apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Executor]CUDAGraph support Speculate Decode (#3769)
Some checks failed
CE Compile Job / ce_job_pre_check (push) Has been cancelled
Details
CE Compile Job / print_ce_job_pre_check_outputs (push) Has been cancelled
Details
CE Compile Job / FD-Clone-Linux (push) Has been cancelled
Details
CE Compile Job / Show Code Archive Output (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8090 (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8689 (push) Has been cancelled
Details
CE Compile Job / CE_UPLOAD (push) Has been cancelled
Details
Deploy GitHub Pages / deploy (push) Has been cancelled
Details

Browse Source 
* success run ngram

* Revert "[Code Simplification] remove cum_offsets (#3410)"

This reverts commit 32b39620bc.

* success run ngram5 tp4 42bs

* success run ngram5 tp4 42bs

* mtp draft commit

* add decorator for target model

* enable draft model in cudagraph v0.5

* revert revrt cum_offset

* enable target model in cudagraph v0.9 And clean debug code

* Revert "success run ngram"

This reverts commit 8351e83993.

* add reverted code

* enable target model in cudagraph v0.9

* solve comment

* fix bid < 0

* Enable Target Model Padding And Draft Model in cudagraph

* solve problem

* delete rebuild padding debug note

* fast compile

* Add capture list for mtp

* success run 256 tp1 mtp

* Enable Lite TP2 Bsz256

* realy enable tp2 bsz 256

* fix problem

* Solve problem for Draft model in cudagraph

* Solve comment

* replace emptytensor as zeros

* Solve comments

* Revert "fast compile"

This reverts commit 834639a7ff.

* fix bug

* fix merge bug

* fix typo

* fix bug

---------

Co-authored-by: lizexu <2694294196@qq.com>
Co-authored-by: littledgg <1658565283@qq.com>
Co-authored-by: zeroRains <linjunlu@zerorains.top>
Co-authored-by: gongshaotian <gstain5555@outlook.com>
This commit is contained in:
RAM
2025-10-09 21:18:29 +08:00
committed by GitHub
parent 7b1689f437
commit aa27b03bc0
19 changed files with 250 additions and 139 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
fastdeploy/model_executor/graph_optimization/graph_optimization_backend.py
View File

@@ -115,7 +115,7 @@ class GraphOptBackend:
self.runnable = runnable
self.fd_config = fd_config
self.max_captre_batch = fd_config.graph_opt_config.cudagraph_capture_sizes[0]
self.max_captre_size = fd_config.graph_opt_config.cudagraph_capture_sizes[0]
if self.fd_config.graph_opt_config.graph_opt_level > 0:
# 1. Prepare cuda graph input buffers (contain output of subgraphs)
@@ -138,9 +138,9 @@ class GraphOptBackend:
)
assert kwargs["forward_meta"].ids_remove_padding is not None
batch_size = kwargs["forward_meta"].ids_remove_padding.shape[0]
real_shape = kwargs["forward_meta"].ids_remove_padding.shape[0]
if (not kwargs["forward_meta"].step_use_cudagraph) or (batch_size > self.max_captre_batch):
if (not kwargs["forward_meta"].step_use_cudagraph) or (real_shape > self.max_captre_size):
return self.runnable(**kwargs)
else:
return self.cudagraph_piecewise_backend.__call__(**kwargs)