[Feature]: support FP8 communication in DDP, FSDP, Gemini (#5928)

* support fp8_communication in the Torch DDP grad comm, FSDP grad comm, and FSDP params comm

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* implement communication hook for FSDP params all-gather

* added unit test for fp8 operators

* support fp8 communication in GeminiPlugin

* update training scripts to support fsdp and fp8 communication

* fixed some minor bugs observed in unit test

* add all_gather_into_tensor_flat_fp8

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* add skip the test if torch < 2.2.0

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* add skip the test if torch < 2.2.0

* add skip the test if torch < 2.2.0

* add fp8_comm flag

* rebase latest fp8 operators

* rebase latest fp8 operators

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

colossalai/booster/plugin/torch_ddp_plugin.py

@@ -177,6 +177,7 @@ class TorchDDPPlugin(DPPluginBase):
         check_reduction: bool = False,
         gradient_as_bucket_view: bool = False,
         static_graph: bool = False,
+        fp8_communication: bool = False,
     ) -> None:
         super().__init__()
         self.ddp_kwargs = dict(
@@ -187,6 +188,7 @@ class TorchDDPPlugin(DPPluginBase):
             gradient_as_bucket_view=gradient_as_bucket_view,
             static_graph=static_graph,
         )
+        self.fp8_communication = fp8_communication
 
     def support_no_sync(self) -> bool:
         return True
@@ -226,6 +228,11 @@ class TorchDDPPlugin(DPPluginBase):
         if optimizer is not None and not isinstance(optimizer, OptimizerWrapper):
             optimizer = OptimizerWrapper(optimizer)
 
+        if self.fp8_communication:
+            from colossalai.quantization.fp8 import fp8_compress_ddp_grad_comm_hook_async
+
+            model.module.register_comm_hook(None, fp8_compress_ddp_grad_comm_hook_async)
+
         return model, optimizer, criterion, dataloader, lr_scheduler
 
     def control_checkpoint_io(self) -> bool: