fix sampling in chat stream

fix slow long sequence inference

remove unused generation configs

remove unused generation configs

Update modeling_qwen.py

update tokenization and readme

deprecate argument stream in model.chat()

update support for flash attn

fix chat streaming

add support for flash attn 2

shard the pytorch_model.bin into parts

update readme

fix kwargs in generate method and update readme

refactor tokenization and update readme

update config and streaming generation

update config about model precision, fix apply_rotary_pos_emb

revert convert_tokens_to_string

update quickusage and example

update readme and fix convert_tokens_to_string

support cpu inference, format file (#9)

fix decoder, and provide an option to remove attack rejection (#8)

Update Readme.md, fix typo

Update modeling_qwen.py, fix logn bug

fix overflow

Update config.json

update fast usage

update humaneval score

update fast usage

Upload wanx_colorful_black.png

implement _convert_id_to_token

fix flash-attention usage

Update README.md

update tokenization_qwen.py

fix report link

update demo link

update tokenization_qwen.py

update react example

disable hosted inference API

update readme

add resource files

upload pytorch_model.bin

Create README.md

initial commit