DiffSynth-Studio · Issues· 561 open

minimax-H3 FL2VA train error

MiniMax-H3这个模型经过了CFG-Distillation，它还能正常SFT吗？会不会SFT更多step以后丧失CFG蒸馏的能力？

怎么能使用minimax h3 turbo lora、我测试了最新的 1.2版本、输出视频都是噪声

FLUX LoRA Fine-Tuning: Border Outpainting and Pixel-Level Misalignment Issues in Image Editing

Silently slower attention kernel when training MiniMax-H3 LoRA on B200

oom on minimax h3

train_scale_cfg

Partial download surfaces as "Cannot detect the model type" rather than a download error

LingBot-Video 30B LoRA training OOM on H100

DMD2训练损失震荡

Minimax h3 多卡如何推理？

[Proposal] 探讨对 Diffusion Language Model 的支持

minimax h3我看新增了 adater 训练，ref2av 目前是不支持这种训练方式吗

训练minimax- h3的nf4模型的lora，使用adapter训练时融合lora会报错

Question about optimal hyper parameters for training Qwen-Image-Edit-2509