Docker fixes and other improvements#19
Open
PolynomialDivision wants to merge 13 commits into
Open
Conversation
Ensures the newest commit is always built.
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
from
32ba027 to
2a12950
Compare
Concurrent requests serialize through a mutex. Previously the second request would block forever and silently return the original input on failure. Now it waits up to 120 seconds and returns HTTP 503 if the server is still busy.
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
from
2a12950 to
2f4a01b
Compare
Contributor
Author
|
Runtested on my unraid, except the new gemma 4 model. |
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
from
2f4a01b to
e5d8e78
Compare
…cpp-2 Drops the git submodule in favor of llama-cpp-2 v0.1.146 from crates.io (utilityai/llama-cpp-rs). The fork was needed when upstream lacked Gemma 3 chat template support. Use token_to_piece instead of the deprecated token_to_bytes / Special API.
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
from
e5d8e78 to
03859c6
Compare
Contributor
Author
|
gemma4 works now |
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
2 times, most recently
from
d58d1ff to
adf2500
Compare
Gemma 4's template hides <start_of_turn> in macros so detection fails. Fall back to hardcoded Gemma format when apply_chat_template returns an error.
PolynomialDivision
force-pushed
the
docker-fixes-and-other-improvements
branch
from
adf2500 to
3de18cf
Compare
Contributor
Author
|
I tried to gorup this PR into several smaller PRs. |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Some small docker fixes, and gpu busy handling.