mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-09-27 04:46:16 +08:00
[Bug fix] Fix batched token condition (#3565)
This commit is contained in:
@@ -156,12 +156,12 @@ class DPLocalScheduler(LocalScheduler):
|
|||||||
required_total_blocks += required_input_blocks + reserved_output_blocks
|
required_total_blocks += required_input_blocks + reserved_output_blocks
|
||||||
if required_total_blocks > available_blocks:
|
if required_total_blocks > available_blocks:
|
||||||
break
|
break
|
||||||
if current_prefill_tokens > max_num_batched_tokens:
|
|
||||||
break
|
|
||||||
|
|
||||||
requests.append(request.raw)
|
requests.append(request.raw)
|
||||||
self.ids_read_cursor += 1
|
self.ids_read_cursor += 1
|
||||||
start_batch_time = time.time()
|
start_batch_time = time.time()
|
||||||
|
if current_prefill_tokens > max_num_batched_tokens:
|
||||||
|
break
|
||||||
if len(requests) >= batch:
|
if len(requests) >= batch:
|
||||||
break
|
break
|
||||||
if (
|
if (
|
||||||
|
Reference in New Issue
Block a user