fix: datasets broken import due to HF package and folder name collision (#1730)

This PR resolves the `datasets` import error that were introduced in
#1712. The error causes the following error message:
```Traceback (most recent call last):
  File "./torchtitan/train.py", line 16, in <module>
    import torchtitan.protocols.train_spec as train_spec_module
  File "./torchtitan/__init__.py", line 12, in <module>
    import torchtitan.experiments  # noqa: F401
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "./torchtitan/experiments/__init__.py", line 7, in <module>
    import torchtitan.experiments.llama4  # noqa: F401
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "./torchtitan/experiments/llama4/__init__.py", line 11, in <module>
    from torchtitan.datasets.hf_datasets import build_hf_dataloader
  File "./torchtitan/datasets/hf_datasets.py", line 12, in <module>
    from datasets import Dataset, load_dataset
ImportError: cannot import name 'Dataset' from 'datasets' (./torchtitan/datasets/__init__.py)
```

Why is this happening? It is because #1712 added an `__init__.py` file
to the `datasets` folder. On the surface the PR looks fine, however, it
causes a collision with HF datasets Python package called `datasets`. So
when we try to import the `Dataset` class, we actually want the HF
datasets package and not the local `datasets` folder. This, in turn,
causes an import error. The solution is simple, change the `__init__.py`
name to something else, in our case I changed it to `common.py` as I
found this name most fitting.

Author: idoh

multinode_trainer.slurm

@@ -59,5 +59,5 @@ CONFIG_FILE=${CONFIG_FILE:-"./torchtitan/models/llama3/train_configs/llama3_8b.t
 dcgmi profile --pause
 # adjust sbatch --ntasks and sbatch --nodes above and --nnodes below
 # to your specific node count, and update target launch file.
-srun torchrun --nnodes 4 --nproc_per_node 8 --rdzv_id 101 --rdzv_backend c10d --rdzv_endpoint "$head_node_ip:29500" ./torchtitan/train.py --job.config_file ${CONFIG_FILE} "$@"
+srun torchrun --nnodes 4 --nproc_per_node 8 --rdzv_id 101 --rdzv_backend c10d --rdzv_endpoint "$head_node_ip:29500" -m torchtitan.train --job.config_file ${CONFIG_FILE} "$@"
 dcgmi profile --resume