README.md

metadata

language:
  - en
license: llama3
library_name: transformers
tags:
  - mathematics
  - TensorBlock
  - GGUF
datasets:
  - hkust-nlp/dart-math-hard
metrics:
  - accuracy
pipeline_tag: text-generation
base_model: hkust-nlp/dart-math-llama3-8b-prop2diff
model-index:
  - name: dart-math-llama3-8b-prop2diff
    results:
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: MATH
          type: hendrycks/competition_math
          split: test
        metrics:
          - type: accuracy
            value: 46.6
            name: Pass@1 (0-shot CoT)
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: GSM8K
          type: openai/gsm8k
          config: main
          split: test
        metrics:
          - type: accuracy
            value: 81.1
            name: Pass@1 (0-shot CoT)
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: CollegeMath
          type: college-math
        metrics:
          - type: accuracy
            value: 28.8
            name: Pass@1 (0-shot CoT)
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: DeepMind-Mathematics
          type: deepmind-mathematics
        metrics:
          - type: accuracy
            value: 48
            name: Pass@1 (0-shot CoT)
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: OlympiadBench-OE_TO_maths_en_COMP
          type: Hothan/OlympiadBench
          config: OE_TO_maths_en_COMP
          split: train
        metrics:
          - type: accuracy
            value: 14.5
            name: Pass@1 (0-shot CoT)
      - task:
          type: text-generation
          name: Mathematical Problem-Solving
        dataset:
          name: TheoremQA
          type: TIGER-Lab/TheoremQA
          split: test
        metrics:
          - type: accuracy
            value: 19.4
            name: Pass@1 (0-shot CoT)

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

hkust-nlp/dart-math-llama3-8b-prop2diff - GGUF

This repo contains GGUF format model files for hkust-nlp/dart-math-llama3-8b-prop2diff.

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

Run them on the TensorBlock client using your local machine ↗

Prompt template

Model file specification

Filename	Quant type	File Size	Description
dart-math-llama3-8b-prop2diff-Q2_K.gguf	Q2_K	3.179 GB	smallest, significant quality loss - not recommended for most purposes
dart-math-llama3-8b-prop2diff-Q3_K_S.gguf	Q3_K_S	3.665 GB	very small, high quality loss
dart-math-llama3-8b-prop2diff-Q3_K_M.gguf	Q3_K_M	4.019 GB	very small, high quality loss
dart-math-llama3-8b-prop2diff-Q3_K_L.gguf	Q3_K_L	4.322 GB	small, substantial quality loss
dart-math-llama3-8b-prop2diff-Q4_0.gguf	Q4_0	4.662 GB	legacy; small, very high quality loss - prefer using Q3_K_M
dart-math-llama3-8b-prop2diff-Q4_K_S.gguf	Q4_K_S	4.693 GB	small, greater quality loss
dart-math-llama3-8b-prop2diff-Q4_K_M.gguf	Q4_K_M	4.921 GB	medium, balanced quality - recommended
dart-math-llama3-8b-prop2diff-Q5_0.gguf	Q5_0	5.600 GB	legacy; medium, balanced quality - prefer using Q4_K_M
dart-math-llama3-8b-prop2diff-Q5_K_S.gguf	Q5_K_S	5.600 GB	large, low quality loss - recommended
dart-math-llama3-8b-prop2diff-Q5_K_M.gguf	Q5_K_M	5.733 GB	large, very low quality loss - recommended
dart-math-llama3-8b-prop2diff-Q6_K.gguf	Q6_K	6.596 GB	very large, extremely low quality loss
dart-math-llama3-8b-prop2diff-Q8_0.gguf	Q8_0	8.541 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/dart-math-llama3-8b-prop2diff-GGUF --include "dart-math-llama3-8b-prop2diff-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/dart-math-llama3-8b-prop2diff-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'