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

[Feature] Support for request-level speculative decoding metrics monitoring. (#5518)

Browse Source 
* support spec metrics monitor per request

* fix bug

* remove debug log

* fix ut bugs
This commit is contained in:
GoldPancake
2025-12-12 12:22:18 +08:00
committed by GitHub
parent 3c1f7b85a4
commit 909059c60a
11 changed files with 154 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/output/test_process_batch_output.py
+3 -6
View File
@@ -142,13 +142,10 @@ class TestTokenProcessorProcessBatchOutput(unittest.TestCase):
processor.num_accepted_tokens = 0
processor.num_emitted_tokens = 0
processor.max_num_emitted_tokens = 0
processor.num_rest_requests_per_head = [
0,
] * MAX_DRAFT_TOKENS
processor.num_accept_requests_per_head = [
0,
] * MAX_DRAFT_TOKENS
processor.speculative_stats_step = 0
processor.total_step_per_request = {}
processor.accept_token_num_per_head_per_request = {}
processor.accept_token_num_per_head = [0] * MAX_DRAFT_TOKENS
# processor._recycle_resources = Mock()