biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix][CI] fix typos (#34934)

Browse Source 
Signed-off-by: 1195343015 <1195343015@qq.com>
Signed-off-by: Jiayi Yan <66017932+1195343015@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Jiayi Yan
2026-03-06 01:05:46 +08:00
committed by GitHub
parent 8c760b6ab6
commit 6a895197fa
98 changed files with 227 additions and 366 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
tests/v1/core/test_scheduler.py
View File

@@ -1116,7 +1116,7 @@ def _step_until_done(
def _step_until_kv_transfer_finished(scheduler: Scheduler, req_ids: list[str]):
"""Cycle requests through a KV transfer cyle."""
"""Cycle requests through a KV transfer cycle."""
# Requests should first transition to WAITING_FOR_REMOTE_KVS
output = scheduler.schedule()
@@ -2714,7 +2714,7 @@ def _assert_right_encoder_inputs(
if expected_total_reqs == 0:
return
# Number of expected enocder inputs should match number of requests
# Number of expected encoder inputs should match number of requests
if expected_encoder_inputs:
assert check_exist and requests is not None # only support expect input exist
assert len(requests) == len(expected_encoder_inputs)
@@ -2964,7 +2964,7 @@ def test_ec_connector_with_partial_cache_hit_multi_round(use_kv_connector):
)
scheduler.update_from_output(output, model_output)
# request1 is finished after outputing 1 token
# request1 is finished after outputting 1 token
# Finish request
scheduler.finish_requests(request1.request_id, RequestStatus.FINISHED_LENGTH_CAPPED)
@@ -3060,14 +3060,14 @@ def test_ec_connector_schedule_multiple_requests(cache_exist, use_kv_connector):
for request in requests:
scheduler.add_request(request)
# Set up to test different encoder cache exsistence scenario after preemption
# Set up to test different encoder cache existence scenario after preemption
# Order of getting encoder cache should be: local cache -> connector-> compute
scheduler.ec_connector.update_state_after_alloc = Mock(
wraps=scheduler.ec_connector.update_state_after_alloc
)
if cache_exist == "local":
# Allocate cache to cache manager manually to mimick
# Allocate cache to cache manager manually to mimic
for req in requests:
scheduler.encoder_cache_manager.allocate(req, 0)
else:
@@ -3384,13 +3384,13 @@ def test_priority_scheduling_ec_connector_preemption_and_resumption(
pooler_output=[],
)
# Finish the requests to make room for the preempted requests to resume
# req_high is finished after outputing 2 tokens
# req_high is finished after outputting 2 tokens
scheduler.update_from_output(output, model_output)
scheduler.finish_requests(
request_high.request_id, RequestStatus.FINISHED_LENGTH_CAPPED
)
# Set up to test different encoder cache exsistence scenario after preemption
# Set up to test different encoder cache existence scenario after preemption
# Order of getting encoder cache should be: local cache -> connector-> compute
# By default, the cache should still exist in local in this test case
if cache_exist != "local":
@@ -3483,7 +3483,7 @@ def test_ec_connector_allocate_encoder_tokens_with_external_load(use_kv_connecto
ec_role="ec_consumer",
)
# Limit the number of availiable slots of EncoderCacheManager
# Limit the number of available slots of EncoderCacheManager
scheduler.encoder_cache_manager = EncoderCacheManager(cache_size=32)
# Create MM request1
@@ -3574,7 +3574,7 @@ def test_ec_connector_allocate_encoder_tokens_with_external_load(use_kv_connecto
)
scheduler.update_from_output(output, model_output)
# request1 is finished after outputing 1 token
# request1 is finished after outputting 1 token
# Finish request
scheduler.finish_requests(request1.request_id, RequestStatus.FINISHED_LENGTH_CAPPED)
assert scheduler.get_num_unfinished_requests() == 1