training-deploy-llava16-on-sagemaker

sft llava-1.6

https://github.com/jackie930/training-deploy-llava16-on-sagemaker

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

✓
CITATION.cff file
Found CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
○
DOI references
○
Academic publication links
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (5.2%) to scientific vocabulary

Last synced: 9 months ago · JSON representation ·

Repository

sft llava-1.6

Basic Info

Host: GitHub
Owner: jackie930
License: apache-2.0
Language: Python
Default Branch: main
Size: 1.01 MB

Statistics

Stars: 1
Watchers: 1
Forks: 0
Open Issues: 0
Releases: 0

Created over 1 year ago · Last pushed over 1 year ago

Metadata Files

Readme Contributing License Code of conduct Citation

README.md

Train and deploy lava-v1.6-mistral-7b-hf on AWS

Overview

本项目提供了在 AWS 上微调和部署 LLaVA-v1.6-Mistral-7B 多模态大语言模型的完整流程。主要特性包括: * 基于自定义数据集进行监督式微调(SFT) * 使用 Amazon SageMaker 进行大规模模型训练 * 支持 VLLM 加速推理的模型部署 * 完整的数据预处理和评估流程

Environments

Python 3.8+
PyTorch 2.0+
推荐使用 CUDA 11.7+
至少 24GB 显存的 GPU (推荐 A10G/H100/A100)

Installation

bash pip install trl

prepare data

here we need to do three steps: * prepare the data into specific format

```bash 将原始数据转换为指定格式:

python preprocess.py --datapath {INPUTCSV} --outputfolder {OUTPUTDATAFOLDER} ``` 参数说明: * datapath: 输入CSV文件路径 * output_folder: 输出文件夹路径

创建 HuggingFace Dataset

将处理后的数据保存为 HuggingFace dataset 格式: bash python prepare.py

Command Line Interface (CLI)

You can use the TRL Command Line Interface (CLI) to quickly get started with Supervised Fine-tuning (SFT) and Direct Preference Optimization (DPO), or vibe check your model with the chat CLI:

SFT:

bash sh train_customer.sh

test the trained model locally (optional)

bash python local_test.py

deploy the trained model as sagemaker endpoint which support VLLM (optional)

follow deply/llavav16mistral7bdeploysagemakerg5xl.ipynb

Owner

Name: Jackie Liu
Login: jackie930
Kind: user
Company: Amazon Web Services

Repositories: 6
Profile: https://github.com/jackie930

Applied Scientist

Citation (CITATION.cff)

cff-version: 1.2.0
title: 'TRL: Transformer Reinforcement Learning'
message: >-
  If you use this software, please cite it using the
  metadata from this file.
type: software
authors:
  - given-names: Leandro
    family-names: von Werra
  - given-names: Younes
    family-names: Belkada
  - given-names: Lewis
    family-names: Tunstall
  - given-names: Edward
    family-names: Beeching
  - given-names: Tristan
    family-names: Thrush
  - given-names: Nathan
    family-names: Lambert
  - given-names: Shengyi
    family-names: Huang
  - given-names: Kashif
    family-names: Rasul
  - given-names: Quentin
    family-names: Gallouédec
repository-code: 'https://github.com/huggingface/trl'
abstract: "With trl you can train transformer language models with Proximal Policy Optimization (PPO). The library is built on top of the transformers library by \U0001F917 Hugging Face. Therefore, pre-trained language models can be directly loaded via transformers. At this point, most decoder and encoder-decoder architectures are supported."
keywords:
  - rlhf
  - deep-learning
  - pytorch
  - transformers
license: Apache-2.0
version: 0.12

GitHub Events

Total

Watch event: 1
Push event: 1

Last Year

Watch event: 1
Push event: 1

Dependencies

docker/trl-latest-gpu/Dockerfile docker

continuumio/miniconda3 latest build
nvidia/cuda 12.2.2-devel-ubuntu22.04 build

docker/trl-source-gpu/Dockerfile docker

continuumio/miniconda3 latest build
nvidia/cuda 12.2.2-devel-ubuntu22.04 build

examples/research_projects/stack_llama_2/scripts/requirements.txt pypi

accelerate *
bitsandbytes *
datasets *
peft *
transformers *
trl *
wandb *

pyproject.toml pypi

requirements.txt pypi

accelerate *
datasets *
rich *
transformers >=4.46.0

setup.py pypi

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Open Source Science