| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
[Modify] hetero-pp support encoder/decoder different micro-batch-size & improve qwen2.5omni data balance on global batch Co-authored-by: huangdabiao<huangdabiao1@huawei.com> # message auto-generated for no-merge-commit merge: !1736 merge test_final_1111 into master [Modify] hetero-pp support encoder/decoder different micro-batch-size & improve qwen2.5omni data balance on global batch Created-by: huangdabiao Commit-by: huangdabiao Merged-by: ascend-robot Description: ## Motivation [Modify] hetero-pp support encoder/decoder different micro-batch-size & improve qwen2.5omni data balance on global batch ## Modification [Modify] hetero-pp support encoder/decoder different micro-batch-size & improve qwen2.5omni data balance on global batch ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [ ] The new code needs to comply with the Clean Code specification. - [ ] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [ ] CLA has been signed and all committers have signed the CLA in this PR. - [ ] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1736 | 7 个月前 | |
[Bugfix] fix wan2.1 feature extract Co-authored-by: 林明哲<linmingzhe3@huawei.com> # message auto-generated for no-merge-commit merge: !1720 merge fix1107 into master [Bugfix] fix wan2.1 feature extract Created-by: LinMingZhe Commit-by: 林明哲 Merged-by: ascend-robot Description: ## Motivation fix wan2.1 feature extract ## Modification fix image encoder check ## Checklist **Before PR**: - [ ] The new code needs to comply with the Clean Code specification. - [ ] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [ ] CLA has been signed and all committers have signed the CLA in this PR. - [ ] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1720 | 7 个月前 | |
| 11 个月前 | ||
[Modify] async offload modify Co-authored-by: liyingxuan<liyingxuan3@huawei.com> # message auto-generated for no-merge-commit merge: !1909 merge async_offload into master [Modify] async offload modify Created-by: liyx616 Commit-by: liyingxuan Merged-by: ascend-robot Description: ## Motivation 优化async offload处理逻辑 ## Modification 1. 原来的代码中，stream和event的管理过于混乱，有一些冗余的wait关系，也缺乏一些必要的wait关系，进行重新梳理，梳理之后统一使用event进行管理 2. 修复反向H2D潜在的显存未及时释放问题 3. 添加必要的注释 4. 移除没有用到的函数 计算流和swap流的依赖关系梳理如下：  ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [ ] The new code needs to comply with the Clean Code specification. - [ ] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [ ] CLA has been signed and all committers have signed the CLA in this PR. - [ ] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1909 | 6 个月前 | |
MM disttrain intervl2 适配自动并行搜索 Co-authored-by: gcw_amOUPDs9<fuyuefeng@huawei.com> # message auto-generated for no-merge-commit merge: !1588 merge master into master MM disttrain intervl2 适配自动并行搜索 Created-by: gcw_amOUPDs9 Commit-by: gcw_amOUPDs9 Merged-by: ascend-robot Description: ## Motivation Please describe the motivation of this PR and the goal you want to achieve through this PR. ## Modification Please briefly describe what modification is made in this PR. ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [x] The new code needs to comply with the Clean Code specification. - [x] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [x] CLA has been signed and all committers have signed the CLA in this PR. - [x] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1588 | 8 个月前 | |
