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

[Feature] [PD] add simple router and refine splitwise deployment (#4709)

Browse Source 
* add simple router and refine splitwise deployment

* fix
This commit is contained in:
Juncai
2025-11-06 14:56:02 +08:00
committed by GitHub
parent 831266da7a
commit 08ca0f6aea
39 changed files with 2397 additions and 171 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
benchmarks/benchmark_dataset.py
View File

@@ -46,7 +46,7 @@ class SampleRequest:
prompt_len: int
expected_output_len: int
response_format: Optional[dict] = None
class BenchmarkDataset(ABC):
"""BenchmarkDataset"""
@@ -299,7 +299,7 @@ class EBChatDataset(BenchmarkDataset):
prompt = entry["messages"][-1].get("content", "")
history_QA = entry.get("messages", [])
response_format = entry.get("response_format")
new_output_len = int(entry.get("max_tokens", 12288))
new_output_len = int(entry.get("max_tokens", output_len if output_len else 12288))
if enable_multimodal_chat:
prompt = self.apply_multimodal_chat_transformation(prompt, None)
@@ -311,7 +311,7 @@ class EBChatDataset(BenchmarkDataset):
prompt_len=0,
history_QA=history_QA,
expected_output_len=new_output_len,
response_format=response_format
response_format=response_format,
)
)
cnt += 1