apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

[TSP] Support qwen3 moe tsp + cudagraph (#4871)
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 
* support qwen3_moe tsp mode

* fix

* fix

* update

* update

* update

* fix

* support external_rmsnorm

* update

* fix
This commit is contained in:
Yuanle Liu
2025-11-10 23:37:51 +08:00
committed by GitHub
parent fb2eb403ab
commit 3dc0ffa46d
28 changed files with 173 additions and 273 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastdeploy/model_executor/models/qwen2_5_vl/qwen2_5_vl.py
+1 -18
View File
@@ -124,9 +124,7 @@ class Qwen2_5_VLModel(nn.Layer):
residual,
)
hidden_states = hidden_states + residual
out = self.norm(hidden_states)
out = self.norm(hidden_states, residual)[0]
return out
@@ -262,21 +260,6 @@ class Qwen2_5_VLForConditionalGeneration(ModelForCasualLM):
return logits
def empty_input_forward(self):
"""
empty_input_forward
"""
fake_hidden_states = paddle.empty(
shape=[0, self.fd_config.model_config.hidden_size],
dtype=paddle.get_default_dtype(),
)
for i in range(
self.fd_config.model_config.moe_layer_start_index,
self.fd_config.model_config.num_hidden_layers,
):
self.ernie.layers[i].mlp.text_fused_moe(fake_hidden_states)
self.ernie.layers[i].mlp.image_fused_moe(fake_hidden_states)
def get_input_embeddings(
self,
ids_remove_padding: paddle.Tensor,