8000 Fix file sizes of LLM's by Smartappli · Pull Request #945 · serge-chat/serge · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Fix file sizes of LLM's #945

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 14 commits into from
Dec 6, 2023
Merged

Fix file sizes of LLM's #945

merged 14 commits into from
Dec 6, 2023

Conversation

Smartappli
Copy link
Contributor
@Smartappli Smartappli commented Dec 4, 2023

Reorganization of models.json file

Fixes #910

Reorganization of models.json file
Add support for SlimOrca 13B
@Smartappli Smartappli marked this pull request as draft December 4, 2023 21:57
Falcon 7B file size bugfix
@Smartappli
Copy link
Contributor Author
Smartappli commented Dec 4, 2023

Models Filesize control

Alfred-40B KO
CodeLlama-7B OK
CodeLlama-7B-Instruct OK
CodeLlama-13B-Instruct OK
Falcon-7B KO
Falcon-7B Instruct KO
Falcon-40B KO
Falcon-40B Instruct KO
LLaMA2-7B-Chat OK
LLaMA2-13B-Chat KO
LLaMA2-70B-Chat OK
Med42-70B OK
Meditron-7B OK
Meditron-70B OK
Mistral-7B OK
Mistral-7B-Instruct OK
Mistral-7B-OpenOrca OK
Neural-Chat-7B-v3-2 OK
Notus-7B-v1 OK
Open_LLaMA-3B-v2 OK
Open_LLaMA-7B-v2 OK
Open_LLaMA-13B-V2 OK
OpenChat-3_5-7B KO
Orca-2-7B OK
Orca-2-13B KO
PsyMedRP-v1-13B KO
PsyMedRP-v1-20B OK
SlimOrca KO
Starling-LM-7B-Alpha OK
Vicuna-7B OK
Vicuna-13B KO
Zephyr-7B-Alpha OK
Zephyr-7B-Beta OK

Falcon 7B filesize updated
Falcon 7B fix
Vicuna-13B fix
SlimOrca bugfix
@gaby
Copy link
Member
gaby commented Dec 5, 2023

I'm not sure what are you changing? @Smartappli

@Smartappli
Copy link
Contributor Author
Smartappli commented Dec 5, 2023

I'm not sure what are you changing? @Smartappli

I'm in the process of testing each module one after the other to make sure it downloads

@gaby
Copy link
Member
gaby commented Dec 5, 2023

@Smartappli You can do that with the huggingface api. I just need to update the sizes. Some of the them are off

@gaby
Copy link
Member
gaby commented Dec 5, 2023

Try this:

from huggingface_hub import list_files_info

print(list(list_files_info(repo_id="TheBloke/CodeLlama-7B-GGUF",
        repo_type="model",
        revision="main",
    ))
)

It should print the size

@Smartappli
Copy link
Contributor Author

You can do that with the huggingface api. I just need to update the sizes. Some of the them are off

we need to update the size of the ones mentioned KO :)

Model filesize updated
Thx Gaby for the snippet :)
@Smartappli Smartappli marked this pull request as ready for review December 5, 2023 01:49
@Smartappli
Copy link
Contributor Author

Wow, they all work:)

@gaby gaby changed the title Update models.json Fix file sizes of LLM's Dec 5, 2023
@gaby
Copy link
Member
gaby commented Dec 5, 2023

@Smartappli Thanks for the help, will review later today. 💪

@Smartappli
Copy link
Contributor Author
Smartappli commented Dec 5, 2023

image

You may need to remove a few decimal places ;)

@gaby
Copy link
Member
gaby commented Dec 5, 2023

Will do, thanks

Add support for Llama-2-70B-OASST
Add support for Llama-2-70B-OASST
tiny correctoin
Add Support for LLaMA 2 7B, 13B, 70B
Smartappli and others added 2 commits December 5, 2023 16:32
Add support for LLaMA 2 7B, 13B, 70B
@gaby gaby merged commit e5be21c into serge-chat:main Dec 6, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

🐛 [Bug]: CodeLlama-7B instruct download stuck
2 participants
0