Spaces:
Sleeping
Sleeping
Converting model to DistilGPT2
Browse files
app.py
CHANGED
@@ -1,28 +1,32 @@
|
|
1 |
import gradio as gr
|
2 |
-
from transformers import
|
|
|
3 |
from langchain.memory import ConversationBufferMemory
|
4 |
-
from langchain.prompts import PromptTemplate
|
5 |
|
6 |
-
# Load the tokenizer and model for
|
7 |
-
tokenizer =
|
8 |
-
model =
|
|
|
|
|
|
|
|
|
9 |
|
10 |
# Set up conversational memory using LangChain's ConversationBufferMemory
|
11 |
memory = ConversationBufferMemory()
|
12 |
|
13 |
# Define the chatbot function with memory
|
14 |
-
def
|
15 |
# Retrieve conversation history and append the current user input
|
16 |
conversation_history = memory.load_memory_variables({})['history']
|
17 |
|
18 |
# Combine the history with the current user input
|
19 |
full_input = f"{conversation_history}\nUser: {input_text}\nAssistant:"
|
20 |
|
21 |
-
# Tokenize the input
|
22 |
-
input_ids = tokenizer.encode(full_input, return_tensors="pt")
|
23 |
|
24 |
-
# Generate the response
|
25 |
-
outputs = model.generate(input_ids, max_length=200, num_return_sequences=1)
|
26 |
|
27 |
# Decode the model output
|
28 |
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
@@ -34,13 +38,14 @@ def chat_with_flan(input_text):
|
|
34 |
|
35 |
# Set up the Gradio interface
|
36 |
interface = gr.Interface(
|
37 |
-
fn=
|
38 |
-
inputs=gr.Textbox(label="Chat with
|
39 |
-
outputs=gr.Textbox(label="
|
40 |
-
title="
|
41 |
-
description="This is a simple chatbot powered by the
|
42 |
)
|
43 |
|
44 |
# Launch the Gradio app
|
45 |
interface.launch()
|
46 |
|
|
|
|
1 |
import gradio as gr
|
2 |
+
from transformers import GPT2Tokenizer, GPT2LMHeadModel
|
3 |
+
import torch
|
4 |
from langchain.memory import ConversationBufferMemory
|
|
|
5 |
|
6 |
+
# Load the tokenizer and model for DistilGPT-2
|
7 |
+
tokenizer = GPT2Tokenizer.from_pretrained("distilgpt2")
|
8 |
+
model = GPT2LMHeadModel.from_pretrained("distilgpt2")
|
9 |
+
|
10 |
+
# Move model to device (GPU if available)
|
11 |
+
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
|
12 |
+
model.to(device)
|
13 |
|
14 |
# Set up conversational memory using LangChain's ConversationBufferMemory
|
15 |
memory = ConversationBufferMemory()
|
16 |
|
17 |
# Define the chatbot function with memory
|
18 |
+
def chat_with_distilgpt2(input_text):
|
19 |
# Retrieve conversation history and append the current user input
|
20 |
conversation_history = memory.load_memory_variables({})['history']
|
21 |
|
22 |
# Combine the history with the current user input
|
23 |
full_input = f"{conversation_history}\nUser: {input_text}\nAssistant:"
|
24 |
|
25 |
+
# Tokenize the input and convert to tensor
|
26 |
+
input_ids = tokenizer.encode(full_input, return_tensors="pt").to(device)
|
27 |
|
28 |
+
# Generate the response using the model
|
29 |
+
outputs = model.generate(input_ids, max_length=200, num_return_sequences=1, pad_token_id=tokenizer.eos_token_id)
|
30 |
|
31 |
# Decode the model output
|
32 |
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
|
|
38 |
|
39 |
# Set up the Gradio interface
|
40 |
interface = gr.Interface(
|
41 |
+
fn=chat_with_distilgpt2,
|
42 |
+
inputs=gr.Textbox(label="Chat with DistilGPT-2"),
|
43 |
+
outputs=gr.Textbox(label="DistilGPT-2's Response"),
|
44 |
+
title="DistilGPT-2 Chatbot with Memory",
|
45 |
+
description="This is a simple chatbot powered by the DistilGPT-2 model with conversational memory, using LangChain.",
|
46 |
)
|
47 |
|
48 |
# Launch the Gradio app
|
49 |
interface.launch()
|
50 |
|
51 |
+
|