Files
FastDeploy/fastdeploy/model_executor/model_loader/default_loader_v1.py
bukejiyu 77514e3e1e
Some checks failed
CE Compile Job / ce_job_pre_check (push) Has been cancelled
CE Compile Job / print_ce_job_pre_check_outputs (push) Has been cancelled
CE Compile Job / FD-Clone-Linux (push) Has been cancelled
CE Compile Job / Show Code Archive Output (push) Has been cancelled
CE Compile Job / BUILD_SM8090 (push) Has been cancelled
CE Compile Job / BUILD_SM8689 (push) Has been cancelled
CE Compile Job / CE_UPLOAD (push) Has been cancelled
Deploy GitHub Pages / deploy (push) Has been cancelled
Publish Job / publish_pre_check (push) Has been cancelled
Publish Job / print_publish_pre_check_outputs (push) Has been cancelled
Publish Job / FD-Clone-Linux (push) Has been cancelled
Publish Job / Show Code Archive Output (push) Has been cancelled
Publish Job / BUILD_SM8090 (push) Has been cancelled
Publish Job / BUILD_SM8689 (push) Has been cancelled
Publish Job / PADDLE_PYPI_UPLOAD_8090 (push) Has been cancelled
Publish Job / PADDLE_PYPI_UPLOAD_8689 (push) Has been cancelled
Publish Job / Run FastDeploy Unit Tests and Coverage (push) Has been cancelled
Publish Job / Run FastDeploy LogProb Tests (push) Has been cancelled
Publish Job / Extracted partial CE model tasks to run in CI. (push) Has been cancelled
Publish Job / Run Base Tests (push) Has been cancelled
Publish Job / Run Accuracy Tests (push) Has been cancelled
[V1 Loader] support weight_only (#3413)
* support wint4/wint8

* delete smoe case

* update ci

* print log
2025-08-23 13:13:41 +08:00

75 lines
2.6 KiB
Python

"""
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
#
# 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 paddle
from paddle import nn
from paddleformers.utils.log import logger
from fastdeploy.config import FDConfig, LoadConfig, ModelConfig
from fastdeploy.model_executor.load_weight_utils import (
fast_weights_iterator,
get_all_safetensors,
measure_time,
)
from fastdeploy.model_executor.model_loader.base_loader import BaseModelLoader
from fastdeploy.model_executor.models.model_base import ModelRegistry
from fastdeploy.platforms import current_platform
class DefaultModelLoaderV1(BaseModelLoader):
"""ModelLoader that can load registered models"""
def __init__(self, load_config: LoadConfig):
super().__init__(load_config)
def download_model(self, model_config: ModelConfig) -> None:
pass
def clean_memory_fragments(self) -> None:
"""clean_memory_fragments"""
if current_platform.is_cuda():
paddle.device.cuda.empty_cache()
paddle.device.synchronize()
@measure_time
def load_weights(self, model, fd_config: FDConfig) -> None:
_, safetensor_files = get_all_safetensors(fd_config.model_config.model)
weights_iterator = fast_weights_iterator(safetensor_files)
model.load_weights(weights_iterator)
self.clean_memory_fragments()
def load_model(self, fd_config: FDConfig) -> nn.Layer:
architectures = fd_config.model_config.architectures[0]
logger.info(f"Starting to load model {architectures}")
context = paddle.LazyGuard()
if fd_config.load_config.dynamic_load_weight:
# register rl model
import fastdeploy.rl # noqa
architectures = architectures + "RL"
with context:
model_cls = ModelRegistry.get_class(architectures)
model = model_cls(fd_config)
model.eval()
# RL model not need set_state_dict
if fd_config.load_config.dynamic_load_weight:
return model
self.load_weights(model, fd_config)
return model