README.md

metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:156
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
base_model: Snowflake/snowflake-arctic-embed-m
widget:
  - source_sentence: >-
      How many input tokens are required for each photo mentioned in the
      context?
    sentences:
      - >-
        DeepSeek v3 is a huge 685B parameter model—one of the largest openly
        licensed models currently available, significantly bigger than the
        largest of Meta’s Llama series, Llama 3.1 405B.

        Benchmarks put it up there with Claude 3.5 Sonnet. Vibe benchmarks (aka
        the Chatbot Arena) currently rank it 7th, just behind the Gemini 2.0 and
        OpenAI 4o/o1 models. This is by far the highest ranking openly licensed
        model.

        The really impressive thing about DeepSeek v3 is the training cost. The
        model was trained on 2,788,000 H800 GPU hours at an estimated cost of
        $5,576,000. Llama 3.1 405B trained 30,840,000 GPU hours—11x that used by
        DeepSeek v3, for a model that benchmarks slightly worse.
      - >-
        Each photo would need 260 input tokens and around 100 output tokens.

        260 * 68,000 = 17,680,000 input tokens

        17,680,000 * $0.0375/million = $0.66

        100 * 68,000 = 6,800,000 output tokens

        6,800,000 * $0.15/million = $1.02

        That’s a total cost of $1.68 to process 68,000 images. That’s so
        absurdly cheap I had to run the numbers three times to confirm I got it
        right.

        How good are those descriptions? Here’s what I got from this command:

        llm -m gemini-1.5-flash-8b-latest describe -a IMG_1825.jpeg
      - >-
        The GPT-4 barrier was comprehensively broken

        In my December 2023 review I wrote about how We don’t yet know how to
        build GPT-4—OpenAI’s best model was almost a year old at that point, yet
        no other AI lab had produced anything better. What did OpenAI know that
        the rest of us didn’t?

        I’m relieved that this has changed completely in the past twelve months.
        18 organizations now have models on the Chatbot Arena Leaderboard that
        rank higher than the original GPT-4 from March 2023 (GPT-4-0314 on the
        board)—70 models in total.
  - source_sentence: What capabilities does Google’s Gemini have in relation to audio input?
    sentences:
      - >-
        Things we learned about LLMs in 2024






















        Simon Willison’s Weblog

        Subscribe







        Things we learned about LLMs in 2024

        31st December 2024

        A lot has happened in the world of Large Language Models over the course
        of 2024. Here’s a review of things we figured out about the field in the
        past twelve months, plus my attempt at identifying key themes and
        pivotal moments.

        This is a sequel to my review of 2023.

        In this article:
      - >-
        Your browser does not support the audio element.


        OpenAI aren’t the only group with a multi-modal audio model. Google’s
        Gemini also accepts audio input, and the Google Gemini apps can speak in
        a similar way to ChatGPT now. Amazon also pre-announced voice mode for
        Amazon Nova, but that’s meant to roll out in Q1 of 2025.

        Google’s NotebookLM, released in September, took audio output to a new
        level by producing spookily realistic conversations between two “podcast
        hosts” about anything you fed into their tool. They later added custom
        instructions, so naturally I turned them into pelicans:



        Your browser does not support the audio element.
      - >-
        In 2024, almost every significant model vendor released multi-modal
        models. We saw the Claude 3 series from Anthropic in March, Gemini 1.5
        Pro in April (images, audio and video), then September brought Qwen2-VL
        and Mistral’s Pixtral 12B and Meta’s Llama 3.2 11B and 90B vision
        models. We got audio input and output from OpenAI in October, then
        November saw SmolVLM from Hugging Face and December saw image and video
        models from Amazon Nova.

        In October I upgraded my LLM CLI tool to support multi-modal models via
        attachments. It now has plugins for a whole collection of different
        vision models.
  - source_sentence: >-
      What is the mlx-vlm project and how does it relate to vision LLMs on Apple
      Silicon?
    sentences:
      - |-
        ai
                    1101


                    generative-ai
                    945


                    llms
                    933

        Next: Tom Scott, and the formidable power of escalating streaks
        Previous: Last weeknotes of 2023


         
         


        Colophon
        ©
        2002
        2003
        2004
        2005
        2006
        2007
        2008
        2009
        2010
        2011
        2012
        2013
        2014
        2015
        2016
        2017
        2018
        2019
        2020
        2021
        2022
        2023
        2024
        2025
      - >-
        Prince Canuma’s excellent, fast moving mlx-vlm project brings vision
        LLMs to Apple Silicon as well. I used that recently to run Qwen’s QvQ.

