-
Notifications
You must be signed in to change notification settings - Fork 3.4k
Pull requests: JustVugg/colibri
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
hyper_connections: parallelise the mHC mix and expand, bit-identical (+3.9%)
#1521
opened Sep 14, 2026 by
rcklomp
Loading…
fix(omp): size the OpenMP team in the four engines that never did
#1517
opened Sep 14, 2026 by
LiRunGuo
Loading…
11 tasks done
fix(k3_repack): a stdout that cannot encode U+2014 kills the resume
#1516
opened Sep 14, 2026 by
kevin9327
Loading…
fix(coli): coli chat hung once an engine stderr line did not fit the locale
#1515
opened Sep 14, 2026 by
kevin9327
Loading…
fix(glm53): refuse an over-long serve prompt with CONTEXT_EXCEEDED
#1514
opened Sep 14, 2026 by
kevin9327
Loading…
vulkan: reserve the expert tier's RAM on integrated GPUs (#653 for the VK path)
#1513
opened Sep 14, 2026 by
huppiflupp
Loading…
tests: make the GPU test suite build under HIP on gfx1151
#1511
opened Sep 14, 2026 by
huppiflupp
Loading…
feat(api): opt-in prompt-injected tool translation for OLMoE and Qwen3.6
#1497
opened Sep 14, 2026 by
basil-k-aji-dev
Loading…
docs: link the GLM-5.3 int4 container from the model section (#1403)
#1474
opened Sep 13, 2026 by
HoneyTyagii
Loading…
feat(glm53): add device-resident HIP backend for MI350P
#1448
opened Sep 12, 2026 by
azharlabs
Loading…
Add GGUF general reader and converter into OlMoE safetensors files
#1425
opened Sep 10, 2026 by
zameo94
Contributor
Loading…
1 of 5 tasks
serve: continue a trailing
assistant turn instead of answering in a new one
#1402
opened Sep 8, 2026 by
enitimeago
Loading…
5 tasks done
perf: COALESCE=1 multi-expert coalesced pread for contiguous shard runs
#1398
opened Sep 8, 2026 by
KyleSanderson
•
Draft
5 tasks done
perf: IDOT_TEAM=1 fuses three OpenMP regions into one team per expert…
#1397
opened Sep 8, 2026 by
KyleSanderson
Loading…
5 tasks done
Opt-in exact verify mode for speculative batches (COLI_EXACT_VERIFY=1)
#1395
opened Sep 7, 2026 by
rybruscoe
Loading…
test(tools): gate the logprob gap check on the engine preamble; add the native-MTP witness
#1357
opened Sep 5, 2026 by
monotophic
Contributor
Loading…
feat(engine): replace the ablation scoring mode with a checked, digest-bound one
#1356
opened Sep 5, 2026 by
monotophic
Contributor
Loading…
feat(tools): offline ablation-evidence checker and an evidence-bound eval harness
#1355
opened Sep 5, 2026 by
monotophic
Contributor
Loading…
OpenAI API Instrumentation (e.g. seed, logprobs, echo, an array prompt)
#1353
opened Sep 5, 2026 by
monotophic
Contributor
Loading…
feat(qwen36): load MLX-affine dense weights and norms from a qpack container
#1343
opened Sep 4, 2026 by
Avicennasis
Contributor
Loading…
qwen36: Vulkan expert tier, and staged device-local uploads for cards without Resizable BAR
#1338
opened Sep 4, 2026 by
crichalchemist
Contributor
Loading…
feat(qwen36): name a checkpoint by its geometry, and a converter that refuses what it cannot place
#1326
opened Sep 3, 2026 by
kreuzzelg
Contributor
Loading…
feat(qwen36): routed experts from a qpack container through bounded Metal slots
#1323
opened Sep 2, 2026 by
Avicennasis
Contributor
Loading…
glm53: size the expert cache around the model, not around MemAvailable
#1321
opened Sep 2, 2026 by
ronaldcklomp
Contributor
Loading…
Previous Next
ProTip!
Type g p on any issue or pull request to go back to the pull request listing page.