nshtrainer 0.26.1__py3-none-any.whl → 0.26.2__py3-none-any.whl

nshtrainer/_hf_hub.py

@@ -188,27 +188,19 @@ class HFHubCallback(NTCallbackBase):
 
         self.config = config
 
-        self._repo_id = None
+        self._repo_id: str | None = None
         self._checksum_to_path_in_repo: dict[str, Path] = {}
 
     @override
     def setup(self, trainer, pl_module, stage):
-        from .trainer.trainer import Trainer
-
-        if not isinstance(trainer, Trainer):
-            raise ValueError(
-                f"HFHubCallback requires a `nshtrainer.Trainer` instance, got {type(trainer)}."
-            )
-
         root_config = cast("BaseConfig", pl_module.hparams)
+        self._repo_id = _repo_name(self.api, root_config)
+
+        if not self.config or not trainer.is_global_zero:
+            return
 
         # Create the repository, if it doesn't exist
-        self._repo_id = self.api.create_repo(
-            repo_id=_repo_name(self.api, root_config),
-            repo_type="model",
-            private=self.config.auto_create.private,
-            exist_ok=True,
-        )
+        self._create_repo_if_not_exists()
 
         # Upload the config and code
         self._save_config(root_config)
@@ -216,17 +208,22 @@ class HFHubCallback(NTCallbackBase):
 
     @override
     def on_checkpoint_saved(self, ckpt_path, metadata_path, trainer, pl_module):
-        root_config = cast("BaseConfig", pl_module.hparams)
-
         # If HF Hub is enabled, then we upload
-        if self.config and trainer.is_global_zero:
-            with self._with_error_handling("save checkpoints"):
-                self._save_checkpoint(
-                    _Upload.from_local_path(ckpt_path, root_config),
-                    _Upload.from_local_path(metadata_path, root_config)
-                    if metadata_path is not None
-                    else None,
-                )
+        if (
+            not self.config
+            or not self.config.save_checkpoints
+            or not trainer.is_global_zero
+        ):
+            return
+
+        with self._with_error_handling("save checkpoints"):
+            root_config = cast("BaseConfig", pl_module.hparams)
+            self._save_checkpoint(
+                _Upload.from_local_path(ckpt_path, root_config),
+                _Upload.from_local_path(metadata_path, root_config)
+                if metadata_path is not None
+                else None,
+            )
 
     @cached_property
     def api(self):
@@ -241,6 +238,33 @@ class HFHubCallback(NTCallbackBase):
             raise ValueError("Repository id has not been initialized.")
         return self._repo_id
 
+    def _create_repo_if_not_exists(self):
+        if not self.config or not self.config.auto_create:
+            return
+
+        # Create the repository, if it doesn't exist
+        with self._with_error_handling("create repository"):
+            from huggingface_hub.utils import RepositoryNotFoundError
+
+            try:
+                # Check if the repository exists
+                self.api.repo_info(repo_id=self.repo_id, repo_type="model")
+                log.info(f"Repository '{self.repo_id}' already exists.")
+            except RepositoryNotFoundError:
+                # Repository doesn't exist, so create it
+                try:
+                    self.api.create_repo(
+                        repo_id=self.repo_id,
+                        repo_type="model",
+                        private=self.config.auto_create.private,
+                        exist_ok=True,
+                    )
+                    log.info(f"Created new repository '{self.repo_id}'.")
+                except Exception:
+                    log.exception(f"Failed to create repository '{self.repo_id}'")
+            except Exception:
+                log.exception(f"Error checking repository '{self.repo_id}'")
+
     def _save_config(self, root_config: "BaseConfig"):
         with self._with_error_handling("upload config"):
             self.api.upload_file(

{nshtrainer-0.26.1.dist-info → nshtrainer-0.26.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nshtrainer
-Version: 0.26.1
+Version: 0.26.2
 Summary: 
 Author: Nima Shoghi
 Author-email: nimashoghi@gmail.com

{nshtrainer-0.26.1.dist-info → nshtrainer-0.26.2.dist-info}/RECORD

@@ -4,7 +4,7 @@ nshtrainer/_checkpoint/loader.py,sha256=5vjg-OFChXJjgiOVv8vnV8nwTscfdDtEdxQRz6uP
 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/_hf_hub.py,sha256=v1DV3Vn6Pdwr2KYI7yL_Xv1dyJLaG7yZhKFWZFy3QFk,13087
 nshtrainer/callbacks/__init__.py,sha256=4qocBDzQbLLhhbIEfvbA3SQB_Dy9ZJH7keMwPay-ZS8,2359
 nshtrainer/callbacks/_throughput_monitor_callback.py,sha256=aJo_11rc4lo0IYOd-kHmPDtzdC4ctgXyRudkRJqH4m4,23184
 nshtrainer/callbacks/actsave.py,sha256=qbnaKts4_dvjPeAaPtv7Ds12_vEWzaHUfg_--49NB9I,4041
@@ -87,6 +87,6 @@ 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-0.26.1.dist-info/METADATA,sha256=tMMpyg1BTKec5d69ziW6XBxDXaI0gSK5tDMPCmj7VCQ,916
-nshtrainer-0.26.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-nshtrainer-0.26.1.dist-info/RECORD,,
+nshtrainer-0.26.2.dist-info/METADATA,sha256=d3vWjdB9FT6fbWJPkyBI-4M18ekg2WcJmJiKasExchM,916
+nshtrainer-0.26.2.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+nshtrainer-0.26.2.dist-info/RECORD,,