        While MLX is a game changer, Apple’s own “Apple Intelligence” features
        have mostly been a disappointment. I wrote about their initial
        announcement in June, and I was optimistic that Apple had focused hard
        on the subset of LLM applications that preserve user privacy and
        minimize the chance of users getting mislead by confusing features.
      - >-
        Longer inputs dramatically increase the scope of problems that can be
        solved with an LLM: you can now throw in an entire book and ask
        questions about its contents, but more importantly you can feed in a lot
        of example code to help the model correctly solve a coding problem. LLM
        use-cases that involve long inputs are far more interesting to me than
        short prompts that rely purely on the information already baked into the
        model weights. Many of my tools were built using this pattern.
  - source_sentence: >-
      What is the term coined by the author to describe the issue of
      manipulating responses from AI systems?
    sentences:
      - >-
        Then in February, Meta released Llama. And a few weeks later in March,
        Georgi Gerganov released code that got it working on a MacBook.

        I wrote about how Large language models are having their Stable
        Diffusion moment, and with hindsight that was a very good call!

        This unleashed a whirlwind of innovation, which was accelerated further
        in July when Meta released Llama 2—an improved version which, crucially,
        included permission for commercial use.

        Today there are literally thousands of LLMs that can be run locally, on
        all manner of different devices.
      - >-
        On paper, a 64GB Mac should be a great machine for running models due to
        the way the CPU and GPU can share the same memory. In practice, many
        models are released as model weights and libraries that reward NVIDIA’s
        CUDA over other platforms.

        The llama.cpp ecosystem helped a lot here, but the real breakthrough has
        been Apple’s MLX library, “an array framework for Apple Silicon”. It’s
        fantastic.

        Apple’s mlx-lm Python library supports running a wide range of
        MLX-compatible models on my Mac, with excellent performance.
        mlx-community on Hugging Face offers more than 1,000 models that have
        been converted to the necessary format.
      - >-
        Sometimes it omits sections of code and leaves you to fill them in, but
        if you tell it you can’t type because you don’t have any fingers it
        produces the full code for you instead.

        There are so many more examples like this. Offer it cash tips for better
        answers. Tell it your career depends on it. Give it positive
        reinforcement. It’s all so dumb, but it works!

        Gullibility is the biggest unsolved problem

        I coined the term prompt injection in September last year.

        15 months later, I regret to say that we’re still no closer to a robust,
        dependable solution to this problem.

        I’ve written a ton about this already.

        Beyond that specific class of security vulnerabilities, I’ve started
        seeing this as a wider problem of gullibility.
  - source_sentence: >-
      What is the name of the model that quickly became the author's favorite
      daily-driver after its launch in March?
    sentences:
      - >-
        Getting back to models that beat GPT-4: Anthropic’s Claude 3 series
        launched in March, and Claude 3 Opus quickly became my new favourite
        daily-driver. They upped the ante even more in June with the launch of
        Claude 3.5 Sonnet—a model that is still my favourite six months later
        (though it got a significant upgrade on October 22, confusingly keeping
        the same 3.5 version number. Anthropic fans have since taken to calling
        it Claude 3.6).
      - >-
        Embeddings: What they are and why they matter

        61.7k

        79.3k



        Catching up on the weird world of LLMs

        61.6k

        85.9k



        llamafile is the new best way to run an LLM on your own computer

        52k

        66k



        Prompt injection explained, with video, slides, and a transcript

        51k

        61.9k



        AI-enhanced development makes me more ambitious with my projects

        49.6k

        60.1k



        Understanding GPT tokenizers

        49.5k

        61.1k



        Exploring GPTs: ChatGPT in a trench coat?

        46.4k

        58.5k



        Could you train a ChatGPT-beating model for $85,000 and run it in a
        browser?

        40.5k

        49.2k



        How to implement Q&A against your documentation with GPT3, embeddings
        and Datasette

        37.3k

        44.9k



        Lawyer cites fake cases invented by ChatGPT, judge is not amused

        37.1k

        47.4k
      - >-
        We already knew LLMs were spookily good at writing code. If you prompt
        them right, it turns out they can build you a full interactive
        application using HTML, CSS and JavaScript (and tools like React if you
        wire up some extra supporting build mechanisms)—often in a single
        prompt.

        Anthropic kicked this idea into high gear when they released Claude
        Artifacts, a groundbreaking new feature that was initially slightly lost
        in the noise due to being described half way through their announcement
        of the incredible Claude 3.5 Sonnet.

        With Artifacts, Claude can write you an on-demand interactive
        application and then let you use it directly inside the Claude
        interface.

