Skip to content

adjust default values #3921

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
lvhan028 merged 2 commits into from
Sep 4, 2025
Merged

adjust default values #3921

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
8a40e27
adjust default values
lvhan028 Sep 1, 2025
9f73edd
update
lvhan028 Sep 3, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 7 additions & 6 deletions lmdeploy/cli/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -526,7 +526,7 @@ def max_log_len(parser):
return parser.add_argument('--max-log-len',
type=int,
default=None,
help='Max number of prompt characters or prompt tokens being'
help='Max number of prompt characters or prompt tokens being '
'printed in log. Default: Unlimited')

@staticmethod
Expand All @@ -552,8 +552,9 @@ def communicator(parser):
return parser.add_argument('--communicator',
type=str,
default='nccl',
choices=['nccl', 'native'],
help='Communication backend for multi-GPU inference')
choices=['nccl', 'native', 'cuda-ipc'],
help='Communication backend for multi-GPU inference. The "native" option is '
'deprecated and serves as an alias for "cuda-ipc"')

@staticmethod
def enable_microbatch(parser):
Expand Down Expand Up @@ -581,9 +582,9 @@ def role(parser):
type=str,
default='Hybrid',
choices=['Hybrid', 'Prefill', 'Decode'],
help='Hybrid for Non-Disaggregated Engine;'
'Prefill for Disaggregated Prefill Engine;'
'Decode for Disaggregated Decode Engine;')
help='Hybrid for Non-Disaggregated Engine; '
'Prefill for Disaggregated Prefill Engine; '
'Decode for Disaggregated Decode Engine')

@staticmethod
def migration_backend(parser):
Expand Down
2 changes: 1 addition & 1 deletion lmdeploy/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -337,7 +337,7 @@ def get_max_batch_size(device_type: str):
"""
assert device_type in ['cuda', 'ascend', 'maca', 'camb']
if device_type == 'cuda':
max_batch_size_map = {'a100': 256, 'a800': 256, 'h100': 512, 'h800': 512}
max_batch_size_map = {'a100': 384, 'a800': 384, 'h100': 1024, 'h800': 1024, 'l20y': 1024, 'h200': 1024}
import torch
device_name = torch.cuda.get_device_name(0).lower()
for name, size in max_batch_size_map.items():
Expand Down