support fp8 per token quant for deepep low latency two stage #76863
+299
−170
Conversation
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
|
你的PR提交成功，感谢你对开源项目的贡献! |
Sunny-bot1
changed the title
carryyu
reviewed
Dec 12, 2025
| sizeof(int4) + (kUseFP8 ? (kHidden + kNumScales * sizeof(float)) | ||
| : (kHidden * sizeof(nv_bfloat16))); | ||
| sizeof(int4) + (kUseFP8 | ||
| ? (kHidden + (kNumScales + 3) / 4 * 4 * sizeof(float)) |
Contributor
There was a problem hiding this comment.
最好不要硬编码成和4对齐，constexpr ALIGN_ELEMS=xxx，类似这种再对齐。
Contributor
Author
There was a problem hiding this comment.
最好不要硬编码成和4对齐，constexpr ALIGN_ELEMS=xxx，类似这种再对齐。
done
Contributor
There was a problem hiding this comment.
这个是由于一次load 16字节，4个float， 所以需要4个float 对齐，不要用环境变量
| auto num_tokens = static_cast<int>(x.size(0)), | ||
| hidden = static_cast<int>(x.size(1)); | ||
| auto num_scales = hidden / 128, num_topk = static_cast<int>(topk_idx.size(1)); | ||
| auto num_scales = num_per_channel == -1 ? 1 : hidden / 128, |
Contributor
There was a problem hiding this comment.
如果引入了num_per_channel，这里是不是改成hidden / num_per_channel
Contributor
Author
There was a problem hiding this comment.
如果引入了num_per_channel，这里是不是改成hidden / num_per_channel
这样的话，per-token的num_per_channel需要传hidden_size进来，参数会有点繁琐
yangjianfengo1
approved these changes
Dec 16, 2025
Contributor
|
LGTM |
lshpku
approved these changes
Dec 16, 2025
gongweibao
approved these changes
Dec 17, 2025
5 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
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Outdated suggestions cannot be applied.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.
PR Category
Performance Optimization
PR Types
New features
Description
为DeepEP ll two stage适配激活per-token量化