README.md · tensorblock/M4-ai_TinyMistral-248M-v3-GGUF at main

metadata

language:
  - en
license: apache-2.0
datasets:
  - Locutusque/TM-DATA-V2
  - LLM360/TxT360
  - mlfoundations/dclm-baseline-1.0
  - Skylion007/openwebtext
  - JeanKaddour/minipile
  - eminorhan/gutenberg_en
tags:
  - TensorBlock
  - GGUF
base_model: M4-ai/TinyMistral-248M-v3
model-index:
  - name: TinyMistral-248M-v3
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 16.39
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 1.78
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 0
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 0
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 5.15
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 1.47
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=M4-ai/TinyMistral-248M-v3
          name: Open LLM Leaderboard

Feedback and support: TensorBlock's Twitter/X, Telegram Group and Discord server

M4-ai/TinyMistral-248M-v3 - GGUF

This repo contains GGUF format model files for M4-ai/TinyMistral-248M-v3.

The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b5165.

Our projects

Awesome MCP Servers	TensorBlock Studio

A comprehensive collection of Model Context Protocol (MCP) servers.	A lightweight, open, and extensible multi-LLM interaction studio.
👀 See what we built 👀	👀 See what we built 👀

Prompt template

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Model file specification

Filename	Quant type	File Size	Description
TinyMistral-248M-v3-Q2_K.gguf	Q2_K	0.105 GB	smallest, significant quality loss - not recommended for most purposes
TinyMistral-248M-v3-Q3_K_S.gguf	Q3_K_S	0.120 GB	very small, high quality loss
TinyMistral-248M-v3-Q3_K_M.gguf	Q3_K_M	0.129 GB	very small, high quality loss
TinyMistral-248M-v3-Q3_K_L.gguf	Q3_K_L	0.137 GB	small, substantial quality loss
TinyMistral-248M-v3-Q4_0.gguf	Q4_0	0.149 GB	legacy; small, very high quality loss - prefer using Q3_K_M
TinyMistral-248M-v3-Q4_K_S.gguf	Q4_K_S	0.149 GB	small, greater quality loss
TinyMistral-248M-v3-Q4_K_M.gguf	Q4_K_M	0.156 GB	medium, balanced quality - recommended
TinyMistral-248M-v3-Q5_0.gguf	Q5_0	0.176 GB	legacy; medium, balanced quality - prefer using Q4_K_M
TinyMistral-248M-v3-Q5_K_S.gguf	Q5_K_S	0.176 GB	large, low quality loss - recommended
TinyMistral-248M-v3-Q5_K_M.gguf	Q5_K_M	0.179 GB	large, very low quality loss - recommended
TinyMistral-248M-v3-Q6_K.gguf	Q6_K	0.204 GB	very large, extremely low quality loss
TinyMistral-248M-v3-Q8_0.gguf	Q8_0	0.264 GB	very large, extremely low quality loss - not recommended

Downloading instruction

Command line

Firstly, install Huggingface Client

pip install -U "huggingface_hub[cli]"

Then, downoad the individual model file the a local directory

huggingface-cli download tensorblock/M4-ai_TinyMistral-248M-v3-GGUF --include "TinyMistral-248M-v3-Q2_K.gguf" --local-dir MY_LOCAL_DIR

If you wanna download multiple model files with a pattern (e.g., *Q4_K*gguf), you can try:

huggingface-cli download tensorblock/M4-ai_TinyMistral-248M-v3-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'