Skip to content
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.

Sign up for GitHub

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

Jump to bottom

ggml : generalize quantize_fns for simpler FP16 handling #286

Closed
ggerganov opened this issue Jun 25, 2023 · 1 comment
Closed

ggml : generalize quantize_fns for simpler FP16 handling #286

ggerganov opened this issue Jun 25, 2023 · 1 comment
Labels
good first issue Good for newcomers refactoring Refactoring

Comments

@ggerganov
Copy link
Owner

This task is described well in ggerganov/llama.cpp#1237

The WIP implementation in that PR might be a bit outdated by now, so one can either attempt to update it or implement it from scratch on top of the current code base.
@ggerganov ggerganov added good first issue Good for newcomers refactoring Refactoring labels Jun 25, 2023
goerch added a commit to goerch/ggml that referenced this issue Jun 28, 2023
@goerch
Fix ggerganov#286
58ed3ad
@goerch goerch mentioned this issue Jun 28, 2023
Closed
goerch added a commit to goerch/ggml that referenced this issue Jul 2, 2023
@goerch
Merge branch 'master' into fix-ggerganov#286
f863199
@ggerganov ggerganov mentioned this issue Jul 5, 2023
Merged
@ggerganov
Copy link
Owner Author

ggerganov commented Jul 5, 2023 •
edited
Loading

@goerch I just synced the unit tests from llama.cpp as you proposed in #317

Will close the issue as completed now.
Maybe in the future we can more similar simplifications for other ops that have quantization branches (e.g. ggml_cpy())

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
good first issue Good for newcomers refactoring Refactoring
Projects
ggml : roadmap
Status: Done
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

[WIP]: Fix #286 goerch/ggml
1 participant
@ggerganov