Hugging Face's logo

tensorblock
/
megatron-gpt2-345m-GGUF

GGUF
English
gpt2
TensorBlock
GGUF
Eval Results
Inference Endpoints
Model card Files Community
megatron-gpt2-345m-GGUF / README.md
morriszms's picture
morriszms
Update README.md
1f834d6 verified
preview code
|
raw
history blame contribute delete
5.47 kB
metadata
language:
  - en
tags:
  - gpt2
  - TensorBlock
  - GGUF
license: apache-2.0
widget:
  - text: >-
      It was a bright cold day in April, and the clocks were striking thirteen.
      Winston Smith,
datasets:
  - wikitext
  - openwebtext
  - spacemanidol/cc-stories
base_model: robowaifudev/megatron-gpt2-345m
model-index:
  - name: megatron-gpt2-345m
    results:
      - task:
          type: text-generation
          name: Text generation
        dataset:
          name: WikiText-103
          type: wikitext
        metrics:
          - type: wikitext
            value: 19.31
            name: Perplexity
          - type: wikitext
            value: 17.151
            name: Perplexity
      - task:
          type: text-generation
          name: Text generation
        dataset:
          name: LAMBADA
          type: lambada
        metrics:
          - type: lambada
            value: 5.509
            name: Perplexity
          - type: lambada
            value: 68.31%
            name: Accuracy
TensorBlock

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

robowaifudev/megatron-gpt2-345m - GGUF

This repo contains GGUF format model files for robowaifudev/megatron-gpt2-345m.

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
megatron-gpt2-345m-Q2_K.gguf Q2_K 0.166 GB smallest, significant quality loss - not recommended for most purposes
megatron-gpt2-345m-Q3_K_S.gguf Q3_K_S 0.188 GB very small, high quality loss
megatron-gpt2-345m-Q3_K_M.gguf Q3_K_M 0.213 GB very small, high quality loss
megatron-gpt2-345m-Q3_K_L.gguf Q3_K_L 0.227 GB small, substantial quality loss
megatron-gpt2-345m-Q4_0.gguf Q4_0 0.231 GB legacy; small, very high quality loss - prefer using Q3_K_M
megatron-gpt2-345m-Q4_K_S.gguf Q4_K_S 0.233 GB small, greater quality loss
megatron-gpt2-345m-Q4_K_M.gguf Q4_K_M 0.252 GB medium, balanced quality - recommended
megatron-gpt2-345m-Q5_0.gguf Q5_0 0.272 GB legacy; medium, balanced quality - prefer using Q4_K_M
megatron-gpt2-345m-Q5_K_S.gguf Q5_K_S 0.272 GB large, low quality loss - recommended
megatron-gpt2-345m-Q5_K_M.gguf Q5_K_M 0.288 GB large, very low quality loss - recommended
megatron-gpt2-345m-Q6_K.gguf Q6_K 0.316 GB very large, extremely low quality loss
megatron-gpt2-345m-Q8_0.gguf Q8_0 0.407 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/megatron-gpt2-345m-GGUF --include "megatron-gpt2-345m-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/megatron-gpt2-345m-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'