Pull Requests · Ascend/MindSpeed-RL - Gitee.com

开源项目 > 人工智能 > 机器学习/深度学习 &&

[resharding] support EP and EP_expend 存在冲突

ascend-cla/yes

zhoubeirong:dev-ep Ascend:dev

!216 2

2025-04-17 15:41

推理异步流水特性支持存在冲突

ci-pipeline-passed

ascend-cla/yes

shishaoyu:dev Ascend:dev

!209 11

2025-04-16 17:56

添加混合奖励和7b R1模型训练示例存在冲突

ci-pipeline-passed

ascend-cla/yes

李尚蔚:dev Ascend:dev

!206 202

2025-04-15 23:57

add format_code_math reward 可合并

ci-pipeline-passed

ascend-cla/yes

白超:415_2 Ascend:master

!204 22

2025-04-15 17:15

添加32b R1模型训练示例可合并

ascend-cla/yes

stat/needs-squash

xiecheng:dev Ascend:dev

!202 38

2025-04-15 15:49

[resharding] support vpp 可合并

ci-pipeline-passed

ascend-cla/yes

stat/needs-squash

zhoubeirong:mindspeed_rl/vpp Ascend:dev

!200 43

2025-04-15 13:37

拒绝采样：增加readme，及qwen2.5-7B、qwen2.5-32B运行配置和脚本可合并

ci-pipeline-passed

ascend-cla/yes

stat/needs-squash

linqihong:rl_2 Ascend:master

!188 48

2025-04-10 20:20

拒绝采样：拒绝采样算法及ut、st 可合并

ci-pipeline-passed

ascend-cla/yes

stat/needs-squash

linqihong:rl_1 Ascend:master

!187 38

2025-04-10 20:19

【轻量级 PR】：update README.md. 可合并

ascend-cla/no

N/A Ascend:master

!177 2

2025-04-07 15:01

特性use_remove_padding 存在冲突

ascend-cla/yes

stat/needs-squash

草稿

Nurxat:master Ascend:master

!174 2

2025-04-03 14:38

1

https://gitee.com/ascend/MindSpeed-RL.git

git@gitee.com:ascend/MindSpeed-RL.git

ascend

MindSpeed-RL

MindSpeed-RL