测试基建:✨ 补齐 FP8 与 TE MoE 对齐测试能力#682
Merged
Merged
Conversation
|
Thanks for your contribution! |
cangtianhuang
enabled auto-merge (squash)
July 16, 2026 12:51
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🔧 主要变更
1. 统一比较异常分类
将
accuracy、accuracy_stable、paddle_device_vs_*、paddle_cinn_vs_dygraph等 tester 中的重复异常处理改为调用统一的report_compare_error,并修复[torch_assert_OOM]的 OOM 分类,提升报告日志的一致性。2. 修复 numpy 与 MoE 边界输入生成
在 embedding、cross entropy、take along axis 等依赖随机索引上界的配置生成中增加 0-size 兜底;MoE routemap 生成同时处理
topk == 0与topk > num_experts,避免 numpy 采样异常并保持tokens_per_expert与 routemap 一致。3. 扩展 FP8 与 TE MoE 转换规则
为
fp8_quant_blockwise增加 Transformer Engine 参考实现和环境变量切换能力,补齐 1x128 / 128x128、scale transpose、input transpose、ue8m0 packed scale 等路径;同时完善 MoE unzipped scale scatter 和 fused FP8 quant 的 ue8m0 输出打包逻辑。📁 改动文件