2.2.0

ascend-robotfix(pytorch):Ensure no PP/VPP stage contains only empty layers during LoRA fine-tuning

337813b0创建于 4月2日历史提交

文件	最后提交记录	最后更新时间
core	[pytorch][bugfix] fixes the accuracy bug for the TND.	4 个月前
features_manager	fix(pytorch):Ensure no PP/VPP stage contains only empty layers during LoRA fine-tuning	2 个月前
inference	[pytorch][bugfix]fix about cleancode	8 个月前
legacy	[pytorch][bugfix]fix about cleancode	8 个月前
mindspore	[mindspore][bugfix]adapt qwen3_235b for mindspore	6 个月前
tasks	[pytorch][bugfix]fix deepseek3 tnd bug in mbs > 1	3 个月前
training	fix(pytorch):Ensure no PP/VPP stage contains only empty layers during LoRA fine-tuning	2 个月前
__init__.py	!3163 [pytorch][feature]switch megatron_adaptor to v2	9 个月前