Files
AstrAI/astrai/parallel/utils.py
T
ViperEkura bcaa2d1ae0 fix: FSDP unwrap_model collective op and None guard
- unshard() and full_tensor() are collective ops, all ranks must participate
- Old code returned None on non-rank-0 before calling unshard, causing deadlock
- Fix: all ranks unshard/full_tensor, only rank-0 keeps the result
- Move create_ref_model to parallel/utils.py, accept executor+model directly
- Guard create_ref_model and sync_old_model against None on non-rank-0
2026-07-29 23:41:10 +08:00

35 lines
1.0 KiB
Python

"""Utility functions for parallel training."""
from typing import TYPE_CHECKING, Callable, Dict, Optional
import torch
import torch.nn as nn
if TYPE_CHECKING:
from astrai.parallel.executor import BaseExecutor
def create_ref_model(
model_fn: Callable[[], nn.Module],
executor: Optional["BaseExecutor"] = None,
model: Optional[nn.Module] = None,
state_dict: Optional[Dict[str, torch.Tensor]] = None,
device: Optional[str] = None,
) -> Optional[nn.Module]:
"""Create a frozen reference model from executor or state dict.
On non-rank-0, returns None (executor.unwrap_model returns None).
"""
if state_dict is None and executor is not None and model is not None:
state_dict = executor.unwrap_model(model)
if state_dict is None:
return None
ref_model = model_fn()
ref_model.load_state_dict(state_dict)
ref_model.requires_grad_(False)
ref_model.eval()
if device is not None:
ref_model = ref_model.to(device=device)
return ref_model