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
affinity	!3071 [pytorch][refactor]refact affinity feature	10 个月前
ai_framework	!3215 [mindspore][master] add mindspore_register_args	9 个月前
arguments	[pytorch][sh] update dpo in 2.2.0	7 个月前
common	!3232 [pytorh][refactor]refactor tp-2d	9 个月前
context_parallel	[pytorch][doc]update TND doc	3 个月前
dataset	!3164 [pytorch][feature] lora finetune support	9 个月前
dpo	[pytorch][sh] update dpo in 2.2.0	7 个月前
evaluation	!3136 [pytorch][refactor]refactor evaluation feature	9 个月前
finetune	[pytorch][bugfix]add valid_parameter of vpp and lora	5 个月前
functional	!3369 [pytorch][optimize]fix profile step setting and qwen3 scripts	8 个月前
high_availability	!3314 [pytorch][feature]mindio_tft fix the communication_wrapper of barrier	8 个月前
inference	!3128 [pytorch][refactor]refactor inference feature	9 个月前
megatron_basic	fix(pytorch):Ensure no PP/VPP stage contains only empty layers during LoRA fine-tuning	2 个月前
models	!3316 [pytorch][model]add qwen3_next model	8 个月前
moe	!3316 [pytorch][model]add qwen3_next model	8 个月前
optimizer	!3064 [pytorch][refactor]Refactor moe and fusions relevent features	10 个月前
pipeline_parallel	[pytorch][bugfix]fix num-layer-list of ckpt mg2hf	6 个月前
tensor_parallel	!3287 [pytorch][bugfix]fix coc in adapter_v2	8 个月前
tokenizer	!3238 [pytorch][model]support bailing_mini 0day	9 个月前
transformer	[pytorch][bugfix] update ling-mini not support mtp	5 个月前
__init__.py	[pytorch][feature] lora/qlora finetune support gemm	8 个月前