Vllm2#27

Open

darrenearl wants to merge 148 commits intomainfrom

Commits on Apr 5, 2024

Initial commit
yanghaojin
authored

Commits on Apr 7, 2024

added dirs and other files
Haojin Yang
committed
added content for readme
Haojin Yang
committed

Commits on Apr 8, 2024

adapted project structure
Haojin Yang
committed

Commits on Apr 9, 2024

added model loading
Haojin Yang
committed

Commits on Apr 10, 2024

added inference codes
Haojin Yang
committed
created simple generation
Haojin Yang
committed

Commits on Apr 11, 2024

Commits on Apr 12, 2024

Commits on Apr 13, 2024

Added an additional library installation information and some minor modifications.
Haojin Yang
committed
Merge pull request #3 from GreenBitAI/feature/evaluation
yanghaojin
authored
improved inference package, readme
Haojin Yang
committed
added draft setup.py
Haojin Yang
committed
Merge pull request #4 from GreenBitAI/feature/chat-cli
yanghaojin
authored

Commits on Apr 15, 2024

improved requirement and reame in project root
Haojin Yang
committed
Merge pull request #5 from GreenBitAI/feature/chat-cli
yanghaojin
authored
added version information into package init.
Haojin Yang
committed
added bitorch engine into the requirement section of the readme file.
Haojin Yang
committed
add zero-shot results
NicoNico6
committed

Commits on Apr 16, 2024

support sft finetune.py
NicoNico6
committed

Commits on Apr 17, 2024

added one constraint in the traing arguments
Haojin Yang
committed

Commits on Apr 18, 2024

Added readme file (not finished). Added optional settings for Galore.
Haojin Yang
committed
fixed error params
Haojin Yang
committed
removed used debug code
Haojin Yang
committed
Galore strategy integrated.
Haojin Yang
committed

Commits on Apr 19, 2024

galore works with low bit qweights.
Haojin Yang
committed

Commits on Apr 20, 2024

Improve readme file for sft package
Haojin Yang
committed
added modified bnb 8-bit optimizer nad galore support
Haojin Yang
committed

Commits on Apr 21, 2024

fixed issues in adamw8bit optim
Haojin Yang
committed

Commits on Apr 22, 2024

Adapted sft/README.md
Haojin Yang
committed
support peft library
NicoNico6
committed
update peft lora command
NicoNico6
committed
add __init__
NicoNico6
committed

Commits on Apr 23, 2024

Added a current limitation into readme. Added batch size arg into finetune.py
Haojin Yang
committed
Merge branch 'feature/sft' of https://github.com/GreenBitAI/green-bit-llm into feature/sft
Haojin Yang
committed
Added missing comments to the classes and functions. Simplified the args parser for sft and inference package.
Haojin Yang
committed
Fixed error import
Haojin Yang
committed

Commits on Apr 24, 2024

Adapted string comparison.
Haojin Yang
committed
Merge branch 'feature/sft' of https://github.com/GreenBitAI/green-bit-llm into feature/sft
Haojin Yang
committed

Commits on Apr 25, 2024

Commits on Apr 26, 2024

Added autoconfig.
Haojin Yang
committed
Corrected the issue of conversation template of phi3
Haojin Yang
committed
Fixed various issues in finetune and peft. Improved readme.
Haojin Yang
committed
Added demo gif images and adapted readme file.
Haojin Yang
committed
Adapted project file path, fixed corresponding errors.
Haojin Yang
committed
Fixed error in inference model args. Added colored info message into evaluation.
Haojin Yang
committed
Use find_packages() in setup.py
Haojin Yang
committed
Merge pull request #6 from GreenBitAI/feature/sft
yanghaojin
authored

Commits on Apr 29, 2024

Commits on May 1, 2024

update evaluation info

nianhui.guo
authored and
nianhui.guo
committed
update evaluation command

nianhui.guo
authored and
nianhui.guo
committed
Merge pull request #7 from GreenBitAI/feature/sft
yanghaojin
authored
Update README.md
yanghaojin
authored
Added changelog file.
Haojin Yang
committed
fix dataset loading issue

nianhui.guo
authored and
nianhui.guo
committed

Commits on May 14, 2024

Commits on May 15, 2024

Added the initial support for a classical gptq model, we will run it using our mpqlinear layer.
Haojin Yang
committed

Commits on May 18, 2024

Commits on May 20, 2024

Commits on May 22, 2024

Updated version.py
Haojin Yang
committed

Commits on May 25, 2024

Merge pull request #14 from GreenBitAI/feature/gptq_model
yanghaojin
authored

Commits on May 26, 2024

Commits on Jun 4, 2024

Commits on Oct 14, 2024

Developing langchain integration
Haojin Yang
committed

Commits on Oct 15, 2024

Added langchain pipeline, chatmodel, corresponding unit tests, and a local rag demo.
Haojin Yang
committed
Local rag demo works; solved issues in unit tests.
Haojin Yang
committed

Commits on Oct 16, 2024

Adjusted the embedding model used in local RAG demo.
Haojin Yang
committed

Commits on Oct 18, 2024

Fixed an error in README of langchain package
Haojin Yang
committed

Commits on Nov 24, 2024

Created fastapi server, routing method.
Haojin Yang
committed

Commits on Nov 25, 2024

Added max_tokens control in pipeline
Haojin Yang
committed

Commits on Nov 26, 2024

Adapted some configuration names.
Haojin Yang
committed
Small changes in curl testing script.
Haojin Yang
committed
Merge branch 'feature/langchain_integration' of https://github.com/GreenBitAI/green-bit-llm into feature/langchain_integration
Haojin Yang
committed

Commits on Dec 1, 2024

Added token usage information into stream methods.
Haojin Yang
committed

Commits on Dec 3, 2024

Finally fixed issue in pipline and chat model class for llama3 and qwen 2.5
Haojin Yang
committed
Merge branch 'feature/langchain_integration' of https://github.com/GreenBitAI/green-bit-llm into feature/langchain_integration
Haojin Yang
committed
Now the fastapi server works!
Haojin Yang
committed
Updated testing script for api server.
Haojin Yang
committed
Merge branch 'feature/langchain_integration' of https://github.com/GreenBitAI/green-bit-llm into feature/langchain_integration
Haojin Yang
committed

Commits on Dec 4, 2024

Merge pull request #24 from GreenBitAI/feature/langchain_integration
yanghaojin
authored

Commits on Dec 24, 2024

Commits on Jan 9, 2025

Commits on Jan 20, 2025

add the evaluate_vllm.py
darrenearl
committed