nshtrainer 0.25.0__py3-none-any.whl → 0.26.1__py3-none-any.whl

nshtrainer/_checkpoint/metadata.py

@@ -1,7 +1,6 @@
 import copy
 import datetime
 import logging
-import shutil
 from collections.abc import Callable
 from pathlib import Path
 from typing import TYPE_CHECKING, Any, ClassVar, cast
@@ -11,7 +10,7 @@ import numpy as np
 import torch
 
 from ..util._environment_info import EnvironmentConfig
-from ..util.path import compute_file_checksum, get_relative_path
+from ..util.path import compute_file_checksum, try_symlink_or_copy
 
 if TYPE_CHECKING:
     from ..model import BaseConfig, LightningModuleBase
@@ -142,21 +141,7 @@ def _link_checkpoint_metadata(checkpoint_path: Path, linked_checkpoint_path: Pat
     # Link the metadata files to the new checkpoint
     path = _metadata_path(checkpoint_path)
     linked_path = _metadata_path(linked_checkpoint_path)
-    try:
-        try:
-            # linked_path.symlink_to(path)
-            # We should store the path as a relative path
-            # to the metadata file to avoid issues with
-            # moving the checkpoint directory
-            linked_path.symlink_to(get_relative_path(linked_path, path))
-        except OSError:
-            # on Windows, special permissions are required to create symbolic links as a regular user
-            # fall back to copying the file
-            shutil.copy(path, linked_path)
-    except Exception:
-        log.exception(f"Failed to link {path} to {linked_path}")
-    else:
-        log.debug(f"Linked {path} to {linked_path}")
+    try_symlink_or_copy(path, linked_path)
 
 
 def _sort_ckpts_by_metadata(

nshtrainer/_checkpoint/saver.py

@@ -5,7 +5,7 @@ from pathlib import Path
 
 from lightning.pytorch import Trainer
 
-from ..util.path import get_relative_path
+from ..util.path import try_symlink_or_copy
 from .metadata import _link_checkpoint_metadata, _remove_checkpoint_metadata
 
 log = logging.getLogger(__name__)
@@ -34,13 +34,7 @@ def _link_checkpoint(
         if metadata:
             _remove_checkpoint_metadata(linkpath)
 
-    try:
-        linkpath.symlink_to(get_relative_path(linkpath, filepath))
-    except OSError:
-        # on Windows, special permissions are required to create symbolic links as a regular user
-        # fall back to copying the file
-        shutil.copy(filepath, linkpath)
-
+    try_symlink_or_copy(filepath, linkpath)
     if metadata:
         _link_checkpoint_metadata(filepath, linkpath)
 

nshtrainer/callbacks/gradient_skipping.py

@@ -1,8 +1,8 @@
-import importlib.util
 import logging
 from typing import Any, Literal, Protocol, runtime_checkable
 
 import torch
+import torchmetrics
 from lightning.pytorch import Callback, LightningModule, Trainer
 from torch.optim import Optimizer
 from typing_extensions import override
@@ -20,19 +20,12 @@ class HasGradSkippedSteps(Protocol):
 
 class GradientSkipping(Callback):
     def __init__(self, config: "GradientSkippingConfig"):
-        if importlib.util.find_spec("torchmetrics") is not None:
-            raise ImportError(
-                "To use the GradientSkipping callback, please install torchmetrics: pip install torchmetrics"
-            )
-
         super().__init__()
         self.config = config
 
     @override
     def setup(self, trainer: Trainer, pl_module: LightningModule, stage: str) -> None:
         if not isinstance(pl_module, HasGradSkippedSteps):
-            import torchmetrics  # type: ignore
-
             pl_module.grad_skipped_steps = torchmetrics.SumMetric()
 
     @override

nshtrainer/util/path.py

@@ -1,8 +1,13 @@
 import hashlib
+import logging
 import os
+import platform
+import shutil
 from pathlib import Path
 from typing import TypeAlias
 
+log = logging.getLogger(__name__)
+
 _Path: TypeAlias = str | Path | os.PathLike
 
 
@@ -68,3 +73,32 @@ def compute_file_checksum(file_path: Path) -> str:
         for byte_block in iter(lambda: f.read(4096), b""):
             sha256_hash.update(byte_block)
     return sha256_hash.hexdigest()
+
+
+def try_symlink_or_copy(
+    file_path: Path,
+    link_path: Path,
+    target_is_directory: bool = False,
+    relative: bool = True,
+):
+    """
+    Symlinks on Unix, copies on Windows.
+    """
+
+    symlink_target = get_relative_path(link_path, file_path) if relative else file_path
+    try:
+        if platform.system() == "Windows":
+            if target_is_directory:
+                shutil.copytree(file_path, link_path)
+            else:
+                shutil.copy(file_path, link_path)
+        else:
+            link_path.symlink_to(
+                symlink_target, target_is_directory=target_is_directory
+            )
+    except Exception:
+        log.exception(f"Failed to create symlink or copy {file_path} to {link_path}")
+        return False
+    else:
+        log.debug(f"Created symlink or copied {file_path} to {link_path}")
+        return True

{nshtrainer-0.25.0.dist-info → nshtrainer-0.26.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nshtrainer
-Version: 0.25.0
+Version: 0.26.1
 Summary: 
 Author: Nima Shoghi
 Author-email: nimashoghi@gmail.com
@@ -22,7 +22,7 @@ Requires-Dist: psutil
 Requires-Dist: pytorch-lightning
 Requires-Dist: tensorboard ; extra == "extra"
 Requires-Dist: torch
-Requires-Dist: torchmetrics ; extra == "extra"
+Requires-Dist: torchmetrics
 Requires-Dist: typing-extensions
 Requires-Dist: wandb ; extra == "extra"
 Requires-Dist: wrapt ; extra == "extra"

