TensorRT-LLMs/tests/functional/test_nccl.py
Kaiyu Xie 75b6210ff4
Kaiyu/update main (#5)
* Update

* Update
2023-10-18 22:38:53 +08:00

147 lines
5.3 KiB
Python

# SPDX-FileCopyrightText: Copyright (c) 2022-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import unittest
from itertools import product
import pytest
import tensorrt as trt
import torch
from cuda import cudart
from parameterized import parameterized
from polygraphy.backend.trt import CreateConfig, EngineFromNetwork
import tensorrt_llm as tllm
from tensorrt_llm import Mapping, Tensor
from tensorrt_llm._ipc_utils import IpcMemory, peer_access
from tensorrt_llm.functional import AllReduceStrategy, allreduce
def custom_name_func(testcase_func, param_num, param):
return "%s_%s" % (
testcase_func.__name__,
parameterized.to_safe_name("_".join(str(x) for x in param.args)),
)
class TestCommunicationPlugin(unittest.TestCase):
def setUp(self):
tllm.logger.set_level('error')
self.world_size = tllm.mpi_world_size()
self.rank = tllm.mpi_rank()
torch.cuda.set_device(self.rank)
cudart.cudaSetDevice(self.rank)
self.reference_tensors = [
torch.full([10000000], i + 1, dtype=torch.float32, device="cuda")
for i in range(self.world_size)
]
self.mapping = Mapping(self.world_size, self.rank, self.world_size,
self.world_size)
@parameterized.expand(list(
product(["bfloat16", 'float16', "float32"], [
AllReduceStrategy.RING, AllReduceStrategy.ONESHOT,
AllReduceStrategy.TWOSHOT
], [64 * 70000, 64 * 70, 64])),
name_func=custom_name_func)
def test_nccl_allreduce(self, dtype: str, strategy: AllReduceStrategy,
size: int):
if self.world_size == 1:
pytest.skip()
workspace = None
ipc_buffers = IpcMemory(self.mapping, IpcMemory.IPC_BUFFERS_SIZE)
ipc_barriers_in = IpcMemory(
self.mapping,
IpcMemory.IPC_BARRIERS_SIZE_PER_GPU * self.mapping.tp_size)
ipc_barriers_out = IpcMemory(
self.mapping,
IpcMemory.IPC_BARRIERS_SIZE_PER_GPU * self.mapping.tp_size)
workspace = torch.tensor(ipc_buffers.serialize() +
ipc_barriers_in.serialize() +
ipc_barriers_out.serialize(),
dtype=torch.int64,
device="cpu")
torch_dtype = tllm._utils.str_dtype_to_torch(dtype)
allreduce_ref = torch.zeros(self.reference_tensors[0][:size].shape,
dtype=torch_dtype,
device="cuda")
for i in range(self.world_size):
allreduce_ref = allreduce_ref + self.reference_tensors[i][:size].to(
torch_dtype)
builder = tllm.Builder()
net = builder.create_network()
net.plugin_config.set_nccl_plugin(dtype)
input = self.reference_tensors[self.rank][:size].to(torch_dtype)
inner_loop = 5
with peer_access(self.mapping):
with tllm.net_guard(net):
network = tllm.default_trtnet()
x = Tensor(name='x',
shape=input.shape,
dtype=tllm.str_dtype_to_trt(dtype))
w = Tensor(name='workspace',
shape=workspace.shape,
dtype=trt.int64)
current = x
for i in range(inner_loop):
current = allreduce(
current, self.mapping.tp_group,
w if strategy != AllReduceStrategy.RING else None, i,
strategy)
output = current.trt_tensor
output.name = 'output'
output.dtype = tllm.str_dtype_to_trt(dtype)
network.mark_output(output)
build_engine = EngineFromNetwork(
(builder.trt_builder, net.trt_network),
config=CreateConfig(
fp16=(dtype == 'float16'),
bf16=(dtype == 'bfloat16'),
precision_constraints='obey',
))
output = torch.zeros_like(input)
stream = torch.cuda.current_stream()
feed_dict = {'x': input, 'workspace': workspace}
session = tllm.runtime.Session.from_engine(build_engine())
session.run(inputs=feed_dict,
outputs={"output": output},
stream=stream.cuda_stream)
torch.cuda.synchronize()
self.assertTrue(
torch.allclose(output.cpu(),
(self.mapping.tp_size**(inner_loop - 1)) *
allreduce_ref.cpu()))
if __name__ == "__main__":
unittest.main()