Llama-2-7b-chat-hf-4bit-AWQ

Model Card for alokabhishek/Llama-2-7b-chat-hf-4bit-AWQ

<!-- Provide a quick summary of what the model is/does. -->

This repo contains 4-bit quantized (using AutoAWQ) model of Meta's meta-llama/Llama-2-7b-chat-hf

AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration is developed by MIT-HAN-Lab

Model Details

• Model creator: Meta
• Original model: Llama-2-7b-chat-hf

About 4 bit quantization using AutoAWQ

• AutoAWQ github repo: AutoAWQ github repo
• MIT-han-lab llm-aws github repo: MIT-han-lab llm-aws github repo

@inproceedings{lin2023awq, title={AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration}, author={Lin, Ji and Tang, Jiaming and Tang, Haotian and Yang, Shang and Chen, Wei-Ming and Wang, Wei-Chen and Xiao, Guangxuan and Dang, Xingyu and Gan, Chuang and Han, Song}, booktitle={MLSys}, year={2024} }

How to Get Started with the Model

Use the code below to get started with the model.

How to run from Python code

First install the package

!pip install autoawq
!pip install accelerate

Import

import torch
import os
from torch import bfloat16
from huggingface_hub import login, HfApi, create_repo
from transformers import AutoTokenizer, pipeline
from awq import AutoAWQForCausalLM

Use a pipeline as a high-level helper

# define the model ID
model_id_llama = "alokabhishek/Llama-2-7b-chat-hf-4bit-AWQ"

# Load model
tokenizer_llama = AutoTokenizer.from_pretrained(model_id_llama, use_fast=True)
model_llama = AutoAWQForCausalLM.from_quantized(model_id_llama, fuse_layer=True, trust_remote_code = False, safetensors = True)

# Set up the prompt and prompt template. Change instruction as per requirements.
prompt_llama = "Tell me a funny joke about Large Language Models meeting a Blackhole in an intergalactic Bar."
fromatted_prompt = f'''[INST] <<SYS>> You are a helpful, and fun loving assistant. Always answer as jestfully as possible. <</SYS>> {prompt_llama} [/INST] '''

tokens = tokenizer_llama(fromatted_prompt, return_tensors="pt").input_ids.cuda()

# Generate output, adjust parameters as per requirements
generation_output = model_llama.generate(tokens, do_sample=True, temperature=1.7, top_p=0.95, top_k=40, max_new_tokens=512)

# Print the output
print(tokenizer_llama.decode(generation_output[0], skip_special_tokens=True))

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

[More Information Needed]

Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

[More Information Needed]

Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

[More Information Needed]

Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]