0
代码
代码
Issues
Pull Requests
讨论
Wiki
分析
0
ascend-robotascend-robotfix(quant): only hif8 add dst_type_max args
0dbb7a63创建于 3 天前历史提交
文件最后提交记录最后更新时间
recipes
fix(quant): only hif8 add dst_type_max args Co-authored-by: Muu<koimuu@163.com> # message auto-generated for no-merge-commit merge: !3514 merge fix-hif8-tensorwise into master fix(quant): only hif8 add dst_type_max args Created-by: Muuyo Commit-by: Muu Merged-by: ascend-robot Description: fix(quant): only hif8 add dst_type_max args See merge request: Ascend/MindSpeed!35143 天前
tensor
feat: mxfp8-32x32 quant Co-authored-by: kyle_zhangchi<zhangchi158@huawei.com> # message auto-generated for no-merge-commit merge: !3471 merge feat_mxfp8-32x32 into master feat: mxfp8-32x32 quant Created-by: kyle_zhangchi Commit-by: kyle_zhangchi Merged-by: ascend-robot Description: ## What this PR does / why we need it? 在Megatron框架下新增mxfp8-32x32量化算子，降低权重显存占用 ## Does this PR introduce *any* user-facing change? --fp8-recipe新增mxfp8-32x32选项 https://gitcode.com/Ascend/MindSpeed/commit/e065cbca6873bfc02661d088b07d90224333e87d?ref=feat_mxfp8-32x32&prId=3471 ## How was this patch tested? 验证文档 https://wiki.huawei.com/domains/170864/wiki/367830/WIKI2026051111046509 See merge request: Ascend/MindSpeed!34717 天前
__init__.py
feat: mxfp8-32x32 quant Co-authored-by: kyle_zhangchi<zhangchi158@huawei.com> # message auto-generated for no-merge-commit merge: !3471 merge feat_mxfp8-32x32 into master feat: mxfp8-32x32 quant Created-by: kyle_zhangchi Commit-by: kyle_zhangchi Merged-by: ascend-robot Description: ## What this PR does / why we need it? 在Megatron框架下新增mxfp8-32x32量化算子，降低权重显存占用 ## Does this PR introduce *any* user-facing change? --fp8-recipe新增mxfp8-32x32选项 https://gitcode.com/Ascend/MindSpeed/commit/e065cbca6873bfc02661d088b07d90224333e87d?ref=feat_mxfp8-32x32&prId=3471 ## How was this patch tested? 验证文档 https://wiki.huawei.com/domains/170864/wiki/367830/WIKI2026051111046509 See merge request: Ascend/MindSpeed!34717 天前
checkpoint.py
fix: FP8 recompute error in verl Co-authored-by: quancs001<quancs@qq.com> # message auto-generated for no-merge-commit merge: !3492 merge fix-fp8-recompute into master fix: FP8 recompute error in verl Created-by: quancs001 Commit-by: quancs001 Merged-by: ascend-robot Description: 修复Verl FP8训练时，条件判断错误导致重计算函数的参数数量不对的问题 See merge request: Ascend/MindSpeed!34928 天前
constants.py
feat: mxfp8-32x32 quant Co-authored-by: kyle_zhangchi<zhangchi158@huawei.com> # message auto-generated for no-merge-commit merge: !3471 merge feat_mxfp8-32x32 into master feat: mxfp8-32x32 quant Created-by: kyle_zhangchi Commit-by: kyle_zhangchi Merged-by: ascend-robot Description: ## What this PR does / why we need it? 在Megatron框架下新增mxfp8-32x32量化算子，降低权重显存占用 ## Does this PR introduce *any* user-facing change? --fp8-recipe新增mxfp8-32x32选项 https://gitcode.com/Ascend/MindSpeed/commit/e065cbca6873bfc02661d088b07d90224333e87d?ref=feat_mxfp8-32x32&prId=3471 ## How was this patch tested? 验证文档 https://wiki.huawei.com/domains/170864/wiki/367830/WIKI2026051111046509 See merge request: Ascend/MindSpeed!34717 天前
fp8.py
feature(fp8): te checkpoint Co-authored-by: Muu<koimuu@163.com> # message auto-generated for no-merge-commit merge: !3162 merge feature_checkpoint into master feature(fp8): te checkpoint Created-by: Muuyo Commit-by: Muu Merged-by: ascend-robot Description: 1. 引入 te checkpoint消除重计算中冗余的量化操作 2. refactor(blockwise): 删除128*128的blockwise策略, 保留1 * 128|128 * 128策略替换 3. perf(hif8): 删除多余的cast 4. fix(delayed): 修复delayed算法 5. refactor(recipe 2x): 重构blockwise和mxfp8策略数据存取, 简化后续算子适配 6. 消除字符串字面量, 采用枚举替代 验证报告: https://wiki.huawei.com/domains/76578/wiki/233229/WIKI202601139775970 See merge request: Ascend/MindSpeed!31624 个月前
metadata.py
feature(fp8): te checkpoint Co-authored-by: Muu<koimuu@163.com> # message auto-generated for no-merge-commit merge: !3162 merge feature_checkpoint into master feature(fp8): te checkpoint Created-by: Muuyo Commit-by: Muu Merged-by: ascend-robot Description: 1. 引入 te checkpoint消除重计算中冗余的量化操作 2. refactor(blockwise): 删除128*128的blockwise策略, 保留1 * 128|128 * 128策略替换 3. perf(hif8): 删除多余的cast 4. fix(delayed): 修复delayed算法 5. refactor(recipe 2x): 重构blockwise和mxfp8策略数据存取, 简化后续算子适配 6. 消除字符串字面量, 采用枚举替代 验证报告: https://wiki.huawei.com/domains/76578/wiki/233229/WIKI202601139775970 See merge request: Ascend/MindSpeed!31624 个月前
padding.py
Support TransformerEngine Co-authored-by: MingzhenWang<wangmingzhen4@huawei.com> Co-authored-by: Muu<koimuu@163.com> Co-authored-by: x30061065<xuyuanhui3@h-partners.com> Co-authored-by: 耿瑞良<gengruiliang@huawei.com> # message auto-generated for no-merge-commit merge: !2947 merge lingqu_master into master Support TransformerEngine Created-by: mingzhenwang Commit-by: mingzhenwang;Muu;MingzhenWang;x30061065;耿瑞良 Merged-by: ascend-robot Description: 1. 支持TELinear层 2. 支持FP8计算，quantmatmul/gmm 3. 支持多种数据类型FP8/HiF8 4. 支持多种量化策略delayed/tensorwise/blockwise/mxfp8 5. TELinear层支持通算融合 See merge request: Ascend/MindSpeed!29476 个月前
reuse.py
feat: mxfp8-32x32 quant Co-authored-by: kyle_zhangchi<zhangchi158@huawei.com> # message auto-generated for no-merge-commit merge: !3471 merge feat_mxfp8-32x32 into master feat: mxfp8-32x32 quant Created-by: kyle_zhangchi Commit-by: kyle_zhangchi Merged-by: ascend-robot Description: ## What this PR does / why we need it? 在Megatron框架下新增mxfp8-32x32量化算子，降低权重显存占用 ## Does this PR introduce *any* user-facing change? --fp8-recipe新增mxfp8-32x32选项 https://gitcode.com/Ascend/MindSpeed/commit/e065cbca6873bfc02661d088b07d90224333e87d?ref=feat_mxfp8-32x32&prId=3471 ## How was this patch tested? 验证文档 https://wiki.huawei.com/domains/170864/wiki/367830/WIKI2026051111046509 See merge request: Ascend/MindSpeed!34717 天前
scale_data.py
FP8 delayed fix Co-authored-by: mingzhenwang<wangmingzhen4@huawei.com> # message auto-generated for no-merge-commit merge: !3242 merge delayed into master FP8 delayed fix Created-by: mingzhenwang Commit-by: mingzhenwang Merged-by: ascend-robot Description: 1.修复delayed模式下流同步未等待amax更新完出现除零nan现象 hif8 delayed验证结果 FP8 因算子饱和模式阻塞，暂时无法验证 https://wiki.huawei.com/domains/76578/wiki/233229/WIKI2026020610085315 See merge request: Ascend/MindSpeed!32423 个月前
state_manager.py
feat: fp8 reuse quant w with te_gmm_mode compatible Co-authored-by: Jia_Austin<dengjia6@huawei.com> # message auto-generated for no-merge-commit merge: !3371 merge fp8_reuse_perf_v2 into master feat: fp8 reuse quant w with te_gmm_mode compatible Created-by: Jia_Austin Commit-by: Jia_Austin Merged-by: ascend-robot Description: What this PR does / why we need it? feat: fp8 reuse quant w with te_gmm_mode compatible; perf/fix: fp8 reuse quant w with te_gmm_mode perf Does this PR introduce any user-facing change? Please describe whether the PR will result in any user-facing usage changes. If there is related documentation, please specify its path. How was this patch tested? Please explain how to verify the correctness and effectiveness of this feature, as well as its usage constraints and limitations. See merge request: Ascend/MindSpeed!33711 个月前