Fabric FSDP with bitsandbytes plugin is not supported

Description & Motivation

While the transformers library supports FSDP with bitsandbytes 4bit quantization (https://huggingface.co/docs/bitsandbytes/main/en/fsdp_qlora), Fabric appears unable to combine the two:

    mode = "nf4"
    plugin = BitsandbytesPrecision(mode=mode, dtype=torch.bfloat16)
    policy = partial(size_based_auto_wrap_policy, min_num_params=1E8)
    # use data parallel with quantization rather than FSDP:
    fabric = Fabric(accelerator="auto", devices="auto",
                    strategy=FSDPStrategy(
                        auto_wrap_policy=policy, cpu_offload=False,
                        state_dict_type="sharded"),
                    plugins=plugin)

gemma-fine-tuning, pid=3426)   File "/home/ubuntu/miniconda3/envs/fabric/lib/python3.10/site-packages/lightning/fabric/strategies/fsdp.py", line 242, in precision
     raise TypeError(f"The FSDP strategy can only work with the `FSDPPrecision` plugin, found {precision}")
 TypeError: The FSDP strategy can only work with the `FSDPPrecision` plugin, found <lightning.fabric.plugins.precision.bitsandbytes.BitsandbytesPrecision object at 0x78b7aa999ba0>

Pitch

Being able to fine-tuning large models with FSDP and quantization is becoming more important as large multi-modal LLMs require a lot of GPU memory for batches (e.g., multi-images per example and batch size >1).

Alternatives

No response

Additional context

No response

cc @lantiga @borda @justusschock

Start by reproducing the supplied Fabric configuration and inspect lightning/fabric/strategies/fsdp.py, especially the FSDPStrategy precision check and the BitsandbytesPrecision entry point. Done means the shown FSDP and bitsandbytes configuration is accepted without the reported TypeError and its behavior is covered by the project's relevant tests.