Skip to content
Other

ToolCall-200M-DPO

Experimental Language Model for Function Calling

1 min read

This model is a checkpoint fine-tuned with DPO (Direct Preference Optimization) to improve format adherence. The base 200M model was trained from scratch on 4 billion tokens using free Kaggle sessions. Experiments show that format optimization reduces hanging and improves refusals for irrelevant queries, though tool selection accuracy drops slightly.

Key features

  • DPO optimization: Focused on strict adherence to structured output formats.
  • Improved refusals: The model makes fewer errors on irrelevant prompts and hits token limits less often.
  • Compact size: Only 200 million parameters for lightweight local tasks.
  • Accuracy limitations: Tool and argument selection accuracy is lower compared to the base version.

What you can do

  • Test responses: Evaluate the model's function calling behavior compared to the SFTv4 version.
  • Pipeline integration: Use it to study how small language models behave under format constraints.

For general use, developers recommend the base SFTv4 version, while this checkpoint is interesting for studying the impact of DPO on tool-calling accuracy.

Read moreDetails
Context: 2,048 tokens
Tool selection accuracy: 66.87% (vs. 72.80% for base)
Exact calls: 21.73% (vs. 23.71% for base)
Correct refusals: 69.17% (vs. 65.42% for base)
Discuss Telegram 0

Will it run?

Runs in F32

Fits without quantisation.

  • F32weights 0.7 GB · ready files1.3 GB
  • BF16weights 0.4 GB0.9 GB
  • FP8 / INT8recommendedweights 0.2 GB0.7 GB
  • 4-bitweights 0.1 GB0.6 GB

Tip: go 8-bit.FP8 / INT8 is within about a percent of full precision in quality, takes half the memory and runs faster on RTX 40/50 and H100. Full precision only pays off for fine-tuning.

What to get for this model

Massed ComputeOpen Massed Compute

Your own GPU: FP8 / INT8 — from RTX 2060 · F32 — from RTX 2060

Massed Compute hourly prices, November 2026. The discounted price applies with the coupon at checkout.

Estimate: weights plus working memory. Long videos, high resolutions and long contexts need more.

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # vaibhav-mugulavalli-2004/ToolCall-200M-DPO
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download vaibhav-mugulavalli-2004/ToolCall-200M-DPO --local-dir ~/models/ToolCall-200M-DPO
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to ToolCall-200M-DPO
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of ToolCall-200M-SFT-v4 (quants, finetunes, LoRAs) will show up under the bell.

Fine-tunes 1All on Hugging Face

About the model

vaibhav-mugulavalli-2004/ToolCall-200M-DPO
Author
vaibhav-mugulavalli-2004
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments