0
代码
代码
Issues
Pull Requests
讨论
Wiki
分析
0
ascend-robotascend-robotfeat: mxfp8-32x32 quant
65d7fe96创建于 7 天前历史提交
文件最后提交记录最后更新时间
__init__.py
feat(fp8): dual mx quant Co-authored-by: Muu<koimuu@163.com> # message auto-generated for no-merge-commit merge: !3208 merge feature_mxquant into master feat(fp8): dual mx quant Created-by: Muuyo Commit-by: Muu Merged-by: ascend-robot Description: 1. MXFP8支持双轴量化 2. 减少Tensorwise策略量化次数 3. Hif8值域更改 4. 重构Linear层Quant -> Fp8 matmul流程 5. 重命名Float8TensorWithTranspose -> Float8Tensor2D 验证文档: https://wiki.huawei.com/domains/76578/wiki/233229/WIKI2026020210022033 See merge request: Ascend/MindSpeed!32083 个月前
float8_block_tensor.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
float8_tensor.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
float8_tensor_cpu.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
mxfp8_tensor.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
mxfp8_tensor_cpu.py
feat(fp8): quant mm add Co-authored-by: Muu<koimuu@163.com> # message auto-generated for no-merge-commit merge: !3272 merge feature_quant_mm_add into master feat(fp8): quant mm add Created-by: Muuyo Commit-by: Muu Merged-by: ascend-robot Description: 1. mxfp8 策略支持 npu_add_quant_matmul 算子 2. tensorwise 支持mc2策略 3. 修复在backward时取FP8Tensor2D device 取不到的问题 自测报告: https://wiki.huawei.com/domains/76578/wiki/233229/WIKI2026030210253506 See merge request: Ascend/MindSpeed!32722 个月前