✨ Add logging for training completion and model saving in training scripts

[qgallouedec]

$ trl sft --model_name_or_path trl-internal-testing/tiny-Qwen2ForCausalLM-2.5 --dataset_name trl-internal-testing/zen --dataset_config standard_language_modeling --report_to none --push_to_hub
INFO 09-09 16:18:52 [__init__.py:235] Automatically detected platform cuda.
The following values were not passed to `accelerate launch` and had defaults used instead:
        `--num_processes` was set to a value of `1`
        `--num_machines` was set to a value of `1`
        `--mixed_precision` was set to a value of `'no'`
        `--dynamo_backend` was set to a value of `'no'`
To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`.
[2025-09-09 16:19:02,976] [INFO] [real_accelerator.py:254:get_accelerator] Setting ds_accelerator to cuda (auto detect)
[2025-09-09 16:19:04,846] [INFO] [logging.py:107:log_dist] [Rank -1] [TorchCheckpointEngine] Initialized with serialization = False
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'eos_token_id': 151645, 'pad_token_id': 151643}.
{'train_runtime': 1.0213, 'train_samples_per_second': 49.938, 'train_steps_per_second': 8.813, 'train_loss': 10.362649705674913, 'entropy': 11.929376496209038, 'num_tokens': 474.0, 'mean_token_accuracy': 0.0, 'epoch': 3.0}
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:01<00:00,  8.81it/s]
✅ Training completed.
💾 Model saved to trainer_output.
🤗 Model pushed to the Hub in https://huggingface.co/qgallouedec/trainer_output.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.