Llama2 Xs 460M Experimental

🚀 LLaMa Lite: Reduced-Scale, Experimental Versions of LLaMA and LLaMa 2

This series of repos offers an open - source reproduction of Meta AI's LLaMA and LLaMa 2 large language models. The models here have significantly reduced sizes, providing an experimental alternative.

🚀 Quick Start

Using with HuggingFace Transformers

The experimental checkpoints can be directly loaded by the Transformers library. The following code snippet shows how to load our experimental model and generate text with it.

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

# model_path = 'ahxt/llama2_xs_460M_experimental'
model_path = 'ahxt/llama1_s_1.8B_experimental'

model = AutoModelForCausalLM.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)
model.eval()

prompt = 'Q: What is the largest bird?\nA:'
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
tokens = model.generate(input_ids, max_length=20)
print( tokenizer.decode(tokens[0].tolist(), skip_special_tokens=True) )
# Q: What is the largest bird?\nA: The largest bird is the bald eagle.

✨ Features

In this series of repos, we present an open - source reproduction of Meta AI's [LLaMA](https://ai.meta.com/blog/large - language - model - llama - meta - ai/) and LLaMa 2 large language models. However, with significantly reduced model sizes, the experimental version of llama1_s has 1.8B parameters, and the experimental version of llama2_xs has 460M parameters. ('s' stands for small, while 'xs' denotes extra small).

📚 Documentation

Dataset and Tokenization

We train our models on part of RedPajama dataset. We use the GPT2Tokenizer to tokenize the text.

📊 Evaluation

We evaluate our models on the MMLU task.

📊 Open LLM Leaderboard Evaluation Results

Detailed results can be found [here](https://huggingface.co/datasets/open - llm - leaderboard/details_ahxt__llama2_xs_460M_experimental)

📞 Contact

This experimental version is developed by Xiaotian Han from Texas A&M University. And these experimental versions are for research only.

📋 Metadata