Command line switch to use F16 for memory_k and memory_v (refactor of #154) #294
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Green-Sky
force-pushed
the
f16_memory_cli
branch
2 times, most recently
from
7b92973 to
82fcf96
Compare
Green-Sky
force-pushed
the
f16_memory_cli
branch
from
82fcf96 to
31edd6f
Compare
ggerganov
approved these changes
Mar 19, 2023
Deadsg
pushed a commit
to Deadsg/llama.cpp
that referenced
this pull request
Dec 19, 2023
….22.0 Bump uvicorn from 0.21.1 to 0.22.0
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
made the changes requested by @ggerganov in #154 . fixes #146
With this change, you can half the
llama_model_load: memory_size = 512.00 MB->memory_size = 256.00 MB(ctx512 7B q4_0)
A non empirical comparison does not seem to degrade the quality of the prediction. but that might not mean anything. (waiting on #270)