[zero] add inference mode and its unit test (#2418)

colossalai/gemini/gemini_mgr.py

@@ -50,6 +50,17 @@ class GeminiManager:
         self._warmup = True
         self._comp_cuda_demand_time = 0
 
+    def reset_attributes(self):
+        self._compute_idx = -1
+        self._h2d_volume = 0
+        self._d2h_volume = 0
+        self._layout_time = 0
+        self._evict_time = 0
+        self._comp_cuda_demand_time = 0
+
+    def is_warmup(self):
+        return self._warmup
+
     def memstats(self):
         """memstats
 
@@ -73,12 +84,7 @@ class GeminiManager:
         if self._mem_stats_collector and self._warmup:
             self._mem_stats_collector.finish_collection()
         self._warmup = False
-        self._compute_idx = -1
-        self._h2d_volume = 0
-        self._d2h_volume = 0
-        self._layout_time = 0
-        self._evict_time = 0
-        self._comp_cuda_demand_time = 0
+        self.reset_attributes()
 
     def adjust_layout(self, chunks: Tuple[Chunk, ...]) -> None:
         """ Adjust the layout of stateful tensors according to the information provided

colossalai/nn/parallel/data_parallel.py

@@ -268,12 +268,35 @@ class ZeroDDP(ColoDDP):
 
         self._logger = get_dist_logger()
 
+    def _post_forward(self):
+        """This function is only triggered for inference.
+        """
+        access_list = list(self.chunk_manager.accessed_chunks)
+        # we need to scatter all accessed chunks and move them to their original places
+        for chunk in access_list:
+            assert chunk.can_release
+            self.chunk_manager.release_chunk(chunk)
+            first_param = next(iter(chunk.tensors_info))
+            self.chunk_manager.move_chunk(chunk, self.grads_device[first_param])
+        assert self.chunk_manager.accessed_mem == 0
+        # reset all recorded attributes
+        self.gemini_manager.reset_attributes()
+
     def forward(self, *args, **kwargs):
+        # check whether we are in a inference mode
+        grad_flag = torch.is_grad_enabled()
+        if not grad_flag:
+            assert not self.gemini_manager.is_warmup(), "You should run a completed iteration as your warmup iter"
+
         args, kwargs = _cast_float(args, torch.half), _cast_float(kwargs, torch.half)
         self.module.zero_grad(set_to_none=True)
         self.gemini_manager.pre_iter(*args)
         with ColoParamOpHookManager.use_hooks(self.param_op_hook):
             outputs = self.module(*args, **kwargs)
+        # scatter chunks in the inference mode
+        if not grad_flag:
+            self._post_forward()
+
         if self.force_outputs_fp32:
             return _cast_float(outputs, torch.float)
         return outputs