[Feature]opensoraplan1.3新增动态DPCP切换功能 Co-authored-by: qusongyun1<qusongyun1@noreply.gitcode.com> # message auto-generated for no-merge-commit merge: !1677 merge dynamicDPCP into master [Feature]opensoraplan1.3新增动态DPCP切换功能 Created-by: qusongyun1 Commit-by: qusongyun1 Merged-by: ascend-robot Description: ## Motivation 当前静态DPCP方案在动态负载下无法充分利用算力，例如在大量短序列和少量长序列的情况下，为了保证不OOM，需要设置较大的CP，然而短序列进行大CP并行会导致性能的下降。本特性新增动态DPCP功能，支持在每轮训练迭代中根据数据特征动态切换DP/CP并行策略。 ## Modification pretrain_sora.py：如果开启了动态DPCP，则优先获取缓存数据 training.py: 在初始化时，新增DPCP并行组的初始化，切换后，将数据在cp组内广播并放入缓存 MindSpeed-MM/mindspeed_mm/utils 中新增dpcp_utils.py文件，所有本特性相关的函数实现均在该文件中 ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [x] The new code needs to comply with the Clean Code specification. - [x] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [x] CLA has been signed and all committers have signed the CLA in this PR. - [x] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1677 | 7 个月前 | |
| 1 年前 | ||
[Bugfix] Fallback to older Hetero-Parallel Data Comm Co-authored-by: mazhuang<mazhuang21@huawei.com> # message auto-generated for no-merge-commit merge: !1706 merge new-comm-group-revert into master [Bugfix] Fallback to older Hetero-Parallel Data Comm Created-by: mazhuang1234 Commit-by: mazhuang Merged-by: ascend-robot Description: ## Motivation 新版通信优化与Hetero-Parallel 流水线并行冲突，需要回调到老版Hetero数据流通信 ## Modification Please briefly describe what modification is made in this PR. ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [ ] The new code needs to comply with the Clean Code specification. - [ ] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [ ] CLA has been signed and all committers have signed the CLA in this PR. - [ ] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1706 | 7 个月前 | |
| 1 年前 | ||
| 9 个月前 | ||
[mindspore] support qwen3vl finetune for megatron backend Co-authored-by: ffmh<fengminghao2@huawei.com> # message auto-generated for no-merge-commit merge: !1785 merge qwen3vl_megatron into master [mindspore] support qwen3vl finetune for megatron backend Created-by: ffmh Commit-by: ffmh Merged-by: ascend-robot Description: ## Motivation Please describe the motivation of this PR and the goal you want to achieve through this PR. 适配 qwen3vl 8B/30B megatron后端 ## Modification Please briefly describe what modification is made in this PR. 添加 layer_spec修改、text模块transformer block/layer添加deepstack相关适配 ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [ ] The new code needs to comply with the Clean Code specification. - [ ] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [ ] CLA has been signed and all committers have signed the CLA in this PR. - [ ] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1785 | 7 个月前 | |
[Feature]Add split&gather function for ViT RingAttention, bugfix for qwen3vl Co-authored-by: zs-derrick1<1434012475@qq.com> # message auto-generated for no-merge-commit merge: !1824 merge ring into master [Feature]Add split&gather function for ViT RingAttention, bugfix for qwen3vl Created-by: zs-derrick1 Commit-by: zs-derrick1 Merged-by: ascend-robot Description: ## Motivation Add split&gather function for ViT RingAttention, bugfix for qwen3vl ## Modification Add split&gather function for ViT RingAttention Bugfix for qwen3vl ## Self-test (Optional) If modifications to this PR may cause/fix function/accuracy/performance DTSs/issues, a self-inspection record needs to be attached. ## BC-breaking (Optional) If there are compatibility issues, such as dependencies on cann/torch_npu versions, they need to be explained in the PR. ## Checklist **Before PR**: - [x] The new code needs to comply with the Clean Code specification. - [x] The PR content is self-checked, and the expression can be clear and the writing standardized **After PR**: - [x] CLA has been signed and all committers have signed the CLA in this PR. - [x] The ci-pipeline is passed, Code Check is passed. See merge request: Ascend/MindSpeed-MM!1824 | 7 个月前 |
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
| 7 个月前 | ||
| 7 个月前 | ||
| 11 个月前 | ||
| 6 个月前 | ||
| 8 个月前 | ||
| 7 个月前 | ||
| 1 年前 | ||
| 7 个月前 | ||
| 1 年前 | ||
| 9 个月前 | ||
| 7 个月前 | ||
| 7 个月前 |