TensorRT-LLMs/tensorrt_llm/models/grok/weight.py
2024-09-03 12:14:23 +02:00

36 lines
1.3 KiB
Python

# SPDX-FileCopyrightText: Copyright (c) 2022-2024 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.
from typing import Union
import numpy as np
import torch
def split(v: Union[np.ndarray, torch.Tensor],
tp_size: int,
tp_rank: int,
dim=0):
if tp_size == 1:
return v
assert len(v.shape) > 1 or dim == 0
if isinstance(v, np.ndarray):
return np.ascontiguousarray(
np.split(v, tp_size, axis=dim)[tp_rank].copy())
else:
assert v.shape[dim] % tp_size == 0, \
'Unable to split: shape={v.shape} (dim={dim}) tp_size={tp_size}.'
split_size = v.shape[dim] // tp_size
return v.split(split_size, dim=dim)[tp_rank].clone().detach()