Merge branch 'main' into support-diffusers-ckpt-gguf

Author: sayakpaul

src/diffusers/hooks/group_offloading.py

@@ -95,7 +95,7 @@ def __init__(
         self.offload_to_disk_path = offload_to_disk_path
         self._is_offloaded_to_disk = False
 
-        if self.offload_to_disk_path:
+        if self.offload_to_disk_path is not None:
             # Instead of `group_id or str(id(self))` we do this because `group_id` can be "" as well.
             self.group_id = group_id if group_id is not None else str(id(self))
             short_hash = _compute_group_hash(self.group_id)
@@ -115,6 +115,12 @@ def __init__(
         else:
             self.cpu_param_dict = self._init_cpu_param_dict()
 
+        self._torch_accelerator_module = (
+            getattr(torch, torch.accelerator.current_accelerator().type)
+            if hasattr(torch, "accelerator")
+            else torch.cuda
+        )
+
     def _init_cpu_param_dict(self):
         cpu_param_dict = {}
         if self.stream is None:
@@ -138,112 +144,76 @@ def _init_cpu_param_dict(self):
 
     @contextmanager
     def _pinned_memory_tensors(self):
-        pinned_dict = {}
         try:
-            for param, tensor in self.cpu_param_dict.items():
-                if not tensor.is_pinned():
-                    pinned_dict[param] = tensor.pin_memory()
-                else:
-                    pinned_dict[param] = tensor
-
+            pinned_dict = {
+                param: tensor.pin_memory() if not tensor.is_pinned() else tensor
+                for param, tensor in self.cpu_param_dict.items()
+            }
             yield pinned_dict
-
         finally:
             pinned_dict = None
 
-    def _transfer_tensor_to_device(self, tensor, source_tensor, current_stream=None):
+    def _transfer_tensor_to_device(self, tensor, source_tensor):
         tensor.data = source_tensor.to(self.onload_device, non_blocking=self.non_blocking)
-        if self.record_stream and current_stream is not None:
-            tensor.data.record_stream(current_stream)
+        if self.record_stream:
+            tensor.data.record_stream(self._torch_accelerator_module.current_stream())
 
-    def _process_tensors_from_modules(self, pinned_memory=None, current_stream=None):
+    def _process_tensors_from_modules(self, pinned_memory=None):
         for group_module in self.modules:
             for param in group_module.parameters():
                 source = pinned_memory[param] if pinned_memory else param.data
-                self._transfer_tensor_to_device(param, source, current_stream)
+                self._transfer_tensor_to_device(param, source)
             for buffer in group_module.buffers():
                 source = pinned_memory[buffer] if pinned_memory else buffer.data
-                self._transfer_tensor_to_device(buffer, source, current_stream)
+                self._transfer_tensor_to_device(buffer, source)
 
         for param in self.parameters:
             source = pinned_memory[param] if pinned_memory else param.data
-            self._transfer_tensor_to_device(param, source, current_stream)
+            self._transfer_tensor_to_device(param, source)
 
         for buffer in self.buffers:
             source = pinned_memory[buffer] if pinned_memory else buffer.data
-            self._transfer_tensor_to_device(buffer, source, current_stream)
+            self._transfer_tensor_to_device(buffer, source)
 
-    def _onload_from_disk(self, current_stream):
+    def _onload_from_disk(self):
         if self.stream is not None:
-            loaded_cpu_tensors = safetensors.torch.load_file(self.safetensors_file_path, device="cpu")
-
-            for key, tensor_obj in self.key_to_tensor.items():
-                self.cpu_param_dict[tensor_obj] = loaded_cpu_tensors[key]
-
-            with self._pinned_memory_tensors() as pinned_memory:
-                for key, tensor_obj in self.key_to_tensor.items():
-                    self._transfer_tensor_to_device(tensor_obj, pinned_memory[tensor_obj], current_stream)
-
-            self.cpu_param_dict.clear()
-
-        else:
-            onload_device = (
-                self.onload_device.type if isinstance(self.onload_device, torch.device) else self.onload_device
-            )
-            loaded_tensors = safetensors.torch.load_file(self.safetensors_file_path, device=onload_device)
-            for key, tensor_obj in self.key_to_tensor.items():
-                tensor_obj.data = loaded_tensors[key]
+            # Wait for previous Host->Device transfer to complete
+            self.stream.synchronize()
 
