[zero] fix error for BEiT models (#2169)

* [zero] fix error for BEiT models

* [ColoParameter] add unpack operation for tuple arguments

* fix bugs

* fix chunkv2 unit testing

* add assertion for gradient state

colossalai/gemini/chunk/chunk.py

@@ -18,9 +18,9 @@ class TensorState(Enum):
 
 
 STATE_TRANS = ((TensorState.FREE, TensorState.HOLD), (TensorState.FREE, TensorState.COMPUTE),
-               (TensorState.HOLD, TensorState.FREE), (TensorState.HOLD, TensorState.COMPUTE),
-               (TensorState.COMPUTE, TensorState.HOLD), (TensorState.COMPUTE, TensorState.HOLD_AFTER_BWD),
-               (TensorState.COMPUTE, TensorState.READY_FOR_REDUCE), (TensorState.HOLD_AFTER_BWD, TensorState.COMPUTE),
+               (TensorState.HOLD, TensorState.FREE), (TensorState.HOLD, TensorState.COMPUTE), (TensorState.COMPUTE,
+                                                                                               TensorState.HOLD),
+               (TensorState.COMPUTE, TensorState.HOLD_AFTER_BWD), (TensorState.HOLD_AFTER_BWD, TensorState.COMPUTE),
                (TensorState.HOLD_AFTER_BWD, TensorState.READY_FOR_REDUCE), (TensorState.READY_FOR_REDUCE,
                                                                             TensorState.HOLD))
 

colossalai/nn/_ops/linear.py

@@ -1,11 +1,13 @@
-import torch.nn.functional as F
-from typing import Optional
-from ._utils import GeneralTensor, convert_to_colo_tensor
-from colossalai.tensor.op_wrapper import colo_op_impl
-from ._utils import reduce_input, reduce_grad
-from colossalai.tensor import ComputePattern, ComputeSpec, ColoTensor, ShardSpec, ReplicaSpec, ColoTensorSpec
-from colossalai.tensor.sharding_spec import ShardingSpec
 from copy import deepcopy
+from typing import Optional
+
+import torch.nn.functional as F
+
+from colossalai.tensor import ColoTensor, ColoTensorSpec, ComputePattern, ComputeSpec, ReplicaSpec, ShardSpec
+from colossalai.tensor.op_wrapper import colo_op_impl
+from colossalai.tensor.sharding_spec import ShardingSpec
+
+from ._utils import GeneralTensor, convert_to_colo_tensor, reduce_grad, reduce_input
 
 
 def colo_linear_1drow(input_tensor: ColoTensor, weight: ColoTensor, bias: Optional[ColoTensor]) -> 'ColoTensor':
@@ -162,10 +164,8 @@ def _has_sharding_spec(tensor):
 
 
 @colo_op_impl(F.linear)
-def colo_linear(input_tensor: GeneralTensor,
-                weight: GeneralTensor,
-                bias: Optional[GeneralTensor] = None) -> 'ColoTensor':
+def colo_linear(input: GeneralTensor, weight: GeneralTensor, bias: Optional[GeneralTensor] = None) -> 'ColoTensor':
     if _has_sharding_spec(weight):
-        return _new_colo_linear_imp(input_tensor, weight, bias)
+        return _new_colo_linear_imp(input, weight, bias)
     else:
-        return colo_linear_imp(input_tensor, weight, bias)
+        return colo_linear_imp(input, weight, bias)

colossalai/nn/parallel/data_parallel.py

@@ -283,7 +283,9 @@ class ZeroDDP(ColoDDP):
             p.grad = None
 
     def _post_backward(self):
-        assert self.chunk_manager.accessed_mem == 0
+        if self.chunk_manager.accessed_mem != 0:
+            raise RuntimeError("ZERO DDP error: the synchronization of gradients doesn't exit properly.",
+                               "The most possible reason is that the model is not compatible with ZeroDDP.")
         self._setup_grads_ptr()
         self._logger.debug(
             f'comp cuda demand time: {self.gemini_manager._comp_cuda_demand_time}, layout time: {self.gemini_manager._layout_time}, evict time: {self.gemini_manager._evict_time}, CPU->CUDA vol: {self.gemini_manager._h2d_volume}B, CUDA->CPU vol: {self.gemini_manager._d2h_volume}'
@@ -304,8 +306,9 @@ class ZeroDDP(ColoDDP):
         empty_grad = torch.empty_like(grad)
         free_storage(empty_grad)
         with torch._C.DisableTorchFunction():
-            self.chunk_manager.trans_tensor_state(p, TensorState.READY_FOR_REDUCE)
             chunk = self.chunk_manager.get_chunk(p)
+            assert chunk.tensors_info[p].state == TensorState.HOLD_AFTER_BWD
+            self.chunk_manager.trans_tensor_state(p, TensorState.READY_FOR_REDUCE)
             chunk.copy_tensor_to_chunk_slice(p, grad)
             reduced = self.chunk_manager.reduce_chunk(chunk)
             if reduced:

colossalai/tensor/colo_parameter.py

@@ -8,8 +8,25 @@ from colossalai.tensor.param_op_hook import ColoParamOpHookManager
 from colossalai.tensor.tensor_spec import ColoTensorSpec
 
