Added FLAGS to replace four params and the value can be adjusted for better speedup #9624
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
|
Thanks for your contribution! |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## develop #9624 +/- ##
===========================================
- Coverage 52.45% 52.21% -0.25%
===========================================
Files 721 723 +2
Lines 115768 114326 -1442
===========================================
- Hits 60728 59695 -1033
+ Misses 55040 54631 -409 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
zhink
force-pushed
the
develop
branch
2 times, most recently
from
ca4e474 to
476a243
Compare
zhink
changed the title
zhink
force-pushed
the
develop
branch
2 times, most recently
from
b87d151 to
ffb986b
Compare
zhink
force-pushed
the
develop
branch
3 times, most recently
from
d37765f to
e8a33ff
Compare
carryyu
approved these changes
Dec 30, 2024
yuanlehome
reviewed
Jan 3, 2025
Comment on lines
225
to
237
| inline uint32_t get_decoder_block_shape_q() { | ||
| static const char* decoder_block_shape_q_env = std::getenv("FLAGS_flag_dec_block_shape_q"); | ||
| static const uint32_t decoder_block_shape_q = | ||
| decoder_block_shape_q_env == nullptr ? 16 : std::stoi(std::string(decoder_block_shape_q_env)); | ||
| return decoder_block_shape_q; | ||
| } | ||
|
|
||
| inline uint32_t get_encoder_block_shape_q() { | ||
| static const char* encoder_block_shape_q_env = std::getenv("FLAGS_flag_block_shape_q"); | ||
| static const uint32_t encoder_block_shape_q = | ||
| encoder_block_shape_q_env == nullptr ? 64 : std::stoi(std::string(encoder_block_shape_q_env)); | ||
| return encoder_block_shape_q; | ||
| } |
Collaborator
There was a problem hiding this comment.
flag命名名以及代码位置可以再完善一下，1. 例如FLAGS_flag_，这个小写的flag是不是多余的？建议与变量名保持一致，2. max_partition_size这个flag和这个存放在一起呗
Contributor
|
谢谢，已收到！
|
DrownFish19
reviewed
Jan 3, 2025
llm/docs/predict/best_practices.md
Outdated
|
|
||
| **Append Attention 优化** | ||
|
|
||
| - `FLAGS_cascade_attention_max_partition_size`：Append Attention decoder计算时对cache_kv进行分chunk的chunk大小，batchsize为1时默认值为128,batchsize大于时512。显示设置时不区分batchsize。 |
Collaborator
There was a problem hiding this comment.
默认值根据batchsize设置，batchsize=1时设置为128，batchsize>1时设置为512。显示设置时不再区分batchsize。
llm/docs/predict/best_practices.md
Outdated
| **Append Attention 优化** | ||
|
|
||
| - `FLAGS_cascade_attention_max_partition_size`：Append Attention decoder计算时对cache_kv进行分chunk的chunk大小，batchsize为1时默认值为128,batchsize大于时512。显示设置时不区分batchsize。 | ||
| - `FLAGS_dec_block_shape_q`：Append Attention decoder计算时对q进行分块的分块大小，默认值为16。设置为16即可。 |
Collaborator
There was a problem hiding this comment.
“设置为16即可。”这里如果没有关联指导，就删除吧。下同。
ZHUI
approved these changes
Jan 6, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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 types
Others
PR changes
Others
Description
Added FLAGS to replace max_partition_size、encoder_block_shape_q、decoder_block_shape_q for append_attention OP. And the value can be adjusted for better speedup.