2.3.0

ascend-robot[pytorch][feature]Decouple the 'variable-seq-lengths' and 'no-pad-to-seq-lengths' parameters

33fc8b89创建于 2025年10月30日历史提交

文件	最后提交记录	最后更新时间
ckpt_convert_hunyuanLarge_hf2mcore.sh	!2828 [pytorch][sh]Add CUDA_DEVICE_MAX_CONNECTIONS Merge pull request !2828 from yanzhixiao/add-environment-variable	1 年前
data_convert_hunyuanLarge_instruction.sh	!2324 【HunyuanLargeMoE】shell files Merge pull request !2324 from zhoubeirong/0218-shell	1 年前
evaluate_hunyuanLarge_389b_A3_ptd.sh	!3162 [pytorch][feature]deprecate the --moe-intermediate-size parameter. Merge pull request !3162 from yanzhixiao/deprecate-args	10 个月前
generate_hunyuanLarge_389b_A3_ptd.sh	!3162 [pytorch][feature]deprecate the --moe-intermediate-size parameter. Merge pull request !3162 from yanzhixiao/deprecate-args	10 个月前
pretrain_hunyuanLarge_389b_A3_ptd.sh	[pytorch][sh] restore the --moe-permute-fusion parameter that was removed in Megatron r0.12.1 Co-authored-by: yanzhixiao<yanzhixiao@h-partners.com> # message auto-generated for no-merge-commit merge: !3550 merge add-permute-fusion into master [pytorch][sh] restore the --moe-permute-fusion parameter that was removed in Megatron r0.12.1 Created-by: yanzhixiao23 Commit-by: yanzhixiao Merged-by: ascend-robot Description: [pytorch][sh] restore the --moe-permute-fusion parameter that was removed in Megatron r0.12.1 See merge request: Ascend/MindSpeed-LLM!3550	8 个月前
tune_hunyuanLarge_389b_A3_ptd.sh	[pytorch][feature]Decouple the 'variable-seq-lengths' and 'no-pad-to-seq-lengths' parameters Co-authored-by: mhh001<mahonghao1@huawei.com> # message auto-generated for no-merge-commit merge: !3579 merge master_pad into master [pytorch][feature]Decouple the 'variable-seq-lengths' and 'no-pad-to-seq-lengths' parameters Created-by: mhh111 Commit-by: mhh001 Merged-by: ascend-robot Description: [pytorch][feature]Decouple the 'variable-seq-lengths' and 'no-pad-to-seq-lengths' parameters See merge request: Ascend/MindSpeed-LLM!3579	8 个月前