 
-def filter_args(func, *args):
-    return [arg for arg in args if func(arg)]
+def filter_colo_parameters(*args, **kwargs):
+    param_list = []
+
+    def get_colo_parameters(element) -> None:
+        if isinstance(element, list) or isinstance(element, tuple):
+            for e in element:
+                get_colo_parameters(e)
+        elif isinstance(element, dict):
+            raise RuntimeError("Found Dict: ColoParameter can't deal with complicated arguments.")
+        elif isinstance(element, ColoParameter):
+            param_list.append(element)
+        return
+
+    for a in args:
+        get_colo_parameters(a)
+    for v in kwargs.values():
+        get_colo_parameters(v)
+
+    return param_list
 
 
 def replace_args(args, kwargs, new_args):
@@ -62,7 +79,7 @@ class ColoParameter(ColoTensor, torch.nn.Parameter):
             if not func.__name__.startswith('__'):
                 if kwargs is None:
                     kwargs = {}
-                params = filter_args(lambda arg: isinstance(arg, ColoParameter), *args, *kwargs.values())
+                params = filter_colo_parameters(*args, **kwargs)
                 if len(params) > 0:
                     with torch._C.DisableTorchFunction():
                         new_args = ColoParamOpHookManager.pre_op(params, *args, *kwargs.values())

colossalai/tensor/colo_tensor.py

@@ -57,7 +57,7 @@ class ColoTensor(torch.Tensor):
     The Colotensor can be initialized with a PyTorch tensor in the following ways.
 
         >>> pg = ProcessGroup()
-        >>> colo_t1 = ColoTensor(torch.randn(2,3), spec = ColoTensorSpec(pg, ReplicaSpec())
+        >>> colo_t1 = ColoTensor(torch.randn(2,3), spec = ColoTensorSpec(pg, ReplicaSpec()))
         >>> # The tensor passed in is a tensor after sharding but not a global tensor.
         >>> shard_spec = ShardSpec(process_group=ProcessGroup(tp=world_size),
         >>>                 dims=[0],

colossalai/tensor/param_op_hook.py

@@ -82,16 +82,26 @@ class ColoParamOpHookManager:
     @staticmethod
     def pre_op(params: List[torch.Tensor], *args: Any) -> list:
         ColoParamOpHookManager._trigger_pre_forward(params)
-        args_info = _get_colo_tensors_info(*args)
-        rets = PreFwdPostBwd.apply(params, *args)
-        return _update_colo_tensors(args_info, *rets)
+        grad_args, rear_args = _get_grad_args(*args)
+        colo_info = _get_colo_tensors_info(*grad_args)
+        rets = PreFwdPostBwd.apply(params, *grad_args)
+        update_args = _update_colo_tensors(colo_info, *rets)
+        if rear_args is None:
+            return update_args
+        else:
+            arg_zero = (tuple(update_args),)
+            return arg_zero + rear_args
 
     @staticmethod
     def post_op(params: List[torch.Tensor], arg: Any) -> Any:
         ColoParamOpHookManager._trigger_post_forward(params)
-        arg_info = _get_colo_tensors_info(arg)
+        colo_info = _get_colo_tensors_info(arg)
         ret = PostFwdPreBwd.apply(params, arg)
-        return _unpack_args(_update_colo_tensors(arg_info, ret))
+        res = _update_colo_tensors(colo_info, ret)
+        if len(res) == 1:
+            return res[0]
+        else:
+            return res
 
     @staticmethod
     def has_hook() -> bool:
@@ -103,7 +113,7 @@ class PreFwdPostBwd(torch.autograd.Function):
     @staticmethod
     def forward(ctx, params, *args):
         ctx.params = params
-        return _unpack_args(args)
+        return args
 
     @staticmethod
     def backward(ctx, *grads):
@@ -124,10 +134,29 @@ class PostFwdPreBwd(torch.autograd.Function):
         return (None,) + grads
 
 
-def _unpack_args(args):
-    if len(args) == 1:
-        return args[0]
-    return args
+def _is_grad_tensor(obj) -> bool:
+    if torch.is_tensor(obj):
+        if obj.grad_fn is not None or obj.requires_grad:
+            return True
+    return False
+
+
+def _get_grad_args(*args):
+    # returns the identical args if there is a grad tensor
+    for obj in args:
+        if _is_grad_tensor(obj):
+            return args, None
+    # otherwise, the first arguement should be a tuple of grad tensors
+    # if there is no grad tensor, the backward of PreFwdPostBwd can't be triggered
+    arg_zero = args[0]
+    if not isinstance(arg_zero, tuple):
+        raise NotImplementedError("Some torch function is incompatible because of its complcated inputs.")
+    check_grad_flag = False
+    for obj in arg_zero:
+        check_grad_flag |= _is_grad_tensor(obj)
+    if not check_grad_flag:
+        raise NotImplementedError("Some torch function is incompatible because of its complcated inputs.")
+    return arg_zero, args[1:]
 
 
 def _get_colo_tensors_info(*args) -> list:

tests/test_gemini/update/test_chunkv2.py

@@ -90,6 +90,7 @@ def exam_chunk_basic(init_device, keep_gathered, pin_memory):
 
     for param in param_list:
         my_chunk.tensor_trans_state(param, TensorState.COMPUTE)
+        my_chunk.tensor_trans_state(param, TensorState.HOLD_AFTER_BWD)
         my_chunk.tensor_trans_state(param, TensorState.READY_FOR_REDUCE)
 
     assert my_chunk.tensor_state_cnter[TensorState.READY_FOR_REDUCE] == 4