423A35C7

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-04-02 02:16:08 +08:00

acac63ef35 [data] fix qwen3vl timestamp (#10338)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-04-01 01:46:00 +08:00

e5e8546493 [misc] fix moe (#10334)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-30 17:06:02 +08:00

97433c53b6 [feat] support LlamaFactory SFT training by HyperParallel FSDP2 backend (#10289)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-27 16:26:00 +08:00

b5afabe3d2 [v1] support ulysses cp for fsdp2 (#10262)

df2e6edb7e [v1] add init on rank0 for fsdp2 (#10264)

d02fcd3588 [ci] add nginx cache config for Ascend NPU CI environment (#10323)

Compare 3 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-27 00:06:01 +08:00

c340aa2a33 [v1] add callbacks (#10255)

1e536733c6 [data] fix mimo-v2 tool call (#10315)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-24 23:06:05 +08:00

97d479fa92 [model] support Qwen3.5 liger kernel (#10313)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-22 05:45:59 +08:00

ffbff33af3 chore: mca workflow compatible with qwen-vl series (#10303)

423A35C7 synced new reference dependabot/npm_and_yarn/demo/frontend/flatted-3.4.2 to 423A35C7/sam2 from mirror 2026-03-20 21:06:02 +08:00

423A35C7 synced commits to dependabot/npm_and_yarn/demo/frontend/flatted-3.4.2 at 423A35C7/sam2 from mirror 2026-03-20 21:06:02 +08:00

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-20 21:06:01 +08:00

833f6027b1 [fix] fit neat_packing & mrope model packing (#10283)

d91d8af89e [data] add SGSC zero-hallucination B2B dataset (NOO-Protocol) (#10284)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-19 02:46:01 +08:00

e67ab9e2f2 fix:MiniCPMVPlugin IndexError in process_messages when training with video (#10276)

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2026-03-19 02:45:58 +08:00

b6a77ad7aa [pytorch3d[ Remove LlffDatasetMapProvider and BlenderDatasetMapProvider

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-18 18:36:04 +08:00

2c4f121817 [fix] handle empty content list in system message (#10291)

487f8b8191 [v1] add qwen3 templates and fix rendering plugin. (#10212)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-18 02:16:01 +08:00

78cad1e332 [fix] unused keys in ray example (#10290)

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2026-03-17 01:46:00 +08:00

52164b8324 Remove Support Ukraine banner from PyTorch3D website

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-17 01:25:58 +08:00

70653026f5 [fix] make position_id_per_seconds configurable for Qwen2OmniPlugin (#10281)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-10 22:06:02 +08:00

246192abd2 [data] correct gpt_oss template format_assistant (#10269)

0258dc14d0 [docker] update npu docker (#10268)

3045adf0ba [fix] fallback to audio_processor when feature_extractor is missing (#10267)

Compare 3 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-10 13:56:00 +08:00

a3d44e3152 [mca] support qwen3.5 (#10265)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-03-09 21:25:59 +08:00

edeb953bc7 [data] convert filter() to list in read_cloud_json to fix broken empty-check (#10260)

d045794387 [docs] fix Python version requirement from 3.10 to >=3.11.0 (#10259)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2026-03-07 04:36:00 +08:00

61cc79aa34 Make _sqrt_positive_part ONNX-exportable