mirror of https://github.com/hpcaitech/ColossalAI
[fix] rm output.data after send fwd;
parent
a48afc4a66
commit
ab643c9af7
|
@ -25,6 +25,24 @@ def _wait_p2p(wait_handles: List[torch.cuda.Event]) -> None:
|
|||
req.wait()
|
||||
|
||||
|
||||
def deallocate_output_tensor(out, deallocate_pipeline_outputs=False):
|
||||
"""Pseudo-deallocate (i.e., set to scalar) the output tensor's '.data' field.
|
||||
|
||||
This method should be called right after the output tensor has been
|
||||
sent to the next pipeline stage. At this point, the output tensor is
|
||||
only useful for its '.grad_fn' field, and not its '.data'.
|
||||
"""
|
||||
if (out is None) or (not deallocate_pipeline_outputs):
|
||||
print(
|
||||
f"(out is None) or (not deallocate_pipeline_outputs): {(out is None) or (not deallocate_pipeline_outputs)}"
|
||||
)
|
||||
return
|
||||
assert isinstance(out, torch.Tensor), "expected Tensor, found %s." % type(out).__name__
|
||||
assert out._base is None, "counter-productive to free a view of another tensor."
|
||||
# out.data = torch.empty((1,), device=out.device, dtype=out.dtype,)
|
||||
out.data.storage().resize_(0)
|
||||
|
||||
|
||||
class ZeroBubbleVPipeScheduler(PipelineSchedule):
|
||||
def __init__(
|
||||
self,
|
||||
|
@ -562,10 +580,13 @@ class ZeroBubbleVPipeScheduler(PipelineSchedule):
|
|||
)
|
||||
# add input and output object for backward b
|
||||
self.input_tensors[model_chunk_id].append(input_obj)
|
||||
self.output_tensors[model_chunk_id].append(output_obj)
|
||||
|
||||
# detached output; for bwd b&w, we only need the graph(grad_fn) of output_obj
|
||||
detached_output_obj = output_obj.clone()
|
||||
deallocate_output_tensor(detached_output_obj, deallocate_pipeline_outputs=True)
|
||||
self.output_tensors[model_chunk_id].append(detached_output_obj)
|
||||
# add output object for backward w
|
||||
self.output_tensors_dw[model_chunk_id].append(output_obj)
|
||||
self.output_tensors_dw[model_chunk_id].append(detached_output_obj)
|
||||
|
||||
# Step3: send fwd
|
||||
# add output to send_fwd_buffer
|
||||
|
|
|
@ -2,8 +2,7 @@ from .albert import *
|
|||
from .bert import *
|
||||
from .blip2 import *
|
||||
from .bloom import *
|
||||
|
||||
# from .chatglm2 import *
|
||||
from .chatglm2 import *
|
||||
from .command import *
|
||||
from .deepseek import *
|
||||
from .falcon import *
|
||||
|
|
|
@ -14,7 +14,6 @@ from colossalai.logging import disable_existing_loggers
|
|||
from colossalai.pipeline.schedule.v_schedule import PipelineGraph, ScheduledNode
|
||||
from colossalai.pipeline.schedule.zero_bubble_pp import ZeroBubbleVPipeScheduler
|
||||
from colossalai.pipeline.stage_manager import PipelineStageManager
|
||||
from colossalai.tensor.d_tensor.api import clear_layout_converter
|
||||
from colossalai.testing import parameterize, rerun_if_address_is_in_use, spawn
|
||||
from tests.kit.model_zoo import model_zoo
|
||||
|
||||
|
@ -701,56 +700,13 @@ def run_with_hybridplugin(test_config):
|
|||
],
|
||||
)
|
||||
def run_with_moehybridplugin(test_config):
|
||||
sub_model_zoo = model_zoo.get_sub_registry("transformers_bert")
|
||||
model_zoo.get_sub_registry("transformers_bert")
|
||||
test_config["use_lazy_init"] = False
|
||||
test_config["pp_size"] = 1 # Do NOT test Pipeline Parallel
|
||||
test_config["initial_scale"] = 2**16 # avoid overflow
|
||||
model_list = [
|
||||
"transformers_bert",
|
||||
]
|
||||
clear_layout_converter()
|
||||
torch.set_default_dtype(torch.bfloat16)
|
||||
for name, (model_fn, data_gen_fn, output_transform_fn, loss_fn, _) in sub_model_zoo.items():
|
||||
data_gen_fn()
|
||||
# print(f"data {data}")
|
||||
# if name in model_list:
|
||||
# (
|
||||
# org_model,
|
||||
# org_optimizer,
|
||||
# sharded_model,
|
||||
# sharded_optimizer,
|
||||
# criterion,
|
||||
# booster,
|
||||
# ) = build_model_from_hybrid_plugin(model_fn, loss_fn, test_config, torch.optim.SGD, torch.optim.SGD)
|
||||
|
||||
# org_loss, org_output, sharded_loss, sharded_output = run_forward_backward_with_hybrid_plugin(
|
||||
# org_model, sharded_model, sharded_optimizer, data_gen_fn, output_transform_fn, criterion, booster
|
||||
# )
|
||||
|
||||
# stage_manager = booster.plugin.stage_manager
|
||||
# tp_group = booster.plugin.tp_group
|
||||
|
||||
# bert = unwrap_model(org_model, "BertModel", "bert")
|
||||
# sharded_bert = unwrap_model(sharded_model, "BertModel", "bert")
|
||||
# weight_layer_for_check = ["encoder.layer[0].output.dense", "encoder.layer[1].output.dense"]
|
||||
|
||||
# org_optimizer.step()
|
||||
# sharded_optimizer.step()
|
||||
|
||||
# # check weights
|
||||
# if test_config["precision"] == "bf16":
|
||||
# atol, rtol = 5e-4, 5e-4
|
||||
# else:
|
||||
# atol, rtol = 5e-4, 5e-4
|
||||
# if stage_manager is None or stage_manager.is_first_stage(ignore_chunk=True):
|
||||
# check_weight(bert, sharded_bert, weight_layer_for_check, tp_group, atol=atol, rtol=rtol, dim=1)
|
||||
# # check optim states
|
||||
# # check_dist_optim_state(org_optimizer, sharded_optimizer.optim)
|
||||
|
||||
# clear_layout_converter()
|
||||
# Randomizer.reset_index()
|
||||
# torch.cuda.empty_cache()
|
||||
# print(f"Bert Model Zoo Test Passed")
|
||||
|
||||
|
||||
# TODO:6) support booster & Hybrid base 4)
|
||||
|
|
Loading…
Reference in New Issue