-
Notifications
You must be signed in to change notification settings - Fork 73
Pull requests: InfiniTensor/InfiniLM
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
feat: Trim cached memory before DeepSeek V2 lm_head on large prefill
#489
opened Jul 9, 2026 by
baominghelly
Contributor
Loading…
1 of 49 tasks
refactor: reduce parameter passing
#487
opened Jul 8, 2026 by
wooway777
Collaborator
Loading…
49 tasks
issue/477 - feat: support partial cache hits for multimodal visual embeddings
#484
opened Jul 7, 2026 by
ma-hang
Contributor
Loading…
1 of 49 tasks
[codex] Fix Kunlun Qwen2.5 greedy inference
#476
opened Jul 6, 2026 by
zhangyue207
Collaborator
•
Draft
Add out-of-tree model plugin interface
#441
opened Jun 17, 2026 by
whjthu
Contributor
Loading…
1 of 48 tasks
chore: expose cache options in test_infer
#438
opened Jun 16, 2026 by
wooway777
Collaborator
Loading…
1 of 48 tasks
feat(fm9g): integrate fused FFN op into csrc decoder layer
#409
opened Jun 3, 2026 by
ZhouBencheng
Loading…
48 tasks
build: add an cmake build option
#391
opened May 20, 2026 by
wooway777
Collaborator
Loading…
31 of 48 tasks
feat: support chunkprefill and prefill cuda graph
#371
opened May 12, 2026 by
Simon12345777
Loading…
48 tasks
refactor: use processor in infer backup
#368
opened May 12, 2026 by
wooway777
Collaborator
Loading…
29 of 48 tasks
refactor: inline InfiniCore into InfiniLM, switch xmake -> CMake
#324
opened Apr 25, 2026 by
zhangyue207
Collaborator
•
Draft
6 tasks
issue/296 - feat: add Worker and ModelRunner for PD disaggregation
#304
opened Apr 15, 2026 by
spike-zhu
Collaborator
Loading…
enable FA and fix total_kv_lengths check in infer_engine.
#256
opened Mar 6, 2026 by
gongchensu
Collaborator
Loading…
Issue/224：add warmup before InfiniLM generation，and use muDNN silu_and_mul to replace elementwise swiglu in moore gpu
#225
opened Feb 11, 2026 by
spike-zhu
Collaborator
Loading…
Previous Next
ProTip!
Add no:assignee to see everything that’s not assigned.