apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-28 18:51:58 +08:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

remove debug code

Browse Source
This commit is contained in:
Deleter-D
2025-09-25 20:58:08 +08:00
parent a46aa06194
commit 03ab5530d8
7 changed files with 6 additions and 108 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
fastdeploy/output/token_processor.py
View File

@@ -158,7 +158,6 @@ class TokenProcessor:
get_output_ep,
get_output_topk,
speculate_get_output,
speculate_get_output_topk,
)
rank_id = self.cfg.parallel_config.local_data_parallel_id
@@ -166,24 +165,9 @@ class TokenProcessor:
try:
is_blocking = True
if self.speculative_decoding:
if self.use_logprobs:
speculate_get_output_topk(
self.output_tokens,
self.output_scores,
self.output_ranks,
K,
rank_id,
is_blocking,
)
print(f"[TokenProcessor] output_tokens: {self.output_tokens}")
print(f"[TokenProcessor] output_scores: {self.output_scores}")
print(f"[TokenProcessor] output_ranks: {self.output_ranks}")
if self.output_tokens[0, 0] == -2:
continue
else:
speculate_get_output(self.output_tokens, rank_id, is_blocking, False)
if self.output_tokens[0] == -2:
continue
speculate_get_output(self.output_tokens, rank_id, is_blocking, False)
if self.output_tokens[0] == -2:
continue
else:
if self.use_logprobs: