Search results

Noisy Reward Guidance

Status: Open.

#34 In KlingAIResearch/VideoAlign;

· Eeelyyy opened on Jan 20, 2026

如何微调VideoReward的checkpoint，而不是从零开始微调Qwen2-VL？

Status: Open.

#33 In KlingAIResearch/VideoAlign;

· serendipity800 opened on Nov 11, 2025

<code>max_length</code> in <code>TrainingConfig</code> doesn't seem to be used？

Status: Open.

#32 In KlingAIResearch/VideoAlign;

· MraDonkey opened on Sep 24, 2025

Multi-GPU Inference on H20 Causes Rotary Position Embedding Dimension Mismatch

Status: Open.

#26 In KlingAIResearch/VideoAlign;

· pikanzz opened on Aug 13, 2025

Code for DPO-finetuning?

Status: Open.

#4 In KlingAIResearch/VideoAlign;

· KasenYoung opened on Mar 15, 2025

will you release the train datasets?

Status: Open.

#1 In KlingAIResearch/VideoAlign;

· wsxwd opened on Feb 13, 2025