tests/test_gemini/update/test_inference.py

@@ -0,0 +1,122 @@
+from functools import partial
+
+import pytest
+import torch
+import torch.distributed as dist
+import torch.multiprocessing as mp
+from torch.nn.parallel import DistributedDataParallel as DDP
+from torch.testing import assert_close
+
+import colossalai
+from colossalai.amp import convert_to_apex_amp
+from colossalai.gemini.chunk import ChunkManager, init_chunk_manager, search_chunk_configuration
+from colossalai.gemini.gemini_mgr import GeminiManager
+from colossalai.nn.optimizer import HybridAdam
+from colossalai.nn.optimizer.zero_optimizer import ZeroOptimizer
+from colossalai.nn.parallel import ZeroDDP
+from colossalai.testing import parameterize, rerun_if_address_is_in_use
+from colossalai.utils import free_port
+from colossalai.utils.cuda import get_current_device
+from colossalai.utils.model.colo_init_context import ColoInitContext, post_process_colo_init_ctx
+from tests.components_to_test import run_fwd_bwd
+from tests.components_to_test.registry import non_distributed_component_funcs
+from tests.test_tensor.common_utils import debug_print, set_seed
+
+
+def check_param(model: ZeroDDP, torch_model: torch.nn.Module):
+    zero_dict = model.state_dict(only_rank_0=False)
+    torch_dict = torch_model.state_dict()
+
+    for key, value in torch_dict.items():
+        # key is 'module.model.PARAMETER', so we truncate it
+        key = key[7:]
+        assert key in zero_dict, "{} not in ZeRO dictionary.".format(key)
+        temp_zero_value = zero_dict[key].to(device=value.device, dtype=value.dtype)
+        # debug_print([0], "max range: ", key, torch.max(torch.abs(value - temp_zero_value)))
+        assert_close(value, temp_zero_value, rtol=1e-3, atol=4e-3)
+
+
+@parameterize('placement_policy', ['cuda', 'cpu', 'auto', 'const'])
+@parameterize('model_name', ['gpt2'])
+def exam_inference(placement_policy, model_name: str):
+    set_seed(19360226)
+    get_components_func = non_distributed_component_funcs.get_callable(model_name)
+    model_builder, train_dataloader, test_dataloader, optimizer_class, criterion = get_components_func()
+
+    torch_model = model_builder().cuda()
+    amp_config = dict(opt_level='O2', keep_batchnorm_fp32=False, loss_scale=128)
+    torch_optim = torch.optim.Adam(torch_model.parameters(), lr=1e-3)
+    torch_model, torch_optim = convert_to_apex_amp(torch_model, torch_optim, amp_config)
+    torch_model = DDP(torch_model, device_ids=[dist.get_rank()])
+
+    init_dev = get_current_device()
+    with ColoInitContext(device=init_dev):
+        model = model_builder()
+
+    for torch_p, p in zip(torch_model.parameters(), model.parameters()):
+        p.data.copy_(torch_p.data)
+
+    world_size = torch.distributed.get_world_size()
+    config_dict, _ = search_chunk_configuration(model, search_range_mb=1, search_interval_byte=100)
+    config_dict[world_size]['chunk_size'] = 5000
+    config_dict[world_size]['keep_gathered'] = False
+    if placement_policy != 'cuda':
+        init_device = torch.device('cpu')
+    else:
+        init_device = None
+    chunk_manager = ChunkManager(config_dict, init_device=init_device)
+    gemini_manager = GeminiManager(placement_policy, chunk_manager)
+    model = ZeroDDP(model, gemini_manager, pin_memory=True)
+
+    optimizer = HybridAdam(model.parameters(), lr=1e-3)
+    zero_optim = ZeroOptimizer(optimizer, model, initial_scale=128)
+
+    model.eval()
+    torch_model.eval()
+
+    set_seed(dist.get_rank() * 3 + 128)
+    train_dataloader = iter(train_dataloader)
+
+    def train_iter():
+        input_ids, label = next(train_dataloader)
+        input_ids, label = input_ids.cuda(), label.cuda()
+        zero_optim.zero_grad()
+        torch_optim.zero_grad()
+        torch_loss = run_fwd_bwd(torch_model, input_ids, label, criterion, torch_optim)
+        loss = run_fwd_bwd(model, input_ids, label, criterion, zero_optim)
+        assert_close(torch_loss, loss)
+        zero_optim.step()
+        torch_optim.step()
+        check_param(model, torch_model)
+
+    def inference_iter():
+        input_ids, label = next(train_dataloader)
+        input_ids, label = input_ids.cuda(), label.cuda()
+        with torch.no_grad():
+            torch_output = torch_model(input_ids)
+            torch_loss = criterion(torch_output.float(), label)
+            zero_output = model(input_ids)
+            zero_loss = criterion(zero_output.float(), label)
+        assert_close(torch_loss, zero_loss)
+
+    train_iter()
+    inference_iter()
+    train_iter()
+
+
+def run_dist(rank, world_size, port):
+    config = {}
+    colossalai.launch(config=config, rank=rank, world_size=world_size, host='localhost', port=port, backend='nccl')
+    exam_inference()
+
+
+@pytest.mark.dist
+@pytest.mark.parametrize('world_size', [1, 4])
+@rerun_if_address_is_in_use()
+def test_inference(world_size):
+    run_func = partial(run_dist, world_size=world_size, port=free_port())
+    mp.spawn(run_func, nprocs=world_size)
+
+
+if __name__ == '__main__':
+    test_inference(1)