Repository navigation
release: 1.12.1 - #1660
Merged
Merged
release: 1.12.1#1660
Conversation
30 pull requests since v1.12.0, 22 from contributors: tokenizers back to the reference (qwen36, OLMoE), brio on DeepSeek V4, max_tokens as a ceiling on qwen36 and qwen38, the dense trunk offered to the VRAM placer with a startup probe that withdraws it when the GPU loses, the SSE4.1 tier and the K1b AVX-512 arms, the GGUF reader, and the README with the dashboard as it is. Version in the seven places the banner and site tests check.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Version bump to 1.12.1 in
c/version.py, the five README banners and the site footer, plus the CHANGELOG section: 30 pull requests since v1.12.0, 22 from contributors.Headlines: qwen36 and OLMoE tokenizers identical to HF
tokenizerson the real vocabularies (#1654, #1656); brio on DeepSeek V4 (#1655);max_tokensas a ceiling on qwen36 and qwen38, which fixescoli chatat the default context (#1659); the dense trunk offered to the VRAM placer with a startup probe that withdraws it when the GPU loses, measured on four Tesla M10 (#1657); the SSE4.1 tier (#1239) and the K1b AVX-512 arms (#1606); the GGUF reader (#1425); the README with the dashboard as it is (#1639, #1644).Contract tests green: every README banner says the declared version, the site shows the version and every family. Dev at the base of this branch: eight engines exact and byte-identical in chat on ds,
make checkgreen.After this lands on
dev: the dev to main PR, then the tag and the Release.