Add support to Fairscale Parallel Layers

Nobody has claimed this yet.

### Description & Motivation

Add support to Fairscale parallel layers:

```python
from fairscale.nn.model_parallel.layers import (
ColumnParallelLinear,
RowParallelLinear,
VocabParallelEmbedding,
)
```

that requires `_MODEL_PARALLEL_GROUP` to be initialized

```
self.tok_embeddings = VocabParallelEmbedding(
File "/home/coder/.local/lib/python3.8/site-packages/fairscale/nn/model_parallel/layers.py", line 118, in __init__
self.num_embeddings, get_model_parallel_rank(), get_model_parallel_world_size()
File "/home/coder/.local/lib/python3.8/site-packages/fairscale/nn/model_parallel/initialize.py", line 157, in get_model_parallel_rank
return torch.distributed.get_rank(group=get_model_parallel_group())
File "/home/coder/.local/lib/python3.8/site-packages/fairscale/nn/model_parallel/initialize.py", line 128, in get_model_parallel_group
assert _MODEL_PARALLEL_GROUP is not None, "model parallel group is not initialized"
AssertionError: model parallel group is not initialized
```

See https://github.com/Lightning-AI/pytorch-lightning/issues/20234 for details.

### Pitch

To support Llama3.1 initialization that requires Fariscale in the original release with parallel layers.

### Alternatives

Get rid of parallel layers or wrap into Lightning (if it is not available already)

### Additional context

_No response_

cc @lantiga @borda

Start by reading Fairscale's model_parallel/layers.py and initialize.py traceback, then review the linked Lightning issue #20234. Done means Llama3.1 initialization can use the parallel layers without the model-parallel-group assertion.

Written by the indexing model from the issue text.