Issues 共 72
[Bug] Chat Template Not Applied Correctly to Recent Models
#551 · Dogacel · 2026-04-30
[Bug] Segfault encountered in training draft model
#539 · jjtql · 2026-04-17
[Bug] There is a gap between the acceptance rates of training and inference
#533 · xinanjiao · 2026-04-14
[Bug] Dflash
#501 · Jim2016713 · 2026-03-16
[Feature] will support qwen3.5-mtp training and qwen3.5-eagle3 training?
#497 · aidenchn · 2026-03-09
Support for target and draft models on different GPUs
#487 · chenxiaodong2002 · 2026-03-03
[Question] Eagle3 training seems to freeze target embeddings but train a new draft lm_head
#478 · Hannibal046 · 2026-03-02
[Bug] DFlash block mask currently relies on Flex Attention only — request for SDPA / generic attention backend support
#474 · baifanxxx · 2026-02-26
[Bug] Fix incorrect padding logic in both backends and wrong target_hidden construction in DFlash training
#471 · hukongyi · 2026-02-13
[Bug] dflash cross block data leak
#470 · TechxGenus · 2026-02-13