate': 0.0004882684074952214, 'num_train_epochs': 4, 'per_device_train_batch_size': 32, 'warmup_steps': 3}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"b17199cf7d1c41ee969d27814d9efe8b","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113176900026802, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_153113-emmid59b"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run beaming-paper-17 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/emmid59b"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [66/66 10:55, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.8145001.876597
201.0222001.064405
300.6246000.902199
400.5075000.854691
500.4362000.856049
600.4078000.836567

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17026547676022635, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▁▁▁▁▁
eval/runtime▁▁▄▃█▅▃
eval/samples_per_second██▅▅▁▄▆
eval/steps_per_second██▄▅▁▄▅
eval_loss▁
train/epoch▁▁▂▂▄▄▅▅▆▆▇▇██
train/global_step▁▁▂▂▃▃▅▅▆▆▇▇███
train/learning_rate█▇▅▄▂▁
train/loss█▃▂▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83657
eval/runtime5.9028
eval/samples_per_second7.285
eval/steps_per_second1.016
eval_loss0.83657
train/epoch6.0
train/global_step66
train/learning_rate4e-05
train/loss0.4078
train/total_flos2090258212601856.0
train/train_loss0.91464
train/train_runtime664.3977
train/train_samples_per_second3.016
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run beaming-paper-17 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/emmid59b
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_153113-emmid59b/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:42:35,259] Trial 6 finished with value: 0.8365665674209595 and parameters: {'learning_rate': 0.000371977101120841, 'num_train_epochs': 6, 'per_device_train_batch_size': 32, 'warmup_steps': 4}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"3718f807c93646f8ba07bbc2d3594547","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113231277947003, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_154235-x78afq0n"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run glittering-monkey-18 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/x78afq0n"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [44/44 07:14, Epoch 4/4]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
103.1951002.187073
201.5745001.620554
300.9015001.120180
400.6909001.036435

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17055463319920083, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▅▂▁▁
eval/runtime▁▆▄█▅
eval/samples_per_second█▃▅▁▄
eval/steps_per_second█▃▅▁▄
eval_loss▁
train/epoch▁▁▃▃▅▅▇▇██
train/global_step▁▁▃▃▅▅▇▇███
train/learning_rate█▆▃▁
train/loss█▃▂▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss1.03644
eval/runtime5.9099
eval/samples_per_second7.276
eval/steps_per_second1.015
eval_loss1.03644
train/epoch4.0
train/global_step44
train/learning_rate2e-05
train/loss0.6909
train/total_flos1395517145776128.0
train/train_loss1.50181
train/train_runtime442.6998
train/train_samples_per_second3.018
train/train_steps_per_second0.099

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run glittering-monkey-18 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/x78afq0n
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_154235-x78afq0n/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 15:50:14,768] Trial 7 finished with value: 1.0364350080490112 and parameters: {'learning_rate': 0.00021352963324526537, 'num_train_epochs': 4, 'per_device_train_batch_size': 32, 'warmup_steps': 5}. Best is trial 4 with value: 0.8308054208755493.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d31e8ddbc07347adb1c8a3798ee36db8","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113219644499218, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_155014-8e8ip35f"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run beaming-fuse-19 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/8e8ip35f"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [120/126 11:45 < 00:35, 0.17 it/s, Epoch 5/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7196001.833074
201.0752001.046847
300.6402000.922950
400.5504000.871228
500.4986000.876081
600.4539000.857398
700.4283000.860341
800.4096000.856458
