423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-12-16 20:00:36 +08:00
Code Issues Packages Projects Releases Wiki Activity

update tests

Browse Source
This commit is contained in:
hiyouga
2024-11-02 12:21:41 +08:00
parent 0a55e60693
commit 93d3b8f43f
23 changed files with 53 additions and 62 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/llamafactory/cli.py
View File

@@ -86,19 +86,19 @@ def main():
elif command == Command.EXPORT:
export_model()
elif command == Command.TRAIN:
force_torchrun = os.environ.get("FORCE_TORCHRUN", "0").lower() in ["true", "1"]
force_torchrun = os.getenv("FORCE_TORCHRUN", "0").lower() in ["true", "1"]
if force_torchrun or get_device_count() > 1:
master_addr = os.environ.get("MASTER_ADDR", "127.0.0.1")
master_port = os.environ.get("MASTER_PORT", str(random.randint(20001, 29999)))
master_addr = os.getenv("MASTER_ADDR", "127.0.0.1")
master_port = os.getenv("MASTER_PORT", str(random.randint(20001, 29999)))
logger.info(f"Initializing distributed tasks at: {master_addr}:{master_port}")
process = subprocess.run(
(
"torchrun --nnodes {nnodes} --node_rank {node_rank} --nproc_per_node {nproc_per_node} "
"--master_addr {master_addr} --master_port {master_port} {file_name} {args}"
).format(
nnodes=os.environ.get("NNODES", "1"),
node_rank=os.environ.get("RANK", "0"),
nproc_per_node=os.environ.get("NPROC_PER_NODE", str(get_device_count())),
nnodes=os.getenv("NNODES", "1"),
node_rank=os.getenv("NODE_RANK", "0"),
nproc_per_node=os.getenv("NPROC_PER_NODE", str(get_device_count())),
master_addr=master_addr,
master_port=master_port,
file_name=launcher.__file__,