metadata

license: apache-2.0
thumbnail: https://i.ibb.co/TvyMrRc/rsz-smol-llama-banner.png
language:
  - en
inference:
  parameters:
    max_new_tokens: 64
    do_sample: true
    temperature: 0.8
    repetition_penalty: 1.15
    no_repeat_ngram_size: 4
    eta_cutoff: 0.0006
    renormalize_logits: true
widget:
  - text: My name is El Microondas the Wise, and
    example_title: El Microondas
  - text: Kennesaw State University is a public
    example_title: Kennesaw State University
  - text: >-
      Bungie Studios is an American video game developer. They are most famous
      for developing the award winning Halo series of video games. They also
      made Destiny. The studio was founded
    example_title: Bungie
  - text: The Mona Lisa is a world-renowned painting created by
    example_title: Mona Lisa
  - text: >-
      The Harry Potter series, written by J.K. Rowling, begins with the book
      titled
    example_title: Harry Potter Series
  - text: >-
      Question: I have cities, but no houses. I have mountains, but no trees. I
      have water, but no fish. What am I?

      Answer:
    example_title: Riddle
  - text: The process of photosynthesis involves the conversion of
    example_title: Photosynthesis
  - text: >-
      Jane went to the store to buy some groceries. She picked up apples,
      oranges, and a loaf of bread. When she got home, she realized she forgot
    example_title: Story Continuation
  - text: >-
      Problem 2: If a train leaves Station A at 9:00 AM and travels at 60 mph,
      and another train leaves Station B at 10:00 AM and travels at 80 mph, when
      will they meet if the distance between the stations is 300 miles?

      To determine
    example_title: Math Problem
  - text: In the context of computer programming, an algorithm is
    example_title: Algorithm Definition
pipeline_tag: text-generation
tags:
  - smol_llama
  - llama2
datasets:
  - JeanKaddour/minipile
  - pszemraj/simple_wikipedia_LM
  - BEE-spoke-data/wikipedia-20230901.en-deduped
  - mattymchen/refinedweb-3m

smol_llama-101M-GQA

A small 101M param (total) decoder model. This is the first version of the model.

768 hidden size, 6 layers
GQA (24 heads, 8 key-value), context length 1024
train-from-scratch

Features

Some cool anecdotes about this model:

this model was pretrained on one GPU for 5 compute-days. You can DIY pretrain too!
0% of the training data (to our knowledge) comes from OpenAI synthetic generation

Notes

This checkpoint is the 'raw' pre-trained model and has not been tuned to a more specific task. It should be fine-tuned before use in most cases.

Checkpoints & Links

smol-er 81M parameter checkpoint with in/out embeddings tied: here
Fine-tuned on pypi to generate Python code - link
For the chat version of this model, please see here

Citation Info

If you find this experiment useful and would like to add some words to your .bib file, it would make us happy.

@misc {beespoke_data_2023,
    author       = { {Peter Szemraj and Vincent Haines} },
    title        = { smol_llama-101M-GQA (Revision 9c9c090) },
    year         = 2023,
    url          = { https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA },
    doi          = { 10.57967/hf/1440 },
    publisher    = { Hugging Face }
}

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	25.32
ARC (25-shot)	23.55
HellaSwag (10-shot)	28.77
MMLU (5-shot)	24.24
TruthfulQA (0-shot)	45.76
Winogrande (5-shot)	50.67
GSM8K (5-shot)	0.83
DROP (3-shot)	3.39