-    def _onload_from_memory(self, current_stream):
-        if self.stream is not None:
-            with self._pinned_memory_tensors() as pinned_memory:
-                self._process_tensors_from_modules(pinned_memory, current_stream)
-        else:
-            self._process_tensors_from_modules(None, current_stream)
+        context = nullcontext() if self.stream is None else self._torch_accelerator_module.stream(self.stream)
+        current_stream = self._torch_accelerator_module.current_stream() if self.record_stream else None
 
-    @torch.compiler.disable()
-    def onload_(self):
-        torch_accelerator_module = (
-            getattr(torch, torch.accelerator.current_accelerator().type)
-            if hasattr(torch, "accelerator")
-            else torch.cuda
-        )
-        context = nullcontext() if self.stream is None else torch_accelerator_module.stream(self.stream)
-        current_stream = torch_accelerator_module.current_stream() if self.record_stream else None
+        with context:
+            # Load to CPU (if using streams) or directly to target device, pin, and async copy to device
+            device = str(self.onload_device) if self.stream is None else "cpu"
+            loaded_tensors = safetensors.torch.load_file(self.safetensors_file_path, device=device)
 
-        if self.offload_to_disk_path:
             if self.stream is not None:
-                # Wait for previous Host->Device transfer to complete
-                self.stream.synchronize()
-
-            with context:
-                if self.stream is not None:
-                    # Load to CPU, pin, and async copy to device for overlapping transfer and compute
-                    loaded_cpu_tensors = safetensors.torch.load_file(self.safetensors_file_path, device="cpu")
-                    for key, tensor_obj in self.key_to_tensor.items():
-                        pinned_tensor = loaded_cpu_tensors[key].pin_memory()
-                        tensor_obj.data = pinned_tensor.to(self.onload_device, non_blocking=self.non_blocking)
-                        if self.record_stream:
-                            tensor_obj.data.record_stream(current_stream)
-                else:
-                    # Load directly to the target device (synchronous)
-                    onload_device = (
-                        self.onload_device.type if isinstance(self.onload_device, torch.device) else self.onload_device
-                    )
-                    loaded_tensors = safetensors.torch.load_file(self.safetensors_file_path, device=onload_device)
-                    for key, tensor_obj in self.key_to_tensor.items():
-                        tensor_obj.data = loaded_tensors[key]
-            return
+                for key, tensor_obj in self.key_to_tensor.items():
+                    pinned_tensor = loaded_tensors[key].pin_memory()
+                    tensor_obj.data = pinned_tensor.to(self.onload_device, non_blocking=self.non_blocking)
+                    if self.record_stream:
+                        tensor_obj.data.record_stream(current_stream)
+            else:
+                onload_device = (
+                    self.onload_device.type if isinstance(self.onload_device, torch.device) else self.onload_device
+                )
+                loaded_tensors = safetensors.torch.load_file(self.safetensors_file_path, device=onload_device)
+                for key, tensor_obj in self.key_to_tensor.items():
+                    tensor_obj.data = loaded_tensors[key]
 
+    def _onload_from_memory(self):
         if self.stream is not None:
             # Wait for previous Host->Device transfer to complete
             self.stream.synchronize()
 
+        context = nullcontext() if self.stream is None else self._torch_accelerator_module.stream(self.stream)
         with context:
-            if self.offload_to_disk_path:
-                self._onload_from_disk(current_stream)
+            if self.stream is not None:
+                with self._pinned_memory_tensors() as pinned_memory:
+                    self._process_tensors_from_modules(pinned_memory)
             else:
-                self._onload_from_memory(current_stream)
+                self._process_tensors_from_modules(None)
 
     def _offload_to_disk(self):
         # TODO: we can potentially optimize this code path by checking if the _all_ the desired
@@ -264,33 +234,36 @@ def _offload_to_disk(self):
             tensor_obj.data = torch.empty_like(tensor_obj.data, device=self.offload_device)
 
