423A35C7

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-12 17:02:18 +08:00

675ce8cc7f [algo] add ASFT (#10174)

ab073f4c13 [v1] add LoRA/Freeze support and merge workflow (#10157)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-11 00:12:28 +08:00

184304b5b4 [model] add liger kernel support for Qwen3-Next (#10176)

d3ebd5678d [model] support GLM-OCR SFT (#10183)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-09 23:42:19 +08:00

1d5e8ebcd0 [v1] init commit for v1 docs (#10145)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-09 15:32:18 +08:00

ea644d04ec [model] support GLM-4.7-Flash SFT (#10173)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-05 05:22:17 +08:00

92fa3df4c4 [trainer] add dpo/kto fsdp fsdp2 support (#10127)

8bedfafa4e [model] support MiniCPM-o-4.5 (#10163)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-03 20:32:16 +08:00

1a02717fa8 [assets] update readme (#10159)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-03 12:22:15 +08:00

e7cb145f5d [logging] Fix race condition in LoggerHandler during multi-GPU training (#10156)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-03 04:12:17 +08:00

b53d7037c2 [model] support youtu-vl model (#10152)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-02-02 20:02:18 +08:00

bf04ca6af8 [deps] adapt to transformers v5 (#10147)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-28 17:22:19 +08:00

762b480131 [feature] support using ray.remote to start distributed training. (#10109)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-27 17:00:38 +08:00

9640f79ae5 [fix] add visual.pos_embed to Qwen3-VL visual model keys (#10139)

7ef19eea00 [v0] Fix reward model training safetensors saving (#10137)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-26 00:00:36 +08:00

f9f11dcb97 [v1] support training with fsdp2 (#9773)

641bfdd482 chore: Update outdated GitHub Actions versions (#10123)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-20 21:10:40 +08:00

e70651ac58 [feat] support all_exhausted_without_replacement in datasets.interleave_datasets (#10112)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-19 20:40:37 +08:00

db2f794f7b [misc] update mcore related docker and mca supported models (#10114)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-17 11:30:34 +08:00

44eadbda1c [v1] fix kernel moe patch (#9867)

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2026-01-15 02:40:34 +08:00

cbcae096a0 Add atol=1e-4 to assertClose calls in test_inverse for Translate

5b1cce56bc Fix for T251460511 ("Your diff, D90498281, broke one test")

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-14 01:50:34 +08:00

9829ae0a77 [ci] using mp to run kernel test (#9754)

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2026-01-13 01:40:34 +08:00

0c3b204375 apply Black 25.11.0 style in fbcode (70/92)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-12 09:00:35 +08:00

958b9c3468 [v1] add sft (#9752)

4d3621e3d3 [model] fixed&added Hunyuan models (#9750)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2026-01-12 00:50:36 +08:00

a296723697 [v1] upgrade batching (#9751)