900.3941000.830379
1000.3715000.847909
1100.3818000.842489
1200.3595000.851102

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:04]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.012 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.35110723430597374, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▃▂▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▂▂▃▄▆█▄█▇▅█▁
eval/samples_per_second█▇▇▆▅▃▁▅▁▂▄▁█
eval/steps_per_second▇▇▇▆▅▃▁▅▁▁▅▁█
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▄▄▄▄▅▅▅▅▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.83038
eval/runtime5.8734
eval/samples_per_second7.321
eval/steps_per_second1.022
eval_loss0.83038
train/epoch5.71
train/global_step120
train/learning_rate2e-05
train/loss0.3595
train/total_flos1821144316674048.0
train/train_loss0.69023
train/train_runtime710.9352
train/train_samples_per_second2.819
train/train_steps_per_second0.177

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run beaming-fuse-19 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/8e8ip35f
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_155014-8e8ip35f/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 16:02:22,467] Trial 8 finished with value: 0.8303791284561157 and parameters: {'learning_rate': 0.0003782307395143863, 'num_train_epochs': 6, 'per_device_train_batch_size': 16, 'warmup_steps': 3}. Best is trial 8 with value: 0.8303791284561157.\n"]},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"d8cdb5224da6481691ab8959404e3d24","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.011113188377607407, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_160222-bob3m06p"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run dazzling-ox-20 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/Education-Chatbot-Optimization"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/bob3m06p"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [140/168 13:43 < 02:47, 0.17 it/s, Epoch 6/8]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7859001.793209
201.0033001.026370
300.6111000.900247
400.5459000.883518
500.4864000.883806
600.4442000.881959
700.4328000.877395
800.4108000.867789
900.4000000.852276
1000.3731000.867085
1100.3862000.850435
1200.3738000.855305
1300.3920000.853369
1400.3604000.869668

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

\n"," \n"," \n"," [6/6 00:05]\n","
\n"," "],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.1705731731337404, max=1.0…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▁▁▁▁▁▁▁▁▁▁▁▁▁
eval/runtime▁▄▅▆▆▆▆▇▇█▆▆▆▇▆
eval/samples_per_second█▅▄▃▃▃▃▂▁▁▃▃▃▂▃
eval/steps_per_second█▅▄▂▃▄▃▂▂▁▂▃▄▂▃
eval_loss▁
train/epoch▁▁▂▂▂▂▃▃▃▃▄▄▄▄▅▅▅▅▆▆▆▆▇▇▇▇████
train/global_step▁▁▂▂▂▂▃▃▃▃▄▄▄▄▅▅▅▅▆▆▆▆▇▇▇▇█████
train/learning_rate█▇▇▆▆▅▅▄▄▃▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.85044
eval/runtime5.8867
eval/samples_per_second7.305
eval/steps_per_second1.019
eval_loss0.85044
train/epoch6.67
train/global_step140
train/learning_rate8e-05
train/loss0.3604
train/total_flos2129326975303680.0
train/train_loss0.64328
train/train_runtime828.8076
train/train_samples_per_second3.224
train/train_steps_per_second0.203

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run dazzling-ox-20 at: https://wandb.ai/szehanz/Education-Chatbot-Optimization/runs/bob3m06p
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_160222-bob3m06p/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"name":"stderr","output_type":"stream","text":["[I 2024-02-19 16:16:28,206] Trial 9 finished with value: 0.8504351377487183 and parameters: {'learning_rate': 0.000457264058410859, 'num_train_epochs': 8, 'per_device_train_batch_size': 16, 'warmup_steps': 5}. Best is trial 8 with value: 0.8303791284561157.