-
Notifications
You must be signed in to change notification settings - Fork 1.4k
Pull requests: jundot/omlx
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
fix: change VLM MTP draft block_size default from 4 to 3
#1909
opened Jun 17, 2026 by
imi4u36d
Contributor
Loading…
fix(thinking): stop budget-forced close from leaking a second </think> (#1864)
#1905
opened Jun 17, 2026 by
richgoodson
Contributor
Loading…
chore(deps-dev): bump xgrammar from 0.2.0 to 0.2.2
dependencies
Pull requests that update a dependency file
python
Pull requests that update python code
#1872
opened Jun 14, 2026 by
dependabot
Bot
Loading…
chore(deps-dev): bump apache-tvm-ffi from 0.1.11 to 0.1.12
dependencies
Pull requests that update a dependency file
python
Pull requests that update python code
#1871
opened Jun 14, 2026 by
dependabot
Bot
Loading…
Native MTP is enabled for this model but the converted weights are missing the mtp.* tensors
#1853
opened Jun 12, 2026 by
RoelVideler
Loading…
feat(profiles): expose profiles as API-visible models (web + macOS UI)
#1838
opened Jun 12, 2026 by
pablomoralesm
Loading…
feat: native video_url support for Qwen3-VL-family VLMs (torch-free)
#1836
opened Jun 12, 2026 by
mdc2122
Loading…
fix(serve): make startup failures visible under launchd (#1814)
#1827
opened Jun 11, 2026 by
richgoodson
Contributor
Loading…
feat(thinking): add guided soft thinking budgets
#1812
opened Jun 11, 2026 by
efortin
Contributor
Loading…
fix(vlm): fix Qwen3OmniMoeProcessor loading
#1801
opened Jun 10, 2026 by
monroewilliams
Contributor
Loading…
fix: add swapaxes support to _TrackedTensor for VLM quantization (#1706)
#1787
opened Jun 10, 2026 by
dodams258
Contributor
Loading…
fix: prefix-cache thinking outputs when the template preserves them (#1003)
#1786
opened Jun 10, 2026 by
popfido
Contributor
Loading…
2 of 3 tasks
fix: honor per-model SpecPrefill settings on every generation path
#1782
opened Jun 10, 2026 by
popfido
Contributor
Loading…
2 of 3 tasks
feat: expose DFlash copyspec_mode (prompt-lookup drafting) as a per-model setting
#1781
opened Jun 10, 2026 by
popfido
Contributor
Loading…
3 of 4 tasks
fix(settings): reject dflash_enabled + specprefill_enabled (#1233)
#1760
opened Jun 9, 2026 by
popfido
Contributor
Loading…
2 tasks done
feat: persist benchmark telemetry with CLI history and GUI viewer
#1738
opened Jun 8, 2026 by
llostinthesauce
Loading…
fix(eval): honor per-model temperature in accuracy benchmark
#1722
opened Jun 7, 2026 by
JimStenstrom
Contributor
Loading…
fix(model_discovery): classify Qwen2ForCausalLM embedding models
#1720
opened Jun 7, 2026 by
JimStenstrom
Contributor
Loading…
chore(deps): update numpy requirement from <2.4,>=1.24.0 to >=1.24.0,<2.5
dependencies
Pull requests that update a dependency file
python
Pull requests that update python code
#1711
opened Jun 7, 2026 by
dependabot
Bot
Loading…
chore(deps): bump actions/checkout from 6.0.2 to 6.0.3
dependencies
Pull requests that update a dependency file
github_actions
Pull requests that update GitHub Actions code
#1710
opened Jun 7, 2026 by
dependabot
Bot
Loading…
feat: CLI model alias management and Bearer token auth for admin API
#1705
opened Jun 6, 2026 by
maxness
Loading…
Previous Next
ProTip!
What’s not been updated in a month: updated:<2026-05-17.