llama2

llama2-7b-glora 🦙

This model was built via parameter-efficient GLoRA finetuning of llama2-7b on the shareGPT dataset. We adapt only the attention layers using GLoRA.

Model Description

The architecture is similar to LLaMA2-7B, but the bias is true for attention layers.

Limitations and Biases

The following language is modified from EleutherAI's GPT-NeoX-20B

This model can produce factually incorrect output, and should not be relied on to produce factually accurate information. This model was trained on various public datasets. While great efforts have been taken to clean the pretraining data, it is possible that this model could generate lewd, biased or otherwise offensive outputs.

How to Use

Install and import the package dependencies:

!pip install -q -U huggingface_hub transformers torch accelerate

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig

Basic model loading:

model = AutoModelForCausalLM.from_pretrained(
    "MBZUAI-LLM/LLaMA2-7B-GLoRA-ShareGPT",
    use_auth_token=True,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained("MBZUAI-LLM/LLaMA2-7B-GLoRA-ShareGPT")

Once loaded, the model and tokenizer can be used with the following code:

def llama_generate(
    model: AutoModelForCausalLM,
    tokenizer: AutoTokenizer,
    prompt: str,
    max_new_tokens: int = 128,
    temperature: float = 0.92,
) -> str:
    """
    Initialize the pipeline
    Uses Hugging Face GenerationConfig defaults
        https://huggingface.co/docs/transformers/v4.29.1/en/main_classes/text_generation#transformers.GenerationConfig
    Args:
        model (transformers.AutoModelForCausalLM): Model for text generation
        tokenizer (transformers.AutoTokenizer): Tokenizer for model
        prompt (str): Prompt for text generation
        max_new_tokens (int, optional): Max new tokens after the prompt to generate. Defaults to 128.
        temperature (float, optional): The value used to modulate the next token probabilities.
            Defaults to 1.0
    """
    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    inputs = tokenizer(
        [prompt],
        return_tensors="pt",
        return_token_type_ids=False,
    ).to(
        device
    )  # tokenize inputs, load on device
    # when running Torch modules in lower precision, it is best practice to use the torch.autocast context manager.
    with torch.autocast("cuda", dtype=torch.bfloat16):
        response = model.generate(
            **inputs,
            max_new_tokens=max_new_tokens,
            temperature=temperature,
            return_dict_in_generate=True,
            eos_token_id=tokenizer.eos_token_id,
            pad_token_id=tokenizer.pad_token_id,
        )
    decoded_output = tokenizer.decode(
        response["sequences"][0],
        skip_special_tokens=True,
    )  # grab output in natural language
    return decoded_output[len(prompt) :]  # remove prompt from output

We can now generate text! For example:

prompt = "You are a helpful assistant. Tell me a recipe for vegan banana bread.\n"
response = llama_generate(
    model,
    tokenizer,
    prompt,
    max_new_tokens=500,
    temperature=0.92,
)
print(response)

Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.

Citation for GLoRA

@misc{chavan2023oneforall,
      title={One-for-All: Generalized LoRA for Parameter-Efficient Fine-tuning}, 
      author={Arnav Chavan and Zhuang Liu and Deepak Gupta and Eric Xing and Zhiqiang Shen},
      year={2023},
      eprint={2306.07967},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}