[Gemini] free and allocate cuda memory by tensor.storage, add grad hook (#2040)

colossalai/gemini/memory_tracer/param_tracer_wrapper.py

@@ -1,9 +1,11 @@
 import torch.nn
 
-from colossalai.tensor.colo_parameter import ColoParameter
 from colossalai.tensor.param_op_hook import ParamOpHookManager
 from colossalai.gemini.ophooks.param_trace_hook import ParamTracerHook
+from colossalai.gemini.tensor_utils import free_storage
 from colossalai.nn.parallel.data_parallel import _cast_float
+from functools import partial
+
 
 __all__ = ['ParamTracerWrapper']
 
@@ -13,17 +15,21 @@ class ParamTracerWrapper():
         super().__init__()
         self.module = module
         self.dtype = dtype
-        self.param_op_hook = ParamTracerHook()
+        self.param_op_hook = ParamTracerHook(dtype)
 
         for p in module.parameters():
-            assert isinstance(p, ColoParameter)
             p.data = p.data.to(dtype)
+            if p.requires_grad:
+                p.register_hook(partial(self.grad_handle))
 
         self._cast_buffers_to_cuda_dtype()
 
     def __call__(self, *args, **kwargs):
         return self.forward(*args, **kwargs)
 
+    def grad_handle(self, grad):
+        free_storage(grad)
+
     def _pre_forward(self):
         self.param_op_hook.mem_monitor.start()
 

colossalai/gemini/ophooks/param_trace_hook.py

@@ -7,6 +7,7 @@ import torch
 
 from colossalai.gemini.memory_tracer import SyncCudaMemoryMonitor
 from colossalai.tensor.param_op_hook import ParamOpHook
+from colossalai.gemini.tensor_utils import free_storage, alloc_storage
 
 
 class TrainingPhase(Enum):
@@ -16,25 +17,26 @@ class TrainingPhase(Enum):
 
 class ParamTracerHook(ParamOpHook):
 
-    def __init__(self) -> None:
+    def __init__(self, dtype: torch.dtype = torch.half) -> None:
         super().__init__()
         self._training_phase = TrainingPhase.FORWARD
         self.mem_monitor = SyncCudaMemoryMonitor()
         self._non_model_data_list = []
         self._model_data_list = []
+        self.dtype = dtype
 
-    def _move_params_to_dev(self, params, dev: str) -> int:
-        assert isinstance(dev, str), f"device should be a str not torch.device"
-        comm_volume = 0
+    def _free_cuda_params(self, params):
         for p in params:
-            if p.data.device.type != dev:
-                p.data = p.data.to(dev)
-                comm_volume += p.data.numel() * p.data.element_size()
-            if p.grad is not None:
-                if p.grad.device.type != dev:
-                    p.grad = p.grad.to(dev)
-                    comm_volume += p.grad.numel() * p.grad.element_size()
-        return comm_volume
+            free_storage(p.data)
+
+    def _allocate_params_on_cuda(self, params):
+        for p in params:
+            cur_dev = p.data.device.type
+            if cur_dev == "cpu":
+                # p.data = p.data.to("cuda")
+                p.data = torch.randn(p.data.shape, device="cuda", dtype=self.dtype)
+            elif cur_dev == "cuda":
+                alloc_storage(p.data)
 
     def sample_model_data(self, params):
         data_volume = 0
@@ -49,12 +51,12 @@ class ParamTracerHook(ParamOpHook):
         cuda_volume = self.mem_monitor.finish()
         if len(self._model_data_list):
             self._non_model_data_list.append(cuda_volume - self._model_data_list[-1])
-        self._move_params_to_dev(params, 'cuda')
+        self._allocate_params_on_cuda(params)
         self.sample_model_data(params)
         self.mem_monitor.start()
 
     def post_op(self, params):
-        self._move_params_to_dev(params, 'cpu')
+        self._free_cuda_params(params)
 
     def pre_forward(self, params: List[torch.Tensor]) -> None:
         self.pre_op(params)

colossalai/gemini/tensor_utils.py

@@ -3,6 +3,20 @@ from colossalai.gemini.stateful_tensor import StatefulTensor
 from typing import Union, Tuple
 
 
+def is_storage_empty(tensor: torch.Tensor) -> bool:
+    return tensor.storage().size() == 0
+
+
+def free_storage(tensor: torch.Tensor) -> None:
+    if not is_storage_empty(tensor):
+        tensor.storage().resize_(0)
+
+
+def alloc_storage(tensor: torch.Tensor) -> None:
+    if is_storage_empty(tensor):
+        tensor.storage().resize_(tensor.numel())
+
+
 def colo_tensor_mem_usage(tensor: Union[torch.Tensor, StatefulTensor]) -> Tuple[int, int]:
     if isinstance(tensor, StatefulTensor):
         t = tensor.payload

tests/test_gemini/test_param_tracer.py

@@ -16,7 +16,7 @@ def run_fwd_bwd(model, data, label, criterion, enable_autocast=False, dtype=torc
     model.backward(loss)
 
 def run_param_wrapper_testing():
-    test_models = ['simple_net']
+    test_models = ['simple_net', 'repeated_computed_layers', 'nested_model']
 
     for model_name in test_models:
         get_components_func = non_distributed_component_funcs.get_callable(model_name)