GitLab

* update test scripts for models with different sizes

* update

* only test after tunning gemm

* chmod +x

* fix typo

* benchmark on a100

* fix typo

* fix typo

* per-token latency percentile in profile_throughput

* fix

* fix

* rename

* make the script accept parameters

* minor fix

* indent

* reformat table

* change to 3000

* minor fix

* minor fix in the profile scripts and docs

* miss arguments

* typo

* fix lint

* update

* update profile scripts

* add top_p, top_k and temperature as input arguments

* fix input_ids

* update profile_throughput

* update profile_restful_api

* update profile_serving

* update

* update

* add progress bar

* remove TODO comments

* update

* remove useless profile_* argument

* remove log level

* change concurrency default value to 64

* update restful_api.md

* update according to review comments

* fix docstring

* fix benchmark serving computation mistake

* fix timestamps computations

* remove speed up

* no mp

* mp seems faster?

* remove

* update

* remove

* fix

* update

* update print log

* typo

* print fist token latency only stream==True

* remove renew_session

* update AsyncEngine

* make IPv6 compatible, safe run for coroutine interrupting

* instance_id -> session_id and fix api_client.py

* update doc

* remove useless faq

* safe ip mapping

* update app.py

* WIP completion

* completion

* update doc

* disable interactive mode for /v1/chat/completions

* docstring

* docstring

* refactor gradio

* update gradio

* udpate

* update doc

* rename

* session_id default -1

* missed two files

* add a APIClient

* add chat func for APIClient

* refine

* add concurrent function

* sequence_start, sequence_end --> interactive_mode

* update doc

* comments

* doc

* better text completion

* remove /v1/embeddings

* comments

* deprecate generate and use /v1/interactive/completions

* /v1/interactive/completion -> /v1/chat/interactive

* embeddings

* rename

* remove wrong arg description

* docstring

* fix

* update cli

* update doc

* strict session_len limit condition

* pass model args to api_server

* fix crash

* update profile_generation.py

* format

* use self.bos_id

* remove sys_instruct

* move tokenizer

* remove Tokenizer in init

* update deploy.py

* make IPv6 compatible, safe run for coroutine interrupting

* instance_id -> session_id and fix api_client.py

* update doc

* remove useless faq

* safe ip mapping

* update app.py

* remove print

* update doc

* fix benchmark serving cannot use Qwen tokenizer

* update benchmark readme

* better profiler

* wait for releasing mem

* remove fire

* remove support for multiple model benchmark

* comments

* output more details

* correct tp

* read data after start processes

* fix hang

* fix exceptions when request_output_len is 0

* support decode

* unit test and benckmark and improve

* remove some drafts

* enable numerical test

* minor

* add some benchmark data

* add more output

* update interface

* remove debugs

* format

* update docstring

* remove print and add benchmark results

* use logits & add main

* fix rb

* dump large

* update test

* update test decode

* add decimal

* add restful api

* refine

* add simple doc

* lint

* add uvicorn requirement

* more args

* add llama2

* docstring

* update doc

* save

* refine

* lint

* better decode

* add v1/embedding

* add GenerateRequest

* add llama2 chat template

* correct profiling

* update documents

* add length judge

* add faq

* update doc and rename req_que to req_queue

* fix md link, use get_logger, fix sequence_end bug

* use another doc link for go to avoid lint error

* add api_client.py

* update doc

* update doc

* update function interface

* update FAQ

* resolve comments

* initial

* add docs

* noqa on docstring

* support no streamer

* typo

* add accel in model name

* fix lint

* fix CSVWRitter typo

* typo

* tmp

* update

* update

* update

* update

* update

* remove

* update

* update

* change to incremental decoding

* update

* works on interlm and vicuna

* support GQA

* remove comment

* update readme, add logger, default tp=1

* remove log

* refactor model.py and support baichuan-7b

* remove model_name

* remove hard session_len

* export tokenizer.py to target dir

* remove model_name from client

* remove model_name

* update

* correct throughput equation

* fix session.response

* update serving.md

* update readme

* update according to review comments

* update

* update

* update

* update

* add profile throughput benchmark

* add output only throughput

* update req/min

* update benckmark readme

* fix lint

---------
Co-authored-by: grimoire <yaoqian@pjlab.org.cn>

* wip

* profile disable tp

* fix profile

* lint

* fix dlpack

* remove comment

* add tp flag

* add session len check

* add eos

* remove tp and session len inputs

* warp tokenizer

* multithread load weight

* update profile

* refactor tokenizer

* remove pre/post process

* remove mpi4py requirement

* remove

* remove bind

* remove mpi requirement

* check backend_tokenizer

* streaming-output

* fix end

* fix profile

* support chinese streaming

* lint

* update chat

* lint

* fix benchmark

---------
Co-authored-by: grimoire <yaoqian@pjlab.org.cn>

* add cal qparams

* support offload inference

* add collect funtions (mod,weight)

* stats kv scales

* update init

* add user guide

* fix hints

* fix comments & support turbomind format

* update user guide

* fix slice kv cache error & support pileval dataset (used in llm-awq)

* fix wrong num heads slice

* update default dataset

* fix conflict

* fix hints

* fix hints

* add gitignore

* wip

* wip

* example finish

* fix include and namespace

* wtf

* install lib

* batchize

* update cmake install

* multithread

* fix comment

* fix

* add mmengine

* bind llamamodel

---------
Co-authored-by: grimoire <yaoqian@pjlab.org.cn>

* rename lmdeploy/serve/fastertransformer to lmdeploy/serve/turbomind

* update

* update

* change llmdeploy to lmdeploy

* update logo

* update readme

* remove constraints on model name

* remove duplicate model converter

* add profile

* get eos and bos from server

* update stop_words

* update sequence_length when the last generated token is eos_id

* fix

* fix

* check-in models

* valicate model_name

* make stop_words as property

* debug profiling

* better stats

* fix assistant reponse

* update profile serving

* update

* update