Support saving models trained with DeepSpeed in Trainer callbacks

### Feature request Trainer callbacks [pass a model](https://github.com/huggingface/transformers/blob/a7cab3c283312b8d4de5df3bbe719971e24f4281/src/transformers/trainer_callback.py#L418) to any registered callback, but this model cannot be saved if training with DeepSpeed Stage 3 (needs access to `Trainer.accelerator` and `Trainer.model_wrapped`) ### Motivation I have a custom callback that logs a model to a tracking server at the end of training, but need access to `Trainer.accelerator` and `Trainer.model_wrapped` in my callback to prevent skipping saving sharded tensors. I believe this may lead to a bug in the official `wandb` callback which uses a ["fake trainer"](https://github.com/huggingface/transformers/blob/b6c9f47fd6f911450024c52e382e544e5d04387a/src/transformers/integrations/integration_utils.py#L850) (I don't use `wandb` so can't confirm, but see an error similar to one reported in another repo if I try the "fake trainer" approach in my custom callback e.g., https://github.com/OpenAccess-AI-Collective/axolotl/issues/1092) ### Your contribution I can contribute this feature, but wanted to get guidance on the design. Roughly * `trainer_callback.CallbackHandler` receives an additional arg, `accelerator`, which will be passed `Trainer.accelerator` on instantiation * `trainer_callback.CallbackHandler` has an attribute `model_wrapped` which gets updated with `Trainer.model_wrapped` by `Trainer` * `trainer_callback.CallbackHandler.call_event` will pass along `self.accelerator` and `self.model_wrapped` when calling callbacks This allows my callback to do something like `state_dict = accelerator.get_state_dict(model_wrapped)` and pass that along to `model.save_pretrained` as to not skip saving sharded tensors If there is a better design or if this is better classified as a bug, please let me know