\n"]}],"source":["def objective(trial):\n","\n"," # Define hyperparameters outside the wandb.init to use them later in the code\n"," learning_rate = trial.suggest_float('learning_rate', 2e-4, 5e-4, log=True)\n"," num_train_epochs = trial.suggest_categorical('num_train_epochs', [4, 6, 8])\n"," per_device_train_batch_size = trial.suggest_categorical('per_device_train_batch_size', [16, 32])\n"," warmup_steps = trial.suggest_int('warmup_steps', 3, 5)\n","\n"," wandb.init(\n"," project=\"Education-Chatbot-Optimization\",\n"," entity=\"szehanz\",\n"," group=\"optuna-optimization\",\n"," job_type=\"hyperparameter_search\",\n"," reinit=True,\n"," config={\n"," \"learning_rate\": learning_rate,\n"," \"num_train_epochs\": num_train_epochs,\n"," \"per_device_train_batch_size\": per_device_train_batch_size,\n"," \"warmup_steps\": warmup_steps\n"," }\n"," )\n","\n"," # Format the current date and time\n"," current_time = datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n"," output_dir = f\"train_out_dir_{current_time}\" # Append the current date and time to the directory name\n","\n"," # Create the output directory\n"," os.makedirs(output_dir, exist_ok=True) # Using exist_ok=True to avoid error if the directory already exists\n","\n","\n"," # Define TrainingArguments with the suggested hyperparameters\n"," training_args = TrainingArguments(\n"," output_dir=output_dir, # Directory for saving output models and checkpoints.\n"," save_strategy=\"steps\", # Save model checkpoints at regular step intervals.\n"," save_steps=10, # Save model checkpoints every 10 steps.\n"," learning_rate=learning_rate, # Initial learning rate for the optimizer.\n"," per_device_train_batch_size=per_device_train_batch_size, # Batch size per device during training.\n"," per_device_eval_batch_size=8, # Batch size per device during evaluation.\n"," num_train_epochs=num_train_epochs, # Total number of training epochs.\n"," warmup_steps=warmup_steps, # Number of warmup steps for the learning rate scheduler.\n"," evaluation_strategy='steps', # Perform evaluation at regular step intervals.\n"," eval_steps=10, # Perform evaluation every 10 steps.\n"," logging_steps=10,\n"," optim='paged_adamw_8bit', # Specifies the optimizer to use.\n"," lr_scheduler_type='linear', # Type of learning rate scheduler.\n"," gradient_accumulation_steps=1, # Number of steps to accumulate gradients before performing an update.\n"," load_best_model_at_end=True, # Load the best model based on evaluation metric at the end of training.\n"," report_to='wandb', # Disable automatic integrations with external reporting tools.\n"," )\n","\n","\n"," # Initialize the Trainer with early stopping callback inside the objective function\n"," trainer = SFTTrainer(\n"," model=model, # Ensure a function or a mechanism to initialize your model\n"," train_dataset=train_dataset,\n"," eval_dataset=val_dataset,\n"," peft_config=peft_config,\n"," dataset_text_field=\"Instruction\",\n"," tokenizer=tokenizer,\n"," args=training_args,\n"," max_seq_length=4096,\n"," callbacks=[EarlyStoppingCallback(early_stopping_patience=3)],\n"," )\n","\n"," # Train the model and evaluate within the objective function\n"," trainer.train()\n"," eval_result = trainer.evaluate()\n","\n"," # Log the primary metric to WandB\n"," wandb.log({\"eval_loss\": eval_result[\"eval_loss\"]})\n","\n"," # Finish the WandB run for this trial\n"," wandb.finish()\n","\n"," # Return the metric to be optimized\n"," return eval_result[\"eval_loss\"]\n","\n","\n","# Run the optimization\n","study = optuna.create_study(direction='minimize')\n","study.optimize(objective, n_trials=10)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"fmdlQTVSHT8e","outputId":"a2935a56-5cad-4dbc-c55c-53b3b5ad1368"},"outputs":[{"name":"stdout","output_type":"stream","text":["Best trial:\n"," Value: 0.8303791284561157\n"," Params: \n"," learning_rate: 0.0003782307395143863\n"," num_train_epochs: 6\n"," per_device_train_batch_size: 16\n"," warmup_steps: 3\n"]}],"source":["# Best trial results\n","print(\"Best trial:\")\n","print(f\" Value: {study.best_trial.value}\")\n","print(\" Params: \")\n","for key, value in study.best_trial.params.items():\n"," print(f\" {key}: {value}\")"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"mKlA_ahVHT8e","outputId":"6365a674-b011-48bb-94ea-7aa9d657d323","colab":{"referenced_widgets":[""]}},"outputs":[{"data":{"text/html":["Tracking run with wandb version 0.16.3"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Run data is saved locally in /home/iot/ITI110/poc-playground/Final project/wandb/run-20240219_161628-5gyifk7s"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Syncing run floating-fish-2 to Weights & Biases (docs)
