423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-22 13:42:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

add DISABLE_TORCHRUN option

Browse Source 
Former-commit-id: 45d8be8f93188e6e9d1e0841c2736dce0c1ef7d2
This commit is contained in:
hiyouga 2024-06-06 23:44:58 +08:00
parent 77b70664e5
commit fd2c64315b
1 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/llamafactory/cli.py
View File

@ -72,7 +72,12 @@ def main():
elif command == Command.EXPORT:
export_model()
elif command == Command.TRAIN:
if get_device_count() > 0:
disable_torchrun = os.environ.get("DISABLE_TORCHRUN", "0").lower() in ["true", "1"]
if disable_torchrun and get_device_count() > 1:
logger.warning("`torchrun` cannot be disabled when device count > 1.")
disable_torchrun = False
if (not disable_torchrun) and (get_device_count() > 0):
master_addr = os.environ.get("MASTER_ADDR", "127.0.0.1")
master_port = os.environ.get("MASTER_PORT", str(random.randint(20001, 29999)))
logger.info("Initializing distributed tasks at: {}:{}".format(master_addr, master_port))