fix on_fit_start (#3616)

* init

* fix call_hook args

pytorch_lightning/trainer/callback_hook.py

@@ -46,10 +46,10 @@ class TrainerCallbackHookMixin(ABC):
         for callback in self.callbacks:
             callback.on_init_end(self)
 
-    def on_fit_start(self, model):
+    def on_fit_start(self):
         """Called when the trainer initialization begins, model has not yet been set."""
         for callback in self.callbacks:
-            callback.on_fit_start(self, model)
+            callback.on_fit_start(self, self.get_model())
 
     def on_fit_end(self):
         """Called when the trainer initialization begins, model has not yet been set."""

pytorch_lightning/trainer/trainer.py

@@ -284,9 +284,6 @@ class Trainer(
         # setup data, etc...
         self.train_loop.setup_fit(model, train_dataloader, val_dataloaders, datamodule)
 
-        # hook
-        self.call_hook('on_fit_start', model)
-
         # hook
         self.data_connector.prepare_data(model)
 
@@ -298,6 +295,10 @@ class Trainer(
         # -------------------------
         self.accelerator_backend = self.accelerator_connector.select_accelerator()
         self.accelerator_backend.setup(model)
+
+        # hook
+        self.call_hook('on_fit_start')
+
         results = self.accelerator_backend.train()
         self.accelerator_backend.teardown()