"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View project at https://wandb.ai/szehanz/huggingface"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run at https://wandb.ai/szehanz/huggingface/runs/5gyifk7s"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","
\n"," \n"," \n"," [126/126 12:16, Epoch 6/6]\n","
\n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n"," \n","
StepTraining LossValidation Loss
102.7158001.857712
201.0773001.051454
300.6475000.913019
400.5472000.881412
500.4899000.886365
600.4574000.855178
700.4284000.860198
800.4072000.863780
900.3950000.834071
1000.3723000.848378
1100.3795000.848452
1200.3588000.857301

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"","version_major":2,"version_minor":0},"text/plain":["VBox(children=(Label(value='0.006 MB of 0.034 MB uploaded\\r'), FloatProgress(value=0.17130191715842674, max=1.…"]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["\n","

Run history:


eval/loss█▂▂▁▁▁▁▁▁▁▁▁
eval/runtime▁▄▄▆▆▆▅█▇▇▆█
eval/samples_per_second█▅▅▃▃▃▄▁▁▂▃▁
eval/steps_per_second█▅▅▃▃▃▄▂▂▂▃▁
train/epoch▁▁▂▂▂▂▃▃▃▃▄▄▅▅▅▅▆▆▆▆▇▇███
train/global_step▁▁▂▂▂▂▃▃▃▃▄▄▅▅▅▅▆▆▆▆▇▇███
train/learning_rate█▇▇▆▅▅▄▄▃▂▂▁
train/loss█▃▂▂▁▁▁▁▁▁▁▁
train/total_flos▁
train/train_loss▁
train/train_runtime▁
train/train_samples_per_second▁
train/train_steps_per_second▁

Run summary:


eval/loss0.8573
eval/runtime5.9051
eval/samples_per_second7.282
eval/steps_per_second1.016
train/epoch6.0
train/global_step126
train/learning_rate2e-05
train/loss0.3588
train/total_flos1913972332118016.0
train/train_loss0.67577
train/train_runtime746.485
train/train_samples_per_second2.685
train/train_steps_per_second0.169

"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":[" View run floating-fish-2 at: https://wandb.ai/szehanz/huggingface/runs/5gyifk7s
Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"],"text/plain":[""]},"metadata":{},"output_type":"display_data"},{"data":{"text/html":["Find logs at: ./wandb/run-20240219_161628-5gyifk7s/logs"],"text/plain":[""]},"metadata":{},"output_type":"display_data"}],"source":["# Use best hyperparameters from the study\n","best_trial = study.best_trial\n","\n","best_learning_rate = best_trial.params['learning_rate']\n","best_num_train_epochs = best_trial.params['num_train_epochs']\n","best_per_device_train_batch_size = best_trial.params['per_device_train_batch_size']\n","best_warmup_steps = best_trial.params['warmup_steps']\n","\n","\n","# Define TrainingArguments with the best hyperparameters for retraining\n","best_training_args = TrainingArguments(\n"," output_dir=\"best_train_out_dir\",\n"," save_strategy=\"steps\",\n"," save_steps=10,\n"," learning_rate=best_learning_rate,\n"," per_device_train_batch_size=best_per_device_train_batch_size,\n"," per_device_eval_batch_size=8,\n"," num_train_epochs=best_num_train_epochs,\n"," warmup_steps=best_warmup_steps,\n"," evaluation_strategy='steps',\n"," eval_steps=10,\n"," logging_steps=10,\n"," optim='paged_adamw_8bit',\n"," lr_scheduler_type='linear',\n"," gradient_accumulation_steps=1,\n"," load_best_model_at_end=True,\n"," report_to='wandb',\n",")\n","\n","# Reinitialize the Trainer with the best hyperparameters\n","best_trainer = SFTTrainer(\n"," model=model,\n"," train_dataset=train_dataset,\n"," eval_dataset=val_dataset,\n"," peft_config=peft_config,\n"," dataset_text_field=\"Instruction\",\n"," tokenizer=tokenizer,\n"," args=best_training_args,\n"," max_seq_length=4096,\n",")\n","\n","# Retrain the model with the best hyperparameters\n","best_trainer.train()\n","\n","\n","# Save trained model\n","best_trainer.model.save_pretrained(new_model)\n","\n","# Finish the WandB run for this trial\n","wandb.finish()"]},{"cell_type":"markdown","metadata":{"id":"_g0fB7P9s0ol"},"source":["Merging the base model with the trained adapter."]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"referenced_widgets":["aafec7a64d034e05b1aaf17bb153136b","1191c9b140394f1aa3952c1cecda8fed","68107c402ec343ffa40e22171e9fe3e9"]},"id":"QQn30cRtAZ-P","outputId":"6508be7b-0a96-494e-bd33-d35c5c331f52"},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"68107c402ec343ffa40e22171e9fe3e9","version_major":2,"version_minor":0},"text/plain":["Loading checkpoint shards: 0%| | 0/2 [00:00
Copy a token from your Hugging Face\ntokens page and paste it below.