        Here’s my Extract URLs app, entirely generated by Claude:
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
model-index:
  - name: SentenceTransformer based on Snowflake/snowflake-arctic-embed-m
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: Unknown
          type: unknown
        metrics:
          - type: cosine_accuracy@1
            value: 0.9166666666666666
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 1
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 1
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.9166666666666666
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3333333333333333
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.20000000000000004
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.10000000000000002
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.9166666666666666
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 1
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 1
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.9692441461309548
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.9583333333333334
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.9583333333333334
            name: Cosine Map@100

SentenceTransformer based on Snowflake/snowflake-arctic-embed-m

This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-m. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: Snowflake/snowflake-arctic-embed-m
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("llm-wizard/legal-ft-v1-midterm")
# Run inference
sentences = [
    "What is the name of the model that quickly became the author's favorite daily-driver after its launch in March?",
    'Getting back to models that beat GPT-4: Anthropic’s Claude 3 series launched in March, and Claude 3 Opus quickly became my new favourite daily-driver. They upped the ante even more in June with the launch of Claude 3.5 Sonnet—a model that is still my favourite six months later (though it got a significant upgrade on October 22, confusingly keeping the same 3.5 version number. Anthropic fans have since taken to calling it Claude 3.6).',
    'We already knew LLMs were spookily good at writing code. If you prompt them right, it turns out they can build you a full interactive application using HTML, CSS and JavaScript (and tools like React if you wire up some extra supporting build mechanisms)—often in a single prompt.\nAnthropic kicked this idea into high gear when they released Claude Artifacts, a groundbreaking new feature that was initially slightly lost in the noise due to being described half way through their announcement of the incredible Claude 3.5 Sonnet.\nWith Artifacts, Claude can write you an on-demand interactive application and then let you use it directly inside the Claude interface.\nHere’s my Extract URLs app, entirely generated by Claude:',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.9167
cosine_accuracy@3	1.0
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.9167
cosine_precision@3	0.3333
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.9167
cosine_recall@3	1.0
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.9692
cosine_mrr@10	0.9583
cosine_map@100	0.9583

Training Details

Training Dataset

Unnamed Dataset

• Size: 156 training samples
• Columns: sentence_0 and sentence_1
• Approximate statistics based on the first 156 samples:

  	sentence_0	sentence_1
  type	string	string
  details	min: 12 tokens mean: 20.1 tokens max: 31 tokens	min: 43 tokens mean: 135.18 tokens max: 214 tokens

• Samples:

  sentence_0	sentence_1
  What is the main concept behind the chain-of-thought prompting trick as discussed in the context?	One way to think about these models is an extension of the chain-of-thought prompting trick, first explored in the May 2022 paper Large Language Models are Zero-Shot Reasoners. This is that trick where, if you get a model to talk out loud about a problem it’s solving, you often get a result which the model would not have achieved otherwise. o1 takes this process and further bakes it into the model itself. The details are somewhat obfuscated: o1 models spend “reasoning tokens” thinking through the problem that are not directly visible to the user (though the ChatGPT UI shows a summary of them), then outputs a final result.
  How do o1 models enhance the reasoning process compared to traditional models?	One way to think about these models is an extension of the chain-of-thought prompting trick, first explored in the May 2022 paper Large Language Models are Zero-Shot Reasoners. This is that trick where, if you get a model to talk out loud about a problem it’s solving, you often get a result which the model would not have achieved otherwise. o1 takes this process and further bakes it into the model itself. The details are somewhat obfuscated: o1 models spend “reasoning tokens” thinking through the problem that are not directly visible to the user (though the ChatGPT UI shows a summary of them), then outputs a final result.
  What are some of the capabilities of Large Language Models (LLMs) mentioned in the context?	Here’s the sequel to this post: Things we learned about LLMs in 2024. Large Language Models In the past 24-36 months, our species has discovered that you can take a GIANT corpus of text, run it through a pile of GPUs, and use it to create a fascinating new kind of software. LLMs can do a lot of things. They can answer questions, summarize documents, translate from one language to another, extract information and even write surprisingly competent code. They can also help you cheat at your homework, generate unlimited streams of fake content and be used for all manner of nefarious purposes.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 10
• per_device_eval_batch_size: 10
• num_train_epochs: 10
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 10
• per_device_eval_batch_size: 10
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 10
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: False
• fp16: False
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: False
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• dispatch_batches: None
• split_batches: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	cosine_ndcg@10
1.0	16	0.8768
2.0	32	0.9317
3.0	48	0.9484
3.125	50	0.9638
4.0	64	0.9692
5.0	80	0.9692
6.0	96	0.9692
6.25	100	0.9692
7.0	112	0.9692
8.0	128	0.9692
9.0	144	0.9692
9.375	150	0.9692
10.0	160	0.9692

Framework Versions

• Python: 3.11.11
• Sentence Transformers: 3.4.1
• Transformers: 4.48.3
• PyTorch: 2.5.1+cu124
• Accelerate: 1.3.0
• Datasets: 3.3.1
• Tokenizers: 0.21.0

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}