     def _offload_to_memory(self):
-        torch_accelerator_module = (
-            getattr(torch, torch.accelerator.current_accelerator().type)
-            if hasattr(torch, "accelerator")
-            else torch.cuda
-        )
         if self.stream is not None:
             if not self.record_stream:
-                torch_accelerator_module.current_stream().synchronize()
+                self._torch_accelerator_module.current_stream().synchronize()
+
             for group_module in self.modules:
                 for param in group_module.parameters():
                     param.data = self.cpu_param_dict[param]
             for param in self.parameters:
                 param.data = self.cpu_param_dict[param]
             for buffer in self.buffers:
                 buffer.data = self.cpu_param_dict[buffer]
-
         else:
             for group_module in self.modules:
-                group_module.to(self.offload_device, non_blocking=self.non_blocking)
+                group_module.to(self.offload_device, non_blocking=False)
             for param in self.parameters:
-                param.data = param.data.to(self.offload_device, non_blocking=self.non_blocking)
+                param.data = param.data.to(self.offload_device, non_blocking=False)
             for buffer in self.buffers:
-                buffer.data = buffer.data.to(self.offload_device, non_blocking=self.non_blocking)
+                buffer.data = buffer.data.to(self.offload_device, non_blocking=False)
+
+    @torch.compiler.disable()
+    def onload_(self):
+        r"""Onloads the group of parameters to the onload_device."""
+        if self.offload_to_disk_path is not None:
+            self._onload_from_disk()
+        else:
+            self._onload_from_memory()
 
     @torch.compiler.disable()
     def offload_(self):
-        r"""Offloads the group of modules to the offload_device."""
+        r"""Offloads the group of parameters to the offload_device."""
         if self.offload_to_disk_path:
             self._offload_to_disk()
         else:
@@ -307,11 +280,9 @@ class GroupOffloadingHook(ModelHook):
 
     _is_stateful = False
 
-    def __init__(
-        self, group: ModuleGroup, next_group: Optional[ModuleGroup] = None, *, config: GroupOffloadingConfig
-    ) -> None:
+    def __init__(self, group: ModuleGroup, *, config: GroupOffloadingConfig) -> None:
         self.group = group
-        self.next_group = next_group
+        self.next_group: Optional[ModuleGroup] = None
         self.config = config
 
     def initialize_hook(self, module: torch.nn.Module) -> torch.nn.Module:
@@ -331,9 +302,23 @@ def pre_forward(self, module: torch.nn.Module, *args, **kwargs):
         if self.group.onload_leader == module:
             if self.group.onload_self:
                 self.group.onload_()
-            if self.next_group is not None and not self.next_group.onload_self:
+
+            should_onload_next_group = self.next_group is not None and not self.next_group.onload_self
+            if should_onload_next_group:
                 self.next_group.onload_()
 
+            should_synchronize = (
+                not self.group.onload_self and self.group.stream is not None and not should_onload_next_group
+            )
+            if should_synchronize:
+                # If this group didn't onload itself, it means it was asynchronously onloaded by the
+                # previous group. We need to synchronize the side stream to ensure parameters
+                # are completely loaded to proceed with forward pass. Without this, uninitialized
+                # weights will be used in the computation, leading to incorrect results
+                # Also, we should only do this synchronization if we don't already do it from the sync call in
+                # self.next_group.onload_, hence the `not should_onload_next_group` check.
+                self.group.stream.synchronize()
+
         args = send_to_device(args, self.group.onload_device, non_blocking=self.group.non_blocking)
         kwargs = send_to_device(kwargs, self.group.onload_device, non_blocking=self.group.non_blocking)
         return args, kwargs
@@ -459,8 +444,8 @@ def pre_forward(self, module, *args, **kwargs):
 
 def apply_group_offloading(
     module: torch.nn.Module,
-    onload_device: torch.device,
-    offload_device: torch.device = torch.device("cpu"),
+    onload_device: Union[str, torch.device],
+    offload_device: Union[str, torch.device] = torch.device("cpu"),
     offload_type: Union[str, GroupOffloadingType] = "block_level",
     num_blocks_per_group: Optional[int] = None,
     non_blocking: bool = False,
@@ -546,6 +531,8 @@ def apply_group_offloading(
         ```
     """
 