Immediately click login after copying\nyour token or it might be stored in plain text in this notebook file.
"}},"14d0ff6e07a148c0b65224af86af6b95":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"187745df1f154e7da09a0c646a22cebb":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_638517f3c94445c7b610b84744859ef8","placeholder":"​","style":"IPY_MODEL_5206f9fc25234586b735fc9c65e6b19a","value":"generation_config.json: 100%"}},"19b8b7692cfd46eca5abb50ac2854262":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"19d3adedf1a245f69a2498b1838f6415":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_31ffba7793b944cd80bec0fb6ac2fbbe","IPY_MODEL_1ae848be349a41239a7ba6efadc82edf","IPY_MODEL_9a33d92ad0604eff9ac56944586facbb"],"layout":"IPY_MODEL_293d425206684064b0572e990a815cee"}},"1ae848be349a41239a7ba6efadc82edf":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_8ae8a9d03adc476a8dd64f09e194907d","max":2,"min":0,"orientation":"horizontal","style":"IPY_MODEL_aeb94dbf3f8a4b43bb07c833af3b98c9","value":2}},"1c6e5dd4ca3446ec8cf8a03b85ad57dc":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_0786fc506a7640a287016991f57d3768","IPY_MODEL_3af05a188f40469dbd52fda55c7a9e22","IPY_MODEL_0af30e36f2524c1084c1237948f3b18e"],"layout":"IPY_MODEL_dfe1c2648a564676bfe1e09bda61d439"}},"1ea12323b1244c768a17a6ed5420f854":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"1f9e8ddb08814b2db7761478eb9069f5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_7982d1d315964d138f31643db445b48a","max":614,"min":0,"orientation":"horizontal","style":"IPY_MODEL_67346bc716384552a69d1afdc844db00","value":614}},"20cc750b1bde4efdb3c3a5b5321222b4":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_a199f20edde447829b17804b17af2fe5","max":188,"min":0,"orientation":"horizontal","style":"IPY_MODEL_447f608108364159ab7ed546ad02559a","value":188}},"21556be54ed34b15b909bf8e7b8fd93a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b8113970ea7245e9890221d4e4cf5e8e","placeholder":"​","style":"IPY_MODEL_d3224d16458249a3bfd29253c2d6a86f","value":"\nPro Tip: If you don't already have one, you can create a dedicated\n'notebooks' token with 'write' access, that you can then easily reuse for all\nnotebooks. "}},"2254fe28315f43a99c3579195c0c0008":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"22e8bae9ff6744bca3e90d46d220106f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"23176fa0130348c5bc1b0deb591709f3":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"236b3417f0e94039a1a449e7ab96738b":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"25543b118c8e423aa7fd8c898d5e1256":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"2924e96aa10346efb39684e5369e2170":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_bce8d1501218410ba8b042aeb3f0fc26","placeholder":"​","style":"IPY_MODEL_405603de026d484ab283f053f4b17c6d","value":"Your token has been saved to /root/.cache/huggingface/token"}},"293d425206684064b0572e990a815cee":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"29f303aa6ac8464aa91124c3fe659379":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2cb9cfbde1e0483c97a2c531e0034adf":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"2dce1978d19e4de3a6a1b1cef6ed518f":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"3076e4abb7fe427fa4fccb43e9f3371e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"30d47a9da70a4cf5a7e597f13d28d526":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"31ffba7793b944cd80bec0fb6ac2fbbe":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_7ee17b5879ec4909b7b616b79d5eab36","placeholder":"​","style":"IPY_MODEL_d8b847adb46d4049a4b43bba2a0e06da","value":"Downloading shards: 100%"}},"32a57f9df45b41da91e1d638f531505d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"3af05a188f40469dbd52fda55c7a9e22":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_82829039182948efb800703e3e21d96d","max":26788,"min":0,"orientation":"horizontal","style":"IPY_MODEL_32a57f9df45b41da91e1d638f531505d","value":26788}},"3e4a06b9b13444e3b82e0c3c26e17b8f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"CheckboxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"CheckboxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"CheckboxView","description":"Add token as git credential?","description_tooltip":null,"disabled":false,"indent":true,"layout":"IPY_MODEL_f91ebc43c1344e8688e2eeb2771c7b65","style":"IPY_MODEL_ea00aa1eb73949fc94083f1d31372915","value":true}},"405603de026d484ab283f053f4b17c6d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"42262413163d49fd8902c275c13da1e6":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_cf370d3015b04b6caa9cbfb8f21dd5e7","placeholder":"​","style":"IPY_MODEL_1ea12323b1244c768a17a6ed5420f854","value":" 9.98G/9.98G [01:29<00:00, 126MB/s]"}},"441c37014e6b4c8995e7a5305f96fa38":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"447f608108364159ab7ed546ad02559a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"46e41557d0bc4cb4a03c6894fcc57d0e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"4cee3e23bcbd4643ba542d94b82dbfd2":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"5206f9fc25234586b735fc9c65e6b19a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"52f2b6adb9904cbeb014387429c0d1ad":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_b61ac64b2a0c4beb8ea209782b9dd09b","IPY_MODEL_9e9c2d712ae84d3aa87fbe433ae0701b","IPY_MODEL_fe4d028a78e64a4d8b47f8bc8b8ec15d"],"layout":"IPY_MODEL_88a7aeaf3b2547ea91cf20ed2f1a3dc0"}},"537d173a4313491f8f89f799f1dc7b2a":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"541ef20ab6f34337a2d6d20098f6fef5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b0f0ac261e364edd99d7b75e747e2c47","placeholder":"​","style":"IPY_MODEL_ff9f726db3434e3184e723d5da884d0a","value":"Token is valid (permission: write)."}},"55e506bd03d24dd7b24129dc64da9c01":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"5e5b95c9801443cdbce9c8e629c33589":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"PasswordModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"PasswordModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"PasswordView","continuous_update":true,"description":"Token:","description_tooltip":null,"disabled":false,"layout":"IPY_MODEL_29f303aa6ac8464aa91124c3fe659379","placeholder":"​","style":"IPY_MODEL_3076e4abb7fe427fa4fccb43e9f3371e","value":""}},"62f1cf19fe204aa4a424248e807ce061":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"638517f3c94445c7b610b84744859ef8":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"6430379d01874ec3a7cf9fea59c42914":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ButtonModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ButtonModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ButtonView","button_style":"","description":"Login","disabled":false,"icon":"","layout":"IPY_MODEL_2dce1978d19e4de3a6a1b1cef6ed518f","style":"IPY_MODEL_e81c501824f94e7d839684fafbc65b31","tooltip":""}},"67346bc716384552a69d1afdc844db00":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"6edf40f558f54d8b82d949f83557d609":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_08400a144d3c497a94ae4d84e72a1067","placeholder":"​","style":"IPY_MODEL_976a3440d2c3423c8be835b0d6f56492","value":"Connecting..."}},"77c0d63f6cd6494daf7e94cfe0397e45":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_4cee3e23bcbd4643ba542d94b82dbfd2","placeholder":"​","style":"IPY_MODEL_25543b118c8e423aa7fd8c898d5e1256","value":" 188/188 [00:00<00:00, 12.3kB/s]"}},"77daad9def9f4254933afdeb7202e5d2":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"7982d1d315964d138f31643db445b48a":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"7c701d53772344f6a057cd43742d47f5":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b9e4e59acd4d468393f42857d5635e25","placeholder":"​","style":"IPY_MODEL_30d47a9da70a4cf5a7e597f13d28d526","value":"Loading checkpoint shards: 100%"}},"7ee17b5879ec4909b7b616b79d5eab36":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"8080eb1edf8f44f69b5f218046715f36":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_e2df35aaf7e0420b9524d533cb122521","max":2,"min":0,"orientation":"horizontal","style":"IPY_MODEL_d90c3a424bac4bb09e99339c28df2eda","value":2}},"82829039182948efb800703e3e21d96d":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"83127281047a429a846b9bb3b3209760":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"86014f569ee34895badef4acfe958051":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"86616a0d2f9d4da1a75e00d092ca9d4d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_a7a17a5883784ed9961437d0102fd9a4","IPY_MODEL_9934540648a444a09f0783843ab2d901","IPY_MODEL_42262413163d49fd8902c275c13da1e6"],"layout":"IPY_MODEL_83127281047a429a846b9bb3b3209760"}},"88a7aeaf3b2547ea91cf20ed2f1a3dc0":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"8ae8a9d03adc476a8dd64f09e194907d":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"963973c1e970410cb2ba97e9fb4ef511":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"976a3440d2c3423c8be835b0d6f56492":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"9934540648a444a09f0783843ab2d901":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_f702ae2d46834e9b934a40abca49a990","max":9976576152,"min":0,"orientation":"horizontal","style":"IPY_MODEL_14d0ff6e07a148c0b65224af86af6b95","value":9976576152}},"9a33d92ad0604eff9ac56944586facbb":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_537d173a4313491f8f89f799f1dc7b2a","placeholder":"​","style":"IPY_MODEL_55e506bd03d24dd7b24129dc64da9c01","value":" 2/2 [01:56<00:00, 52.47s/it]"}},"9c7550ec29c34ef2b674a3667ebd740e":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_236b3417f0e94039a1a449e7ab96738b","placeholder":"​","style":"IPY_MODEL_77daad9def9f4254933afdeb7202e5d2","value":"config.json: 100%"}},"9e4d26b87e1243f3ae744ddbbe2c5b51":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"9e9c2d712ae84d3aa87fbe433ae0701b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"FloatProgressModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"FloatProgressModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"ProgressView","bar_style":"success","description":"","description_tooltip":null,"layout":"IPY_MODEL_86014f569ee34895badef4acfe958051","max":3500296424,"min":0,"orientation":"horizontal","style":"IPY_MODEL_22e8bae9ff6744bca3e90d46d220106f","value":3500296424}},"a066a28914bc46cc85ebcb7d90d27fbe":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a199f20edde447829b17804b17af2fe5":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a39b2a49ff814181aa0df84c57e9245b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"a7a17a5883784ed9961437d0102fd9a4":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_ae173a50f4374c7fa55d8e360d30d4d1","placeholder":"​","style":"IPY_MODEL_b93a21bcaf6344c49a0ab8e2a887922c","value":"model-00001-of-00002.safetensors: 100%"}},"a7ef8ff133144d4b9817800e5b4739a4":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"a9628200c0a34cccacdb0ea66e200b87":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_0cda6495ef4b4c9ab3eeff2392d89163","placeholder":"​","style":"IPY_MODEL_9e4d26b87e1243f3ae744ddbbe2c5b51","value":" 2/2 [00:58<00:00, 26.87s/it]"}},"a96de5c1f80f43d8b0b689d5db0fa248":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"ae173a50f4374c7fa55d8e360d30d4d1":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"aeb94dbf3f8a4b43bb07c833af3b98c9":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"b0f0ac261e364edd99d7b75e747e2c47":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b4ed4d5e266a4cd5b9011436b4e92951":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_7c701d53772344f6a057cd43742d47f5","IPY_MODEL_8080eb1edf8f44f69b5f218046715f36","IPY_MODEL_a9628200c0a34cccacdb0ea66e200b87"],"layout":"IPY_MODEL_2254fe28315f43a99c3579195c0c0008"}},"b577aeffe358446ea41e4b3c58b80641":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_441c37014e6b4c8995e7a5305f96fa38","placeholder":"​","style":"IPY_MODEL_a39b2a49ff814181aa0df84c57e9245b","value":" 614/614 [00:00<00:00, 15.7kB/s]"}},"b61ac64b2a0c4beb8ea209782b9dd09b":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_a066a28914bc46cc85ebcb7d90d27fbe","placeholder":"​","style":"IPY_MODEL_a96de5c1f80f43d8b0b689d5db0fa248","value":"model-00002-of-00002.safetensors: 100%"}},"b634ae14caf84bc99dcafd3a70b7a383":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b64f26ac024c46eabfc4728586369130":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_a7ef8ff133144d4b9817800e5b4739a4","placeholder":"​","style":"IPY_MODEL_ba4cf32b2f71428282721e7818b34a5a","value":"Login successful"}},"b8113970ea7245e9890221d4e4cf5e8e":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"b93a21bcaf6344c49a0ab8e2a887922c":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"b9e4e59acd4d468393f42857d5635e25":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"ba4cf32b2f71428282721e7818b34a5a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"bce8d1501218410ba8b042aeb3f0fc26":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"c3d08bc595a74c3180a7a83afc569584":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"VBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"VBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"VBoxView","box_style":"","children":["IPY_MODEL_541ef20ab6f34337a2d6d20098f6fef5","IPY_MODEL_fe46e1cf697f4b1fab764104be32da95","IPY_MODEL_2924e96aa10346efb39684e5369e2170","IPY_MODEL_b64f26ac024c46eabfc4728586369130"],"layout":"IPY_MODEL_cad60b6f14f249c187d573dd3a4428e0"}},"ca8d07b9b454471e883d43e03b475a9f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"cad60b6f14f249c187d573dd3a4428e0":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":"center","align_self":null,"border":null,"bottom":null,"display":"flex","flex":null,"flex_flow":"column","grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":"50%"}},"cf370d3015b04b6caa9cbfb8f21dd5e7":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"d3224d16458249a3bfd29253c2d6a86f":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d8b847adb46d4049a4b43bba2a0e06da":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"d90c3a424bac4bb09e99339c28df2eda":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ProgressStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ProgressStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","bar_color":null,"description_width":""}},"de756c426cf0492bb122a45b94d4bbe7":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"df0d8231953b44f1bf8e6d2cd39b2f66":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HBoxModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HBoxModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HBoxView","box_style":"","children":["IPY_MODEL_9c7550ec29c34ef2b674a3667ebd740e","IPY_MODEL_1f9e8ddb08814b2db7761478eb9069f5","IPY_MODEL_b577aeffe358446ea41e4b3c58b80641"],"layout":"IPY_MODEL_19b8b7692cfd46eca5abb50ac2854262"}},"dfe1c2648a564676bfe1e09bda61d439":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"e2df35aaf7e0420b9524d533cb122521":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"e81c501824f94e7d839684fafbc65b31":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"ButtonStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"ButtonStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","button_color":null,"font_weight":""}},"ea00aa1eb73949fc94083f1d31372915":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}},"f702ae2d46834e9b934a40abca49a990":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"f81ada25e7ff4f5da6b3f6c6e73590e4":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"f91ebc43c1344e8688e2eeb2771c7b65":{"model_module":"@jupyter-widgets/base","model_module_version":"1.2.0","model_name":"LayoutModel","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"fe46e1cf697f4b1fab764104be32da95":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"LabelModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"LabelModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"LabelView","description":"","description_tooltip":null,"layout":"IPY_MODEL_f81ada25e7ff4f5da6b3f6c6e73590e4","placeholder":"​","style":"IPY_MODEL_62f1cf19fe204aa4a424248e807ce061","value":"Your token has been saved in your configured git credential helpers (store)."}},"fe4d028a78e64a4d8b47f8bc8b8ec15d":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"HTMLModel","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"HTMLModel","_view_count":null,"_view_module":"@jupyter-widgets/controls","_view_module_version":"1.5.0","_view_name":"HTMLView","description":"","description_tooltip":null,"layout":"IPY_MODEL_b634ae14caf84bc99dcafd3a70b7a383","placeholder":"​","style":"IPY_MODEL_46e41557d0bc4cb4a03c6894fcc57d0e","value":" 3.50G/3.50G [00:26<00:00, 176MB/s]"}},"ff9f726db3434e3184e723d5da884d0a":{"model_module":"@jupyter-widgets/controls","model_module_version":"1.5.0","model_name":"DescriptionStyleModel","state":{"_model_module":"@jupyter-widgets/controls","_model_module_version":"1.5.0","_model_name":"DescriptionStyleModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"StyleView","description_width":""}}}}},"nbformat":4,"nbformat_minor":0} \ No newline at end of file