[FSDP] Upstream fairseq big changes (#956)

* made gradient predivide factor configurable

* fix lints

Co-authored-by: Your Name <you@example.com>

fairscale/nn/data_parallel/fully_sharded_data_parallel.py

@@ -331,6 +331,7 @@ def __init__(
         cpu_offload: bool = False,
         offload_config: Optional[OffloadConfig] = None,
         state_dict_on_rank_0_only: bool = False,
+        gradient_predivide_factor: Optional[float] = None,
     ):
         try:
             import torch._C
@@ -399,7 +400,9 @@ def __init__(
         # Experimental feature for now. Use at your own risk.
         self.ssd_offload = True if offload_config and offload_config.offload_type == "ssd_offload" else False
 
-        self.gradient_predivide_factor: float = self._get_gradient_predivide_factor(self.world_size)
+        self.gradient_predivide_factor: float = gradient_predivide_factor or self._get_gradient_predivide_factor(
+            self.world_size
+        )
         self.gradient_postdivide_factor: float = self.world_size / self.gradient_predivide_factor
 
         self.numel_padded_per_param: List[int] = []