Tasks	Version	Filter	Metric	Value		Stderr
arc_challenge	Yaml	none	acc	0.2619	±	0.0128
		none	acc_norm	0.2892	±	0.0133
arc_easy	Yaml	none	acc	0.4777	±	0.0102
		none	acc_norm	0.4461	±	0.0102
boolq	Yaml	none	acc	0.6297	±	0.0084
hellaswag	Yaml	none	acc	0.3934	±	0.0049
		none	acc_norm	0.4930	±	0.0050
openbookqa	Yaml	none	acc	0.2120	±	0.0183
		none	acc_norm	0.3260	±	0.0210
piqa	Yaml	none	acc	0.6915	±	0.0108
		none	acc_norm	0.6877	±	0.0108
winogrande	Yaml	none	acc	0.5714	±	0.0139

Llamafactory EVAL

!CUDA_VISIBLE_DEVICES=0 python src/evaluate.py
--model_name_or_path Deathsquad10/TinyLlama-Remix
--template vanilla
--task mmlu
--split test
--lang en
--n_shot 5
--use_unsloth
--batch_size 1

       Average: 26.29
       STEM: 27.10
       Social Sciences: 25.48
       Humanities: 25.62
       Other: 27.26

!CUDA_VISIBLE_DEVICES=0 python src/evaluate.py
--model_name_or_path Deathsquad10/TinyLlama-Remix
--template vanilla
--task cmmlu
--split test
--lang en
--n_shot 5
--use_unsloth
--batch_size 2

      Average: 24.98
      STEM: 25.52
      Social Sciences: 24.70
      Humanities: 24.59
      Other: 25.19

https://github.com/jzhang38/TinyLlama

The TinyLlama project aims to pretrain a 1.1B Llama model on 3 trillion tokens. With some proper optimization, we can achieve this within a span of "just" 90 days using 16 A100-40G GPUs 🚀🚀. The training has started on 2023-09-01.

We adopted exactly the same architecture and tokenizer as Llama 2. This means TinyLlama can be plugged and played in many open-source projects built upon Llama. Besides, TinyLlama is compact with only 1.1B parameters. This compactness allows it to cater to a multitude of applications demanding a restricted computation and memory footprint.

This Model

This is the chat model finetuned on top of TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T. We follow HF's Zephyr's training recipe. The model was " initially fine-tuned on a variant of the UltraChat dataset, which contains a diverse range of synthetic dialogues generated by ChatGPT. We then further aligned the model with 🤗 TRL's DPOTrainer on the openbmb/UltraFeedback dataset, which contain 64k prompts and model completions that are ranked by GPT-4."

How to use

You will need the transformers>=4.34 Do check the TinyLlama github page for more information.

python

# Install transformers from source - only needed for versions <= v4.34
# pip install git+https://github.com/huggingface/transformers.git
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="auto")
# We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
messages = [
    {
        "role": "system",
        "content": "You are a friendly chatbot who always responds in the style of a pirate",
    },
    {"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
# <|system|>
# You are a friendly chatbot who always responds in the style of a pirate.</s>
# <|user|>
# How many helicopters can a human eat in one sitting?</s>
# <|assistant|>
# ...