+    onload_device = torch.device(onload_device) if isinstance(onload_device, str) else onload_device
+    offload_device = torch.device(offload_device) if isinstance(offload_device, str) else offload_device
     offload_type = GroupOffloadingType(offload_type)
 
     stream = None
@@ -633,7 +620,7 @@ def _apply_group_offloading_block_level(module: torch.nn.Module, config: GroupOf
     # Apply group offloading hooks to the module groups
     for i, group in enumerate(matched_module_groups):
         for group_module in group.modules:
-            _apply_group_offloading_hook(group_module, group, None, config=config)
+            _apply_group_offloading_hook(group_module, group, config=config)
 
     # Parameters and Buffers of the top-level module need to be offloaded/onloaded separately
     # when the forward pass of this module is called. This is because the top-level module is not
@@ -662,9 +649,9 @@ def _apply_group_offloading_block_level(module: torch.nn.Module, config: GroupOf
         group_id=f"{module.__class__.__name__}_unmatched_group",
     )
     if config.stream is None:
-        _apply_group_offloading_hook(module, unmatched_group, None, config=config)
+        _apply_group_offloading_hook(module, unmatched_group, config=config)
     else:
-        _apply_lazy_group_offloading_hook(module, unmatched_group, None, config=config)
+        _apply_lazy_group_offloading_hook(module, unmatched_group, config=config)
 
 
 def _apply_group_offloading_leaf_level(module: torch.nn.Module, config: GroupOffloadingConfig) -> None:
@@ -693,7 +680,7 @@ def _apply_group_offloading_leaf_level(module: torch.nn.Module, config: GroupOff
             onload_self=True,
             group_id=name,
         )
-        _apply_group_offloading_hook(submodule, group, None, config=config)
+        _apply_group_offloading_hook(submodule, group, config=config)
         modules_with_group_offloading.add(name)
 
     # Parameters and Buffers at all non-leaf levels need to be offloaded/onloaded separately when the forward pass
@@ -740,7 +727,7 @@ def _apply_group_offloading_leaf_level(module: torch.nn.Module, config: GroupOff
             onload_self=True,
             group_id=name,
         )
-        _apply_group_offloading_hook(parent_module, group, None, config=config)
+        _apply_group_offloading_hook(parent_module, group, config=config)
 
     if config.stream is not None:
         # When using streams, we need to know the layer execution order for applying prefetching (to overlap data transfer
@@ -762,13 +749,12 @@ def _apply_group_offloading_leaf_level(module: torch.nn.Module, config: GroupOff
             onload_self=True,
             group_id=_GROUP_ID_LAZY_LEAF,
         )
-        _apply_lazy_group_offloading_hook(module, unmatched_group, None, config=config)
+        _apply_lazy_group_offloading_hook(module, unmatched_group, config=config)
 
 
 def _apply_group_offloading_hook(
     module: torch.nn.Module,
     group: ModuleGroup,
-    next_group: Optional[ModuleGroup] = None,
     *,
     config: GroupOffloadingConfig,
 ) -> None:
@@ -777,14 +763,13 @@ def _apply_group_offloading_hook(
     # We may have already registered a group offloading hook if the module had a torch.nn.Parameter whose parent
     # is the current module. In such cases, we don't want to overwrite the existing group offloading hook.
     if registry.get_hook(_GROUP_OFFLOADING) is None:
-        hook = GroupOffloadingHook(group, next_group, config=config)
+        hook = GroupOffloadingHook(group, config=config)
         registry.register_hook(hook, _GROUP_OFFLOADING)
 
 
 def _apply_lazy_group_offloading_hook(
     module: torch.nn.Module,
     group: ModuleGroup,
-    next_group: Optional[ModuleGroup] = None,
     *,
     config: GroupOffloadingConfig,
 ) -> None:
@@ -793,7 +778,7 @@ def _apply_lazy_group_offloading_hook(
     # We may have already registered a group offloading hook if the module had a torch.nn.Parameter whose parent
     # is the current module. In such cases, we don't want to overwrite the existing group offloading hook.
     if registry.get_hook(_GROUP_OFFLOADING) is None:
-        hook = GroupOffloadingHook(group, next_group, config=config)
+        hook = GroupOffloadingHook(group, config=config)
         registry.register_hook(hook, _GROUP_OFFLOADING)
 
     lazy_prefetch_hook = LazyPrefetchGroupOffloadingHook()