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

remove input_ids from ForwardMeta (#4793)
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
This commit is contained in:
周周周
2025-11-05 11:55:51 +08:00
committed by GitHub
parent 9676cc87d6
commit 876e4a8935
13 changed files with 26 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tests/graph_optimization/test_cuda_graph_spec_decode.py
View File

@@ -42,13 +42,13 @@ class TestCase1SubLayer1(paddle.nn.Layer):
def forward(self, ids_remove_padding, forward_meta: ForwardMeta):
"""Sub layer1 forward pass"""
output = paddle.add(forward_meta.input_ids, forward_meta.input_ids)
output = paddle.add(forward_meta.ids_remove_padding, forward_meta.ids_remove_padding)
return output
def forward_correct(self, ids_remove_padding, forward_meta: ForwardMeta):
"""Sub layer1 Correct forward pass"""
output = paddle.add(forward_meta.input_ids, forward_meta.input_ids)
output = paddle.add(forward_meta.ids_remove_padding, forward_meta.ids_remove_padding)
return output
@@ -70,9 +70,7 @@ class TestModel1(paddle.nn.Layer):
sublayer1_output = self.sublayer1(ids_remove_padding=ids_remove_padding, forward_meta=sub_meta1)
# sublayer2 use cuda graph
sub_meta2 = ForwardMeta(
input_ids=sublayer1_output, ids_remove_padding=sublayer1_output, step_use_cudagraph=True
)
sub_meta2 = ForwardMeta(ids_remove_padding=sublayer1_output, step_use_cudagraph=True)
sublayer2_output = self.sublayer2(ids_remove_padding=sublayer1_output, forward_meta=sub_meta2)
return sublayer2_output
@@ -86,7 +84,7 @@ class TestModel1(paddle.nn.Layer):
)
# sublayer2 not use cuda graph
sub_meta2 = ForwardMeta(input_ids=sublayer1_output, ids_remove_padding=sublayer1_output)
sub_meta2 = ForwardMeta(ids_remove_padding=sublayer1_output)
sublayer2_output = self.sublayer2.forward_correct(ids_remove_padding=sublayer1_output, forward_meta=sub_meta2)
return sublayer2_output
@@ -122,7 +120,7 @@ class TestCUDAGrpahSpecDecode(unittest.TestCase):
# Run Test Case1
test_model1 = TestModel1(fd_config=fd_config)
input_tensor1 = paddle.ones([1, 32768])
forward_meta1 = ForwardMeta(input_ids=input_tensor1, ids_remove_padding=input_tensor1, step_use_cudagraph=True)
forward_meta1 = ForwardMeta(ids_remove_padding=input_tensor1, step_use_cudagraph=True)
# Trigger Capture
_ = test_model1(ids_remove_padding=input_tensor1, forward_meta=forward_meta1)