README.md

metadata

language:
  - en
license: apache-2.0
library_name: transformers
datasets:
  - cerebras/SlimPajama-627B
metrics:
  - accuracy
base_model: keeeeenw/MicroLlama
tags:
  - TensorBlock
  - GGUF
model-index:
  - name: MicroLlama
    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: 19.85
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=keeeeenw/MicroLlama
          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: 2.83
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=keeeeenw/MicroLlama
          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=keeeeenw/MicroLlama
          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: 1.45
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=keeeeenw/MicroLlama
          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: 4.79
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=keeeeenw/MicroLlama
          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: 1.53
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=keeeeenw/MicroLlama
          name: Open LLM Leaderboard

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

keeeeenw/MicroLlama - GGUF

This repo contains GGUF format model files for keeeeenw/MicroLlama.

The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4011.

Prompt template

Model file specification

Filename	Quant type	File Size	Description
MicroLlama-Q2_K.gguf	Q2_K	0.117 GB	smallest, significant quality loss - not recommended for most purposes
MicroLlama-Q3_K_S.gguf	Q3_K_S	0.135 GB	very small, high quality loss
MicroLlama-Q3_K_M.gguf	Q3_K_M	0.145 GB	very small, high quality loss
MicroLlama-Q3_K_L.gguf	Q3_K_L	0.155 GB	small, substantial quality loss
MicroLlama-Q4_0.gguf	Q4_0	0.168 GB	legacy; small, very high quality loss - prefer using Q3_K_M
MicroLlama-Q4_K_S.gguf	Q4_K_S	0.169 GB	small, greater quality loss
MicroLlama-Q4_K_M.gguf	Q4_K_M	0.177 GB	medium, balanced quality - recommended
MicroLlama-Q5_0.gguf	Q5_0	0.200 GB	legacy; medium, balanced quality - prefer using Q4_K_M
MicroLlama-Q5_K_S.gguf	Q5_K_S	0.200 GB	large, low quality loss - recommended
MicroLlama-Q5_K_M.gguf	Q5_K_M	0.204 GB	large, very low quality loss - recommended
MicroLlama-Q6_K.gguf	Q6_K	0.233 GB	very large, extremely low quality loss
MicroLlama-Q8_0.gguf	Q8_0	0.302 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/MicroLlama-GGUF --include "MicroLlama-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/MicroLlama-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'