423A35C7

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-27 01:00:34 +08:00

a1b1931b4a [breaking] migrate from setuptools to uv (#9673)

3c17f2722c [model] Update ernie_vl to adapt new version (#9665)

a882e2d5fc [assets] Add GitHub Copilot instructions for repository (#9675)

Compare 3 commits »

423A35C7 synced new reference hiyouga/misc to 423A35C7/LLaMA-Factory from mirror 2025-12-27 01:00:34 +08:00

423A35C7 synced commits to hiyouga/misc at 423A35C7/LLaMA-Factory from mirror 2025-12-27 01:00:34 +08:00

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-25 08:10:36 +08:00

a754604c11 [misc] fix accelerator (#9661)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-24 07:40:35 +08:00

6a2eafbae3 [feat] Models trained and inferred with Mxfp4 are dequantized by default (#9652)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-23 23:30:36 +08:00

84485406b7 [ci] disable pip cache for ci (#9654)

1c8a42d2f8 [v1&WIP] dataloader init (#9645)

7901b2f32e [model] efficient tuning for gpt-oss (#9354)

Compare 3 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-22 06:40:35 +08:00

1f1f5a7d1b [ci] remove docker cache (#9640)

6ef9854713 [misc] fix cache & pin transformers to 4.57.1 (#9638)

Compare 2 commits »

423A35C7 synced and deleted reference refs/tags/hiyouga/cache at 423A35C7/LLaMA-Factory from mirror 2025-12-22 06:40:35 +08:00

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-21 22:30:35 +08:00

4923f52a28 [model] support MiMo-V2-Flash model (#9637)

423A35C7 synced new reference hiyouga/cache to 423A35C7/LLaMA-Factory from mirror 2025-12-21 22:30:35 +08:00

423A35C7 synced commits to hiyouga/cache at 423A35C7/LLaMA-Factory from mirror 2025-12-21 22:30:35 +08:00

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-20 22:00:36 +08:00

0894b4f37e [misc] lint (#9636)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-20 05:40:34 +08:00

b0d49e137f [misc] Support split eval_dataset when explict set "predict_with_generate" (#9604)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-19 21:30:37 +08:00

ddd7dcc722 [data] Fix the video frame sampling issue #9620 (#9634)

5204cd2bca [misc] add version check for moe (#9633)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-19 05:10:35 +08:00

8c74dca76a [feat] Models trained and inferred with FP8 are dequantized by default (#9627)

e8deda53a1 [example] add Qwen3 series examples (#9624)

a769fb94b9 [feat] support ktransformers for dpo (#9621)

964569751f [kt] refactor ktransformers integration (#9632)

Compare 4 commits »

423A35C7 synced commits to main at 423A35C7/pytorch3d from mirror 2025-12-18 05:10:34 +08:00

f5f6b78e70 Add initial CUDA 13.0 support for pulsar and pycuda modules

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-17 04:10:36 +08:00

9fd4b094d4 [model] support VibeThinker models (#9616)

18c21bce5a [test] add allreduce test on npu (#9619)

Compare 2 commits »

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-15 19:30:36 +08:00

a0179772ab [example] add deepspeed autotp config and example (#9602)

423A35C7 synced commits to main at 423A35C7/LLaMA-Factory from mirror 2025-12-14 19:06:26 +08:00

aeda079014 [v1] model loader (#9613)

423A35C7 synced and deleted reference refs/tags/hiyouga/mo at 423A35C7/LLaMA-Factory from mirror 2025-12-14 19:06:26 +08:00