Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / dachengli1/longchat issues and pull requests

#44 - Fix BibTeX

Pull Request - State: open - Opened by sylchw 5 months ago

#42 - How to prepare the training data

Issue - State: open - Opened by ycsun1972 about 1 year ago - 2 comments

#41 - Added additional testcases

Pull Request - State: closed - Opened by AnzeXie about 1 year ago

#40 - dummy conversations seem to be short

Issue - State: closed - Opened by Arist12 about 1 year ago - 2 comments

#39 - Inference is very slow on long text input

Issue - State: open - Opened by Colafei0406 about 1 year ago - 1 comment

#38 - support vllm & lightllm

Pull Request - State: closed - Opened by andy-yang-1 about 1 year ago - 3 comments

#37 - license

Issue - State: closed - Opened by yzhang123 about 1 year ago - 1 comment

#36 - Release LongChat v1.5

Pull Request - State: closed - Opened by DachengLi1 about 1 year ago

#34 - Update Anthropic Client

Issue - State: open - Opened by krrishdholakia over 1 year ago - 2 comments

#33 - flash attention rename

Issue - State: closed - Opened by Arist12 over 1 year ago - 1 comment

#32 - Do you support Llama-2-13b model data？

Issue - State: open - Opened by brewswang over 1 year ago

#31 - train ValueError

Issue - State: open - Opened by brewswang over 1 year ago

#30 - flash_attn installed, but got ImportErrorImportError

Issue - State: closed - Opened by brewswang over 1 year ago - 4 comments

#29 - Output token limit

Issue - State: open - Opened by MoppyDu97 over 1 year ago

#28 - OOM issue

Issue - State: closed - Opened by WeixuanXiong over 1 year ago - 4 comments

#27 - Why the use of flash attention in the inference stage will lead to slower？

Issue - State: closed - Opened by xyfZzz over 1 year ago - 2 comments

#26 - Maybe a bug in the preprocess?

Issue - State: open - Opened by Richar-Du over 1 year ago - 3 comments

#25 - About the print message

Issue - State: open - Opened by lucasjinreal over 1 year ago - 2 comments

#23 - Longchat inference configuration

Issue - State: open - Opened by SeekWrldTea over 1 year ago - 1 comment

#22 - Use attention_mask and use_cache for xformers

Pull Request - State: closed - Opened by fahadh4ilyas over 1 year ago

#21 - Xformers Monkey Patch Compatibility

Issue - State: open - Opened by fahadh4ilyas over 1 year ago - 1 comment

#20 - Support for other model like Baichuan

Issue - State: open - Opened by lucasjinreal over 1 year ago

#19 - About the learning rate

Issue - State: open - Opened by lucasjinreal over 1 year ago - 1 comment

#18 - Will it support qlora?

Issue - State: open - Opened by lw3259111 over 1 year ago - 1 comment

#17 - The purpose of pretrain script?

Issue - State: closed - Opened by fahadh4ilyas over 1 year ago - 2 comments

#16 - why not reuse fschat code?

Issue - State: closed - Opened by lucasjinreal over 1 year ago - 8 comments

#15 - Monkey Patch Xformers use `past_key_value` but `use_cache` can't be `True`?

Issue - State: closed - Opened by fahadh4ilyas over 1 year ago - 9 comments

#14 - longchat-13b-16k chat not work

Issue - State: open - Opened by ahkimkoo over 1 year ago - 9 comments

#13 - Implementation for issue 6&7

Pull Request - State: closed - Opened by AnzeXie over 1 year ago
Labels: enhancement

#12 - Web GUI for longchat

Issue - State: open - Opened by VVNMA over 1 year ago - 3 comments

#11 - Multi-node training?

Issue - State: closed - Opened by XueFuzhao over 1 year ago - 1 comment

#10 - Load the model for inference?

Issue - State: closed - Opened by fahadh4ilyas over 1 year ago - 4 comments

#9 - OutOfMemoryError: CUDA out of memory.

Issue - State: open - Opened by brewswang over 1 year ago - 5 comments

#8 - Can inference be run on consumer hardware?

Issue - State: open - Opened by GrahamboJangles over 1 year ago - 8 comments

#7 - Add scripts on querying closed sourced models

Issue - State: closed - Opened by DachengLi1 over 1 year ago - 1 comment

#6 - Add scripts to generate more testcases

Issue - State: closed - Opened by DachengLi1 over 1 year ago - 1 comment
Labels: enhancement

#5 - How was the 18k dataset prepared?

Issue - State: closed - Opened by musabgultekin over 1 year ago - 3 comments
Labels: good first issue

#4 - How to use 3090 to train 16k model?

Issue - State: open - Opened by aresa7796 over 1 year ago - 7 comments
Labels: good first issue

#3 - Add support for flash attention with use_cache

Issue - State: open - Opened by DachengLi1 over 1 year ago - 1 comment
Labels: enhancement

#2 - unsupervised pre-training on the model

Issue - State: closed - Opened by wqn1 over 1 year ago - 5 comments
Labels: good first issue

#1 - [DO NOT MERGE] Add ZeroScrolls & Scrolls Evals

Pull Request - State: open - Opened by Ying1123 over 1 year ago