Skip to content
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

Add Storage Subdirectory #130

Merged
merged 5 commits into from
May 23, 2023
Merged

Add Storage Subdirectory #130

Changes from 2 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
abef4fe
Add stored subdirectory
May 22, 2023
96639e2
Create subdirectory folder from emote
May 22, 2023
2d97976
Implement suggested changes
May 22, 2023
3d41194
Add docs and replace assert with raise
May 22, 2023
130415c
fmt
May 22, 2023
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
17 changes: 15 additions & 2 deletions emote/callbacks.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
import logging
import os
import time

from collections import deque
Expand Down Expand Up @@ -377,13 +378,15 @@ def __init__(
checkpoint_index: int = 0,
optimizers: Optional[List[optim.Optimizer]] = None,
networks: Optional[List[nn.Module]] = None,
storage_subdirectory: str = "checkpoints"
):
super().__init__(cycle=checkpoint_interval)
self._cbs = callbacks
self._path = path
self._checkpoint_index = checkpoint_index
self._opts: List[optim.Optimizer] = optimizers if optimizers else []
self._nets: List[nn.Module] = networks if networks else []
self._storage_subdirectory = storage_subdirectory

def end_cycle(self):
state_dict = {}
Expand All @@ -393,7 +396,11 @@ def end_cycle(self):
state_dict["training_state"] = {
"checkpoint_index": self._checkpoint_index,
}
torch.save(state_dict, f"{self._path}.{self._checkpoint_index}.tar")
name = f"checkpoint_{self._checkpoint_index}.tar"
folder_path = os.path.join(self._path, self._storage_subdirectory)
os.makedirs(folder_path, exist_ok=True)
final_path = os.path.join(folder_path, name)
torch.save(state_dict, final_path)
self._checkpoint_index += 1


Expand Down Expand Up @@ -427,6 +434,7 @@ def __init__(
reset_training_steps: bool = False,
optimizers: Optional[List[optim.Optimizer]] = None,
networks: Optional[List[nn.Module]] = None,
storage_subdirectory: str = "checkpoints"
):
super().__init__()
self._cbs = callbacks
Expand All @@ -435,9 +443,14 @@ def __init__(
self._reset_training_steps = reset_training_steps
self._opts: List[optim.Optimizer] = optimizers if optimizers else []
self._nets: List[nn.Module] = networks if networks else []
self._stored_subdirectory = storage_subdirectory

def begin_training(self):
state_dict: dict = torch.load(f"{self._path}.{self._checkpoint_index}.tar")
name = f"checkpoint_{self._checkpoint_index}.tar"
folder_path = os.path.join(self._path, self._stored_subdirectory)
os.makedirs(folder_path, exist_ok=True)
final_path = os.path.join(folder_path, name)
state_dict: dict = torch.load(final_path)
for cb, state in zip(self._cbs, state_dict["callback_state_dicts"]):
cb.load_state_dict(state)
for net, state in zip(self._nets, state_dict["network_state_dicts"]):
Expand Down