{nshtrainer-0.25.0.dist-info → nshtrainer-0.26.1.dist-info}/RECORD

@@ -1,8 +1,8 @@
 nshtrainer/__init__.py,sha256=39loiLLXbaGiozEsAn8mPHopxaPsek8JsgR9DD2gxtY,583
 nshtrainer/_callback.py,sha256=A1zLsTy4b_wOYnInLLXGSRdHzT2yNa6mPEql-ozm0u0,1013
 nshtrainer/_checkpoint/loader.py,sha256=5vjg-OFChXJjgiOVv8vnV8nwTscfdDtEdxQRz6uPfDE,14158
-nshtrainer/_checkpoint/metadata.py,sha256=BpxC3VGrgSDHvZpc40A2icjBEDvRNGEsTis9YkFY8Kc,5341
-nshtrainer/_checkpoint/saver.py,sha256=fvRKGI5aeXtsHBOIO4cwGe__wmO-6DiD0-744VASYA4,1500
+nshtrainer/_checkpoint/metadata.py,sha256=hxZwwsUKVbBtt4wjqcKZbObx0PuO-qCdF3BTdnyqaQo,4711
+nshtrainer/_checkpoint/saver.py,sha256=1loCDYDy_Cay37uKs_wvxnkwvr41WMmga85qefct80Q,1271
 nshtrainer/_experimental/__init__.py,sha256=pEXPyI184UuDHvfh4p9Kg9nQZQZI41e4_HvNd4BK-yg,81
 nshtrainer/_hf_hub.py,sha256=0K3uWa8hd2KyGuUYM7OXARcA7vuUiWWGSlP2USysY7o,12066
 nshtrainer/callbacks/__init__.py,sha256=4qocBDzQbLLhhbIEfvbA3SQB_Dy9ZJH7keMwPay-ZS8,2359
@@ -17,7 +17,7 @@ nshtrainer/callbacks/checkpoint/on_exception_checkpoint.py,sha256=ctT88EGT22_t_6
 nshtrainer/callbacks/early_stopping.py,sha256=VWuJz0oN87b6SwBeVc32YNpeJr1wts8K45k8JJJmG9I,4617
 nshtrainer/callbacks/ema.py,sha256=8-WHmKFP3VfnzMviJaIFmVD9xHPqIPmq9NRF5xdu3c8,12131
 nshtrainer/callbacks/finite_checks.py,sha256=gJC_RUr3ais3FJI0uB6wUZnDdE3WRwCix3ppA3PwQXA,2077
-nshtrainer/callbacks/gradient_skipping.py,sha256=pqu5AELx4ctJxR2Y7YSSiGd5oGauVCTZFCEIIS6s88w,3665
+nshtrainer/callbacks/gradient_skipping.py,sha256=EBNkANDnD3BTszWjnG-jwY8FEj-iRqhE3e1x5LQF6M8,3393
 nshtrainer/callbacks/interval.py,sha256=smz5Zl8cN6X6yHKVsMRS2e3SEkzRCP3LvwE1ONvLfaw,8080
 nshtrainer/callbacks/log_epoch.py,sha256=fTa_K_Y8A7g09630cG4YkDE6AzSMPkjb9bpPm4gtqos,1120
 nshtrainer/callbacks/norm_logging.py,sha256=T2psu8mYsw9iahPKT6aUPjkGrZ4TIzm6_UUUmE09GJs,6274
@@ -82,11 +82,11 @@ nshtrainer/trainer/trainer.py,sha256=Zwdcqfmrr7yuonsp4VrNOget8wkaZY9lf-_yeJ94lkk
 nshtrainer/util/_environment_info.py,sha256=gIdq9TJgzGCdcVzZxjHcwYasJ_HmEGVHbvE-KJVVtWs,24187
 nshtrainer/util/_useful_types.py,sha256=dwZokFkIe7M5i2GR3nQ9A1lhGw06DMAFfH5atyquqSA,8000
 nshtrainer/util/environment.py,sha256=AeW_kLl-N70wmb6L_JLz1wRj0kA70xs6RCmc9iUqczE,4159
-nshtrainer/util/path.py,sha256=RUkIOrlj9b8zPPXE3JLhdihBNitJSUWw1whZ33u-2Yk,2005
+nshtrainer/util/path.py,sha256=jAEjF1qp8Aii32L5lWG4UFgVyQAFkHOMYEc_TC2hDx8,2947
 nshtrainer/util/seed.py,sha256=Or2wMPsnQxfnZ2xfBiyMcHFIUt3tGTNeMMyOEanCkqs,280
 nshtrainer/util/slurm.py,sha256=rofIU26z3SdL79SF45tNez6juou1cyDLz07oXEZb9Hg,1566
 nshtrainer/util/typed.py,sha256=NGuDkDzFlc1fAoaXjOFZVbmj0mRFjsQi1E_hPa7Bn5U,128
 nshtrainer/util/typing_utils.py,sha256=8ptjSSLZxlmy4FY6lzzkoGoF5fGNClo8-B_c0XHQaNU,385
-nshtrainer-0.25.0.dist-info/METADATA,sha256=Rqdeh2yp2AhZ_nOHlD47v5YPDrLc2fHN6WGwqJnDv04,935
-nshtrainer-0.25.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-nshtrainer-0.25.0.dist-info/RECORD,,
+nshtrainer-0.26.1.dist-info/METADATA,sha256=tMMpyg1BTKec5d69ziW6XBxDXaI0gSK5tDMPCmj7VCQ,916
+nshtrainer-0.26.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+nshtrainer-0.26.1.dist-info/RECORD,,