444 lines
18 KiB
Python
444 lines
18 KiB
Python
#
|
|
# Copyright (c) 2025 Huawei Technologies Co., Ltd. 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 gc
|
|
import json
|
|
import time
|
|
from copy import deepcopy
|
|
|
|
import torch
|
|
from torch import nn
|
|
from vllm.config import LoadConfig, ModelConfig, VllmConfig
|
|
from vllm.logger import logger
|
|
from vllm.model_executor.model_loader import register_model_loader
|
|
from vllm.model_executor.model_loader.base_loader import BaseModelLoader
|
|
from vllm.model_executor.model_loader.default_loader import DefaultModelLoader
|
|
from vllm.model_executor.model_loader.utils import initialize_model, process_weights_after_loading
|
|
from vllm.utils.torch_utils import set_default_torch_dtype
|
|
|
|
from .interaction.elastic import ElasticServer
|
|
from .load import elastic_load
|
|
from .utils import find_free_port, is_valid_path_prefix
|
|
|
|
DRAFT_PORT_OFFSET = 10000
|
|
|
|
try:
|
|
# Older vLLM versions may not expose the current-config accessor.
|
|
from vllm.config import get_current_vllm_config
|
|
except ImportError:
|
|
get_current_vllm_config = None
|
|
|
|
|
|
@register_model_loader("netloader")
|
|
class ModelNetLoaderElastic(BaseModelLoader):
|
|
"""
|
|
A model loader that uses elastic loading for loading weights.
|
|
"""
|
|
|
|
source: list[dict] | None
|
|
model_path: str | None
|
|
listen_port: int | None
|
|
int8_cache: str
|
|
int8_cache_name: list[str] | None
|
|
output_prefix: str | None
|
|
|
|
def __init__(self, load_config: LoadConfig):
|
|
"""
|
|
Initializes the ModelNetLoaderElastic with configuration.
|
|
|
|
Parameters:
|
|
- load_config: Configuration for loading the model.
|
|
"""
|
|
super().__init__(load_config)
|
|
|
|
config = None
|
|
|
|
# Try to read config file at first
|
|
extra = load_config.model_loader_extra_config
|
|
|
|
if extra is not None and not isinstance(extra, dict):
|
|
err_msg = "NetLoader requires --model-loader-extra-config to be a JSON object."
|
|
logger.error(err_msg)
|
|
raise RuntimeError(err_msg)
|
|
|
|
if extra and "CONFIG_FILE" in extra:
|
|
try:
|
|
logger.info("Reading configs in file %s ...", load_config.model_loader_extra_config["CONFIG_FILE"])
|
|
with open(extra["CONFIG_FILE"]) as f:
|
|
config = json.load(f)
|
|
except FileNotFoundError:
|
|
logger.error("CONFIG_FILE not found")
|
|
except json.JSONDecodeError:
|
|
logger.error("CONFIG_FILE is not a valid JSON file")
|
|
except Exception as e:
|
|
logger.error("Unexpected error while reading CONFIG_FILE: %s", e)
|
|
|
|
if config is None and extra:
|
|
logger.info("Reading configs in model_loader_extra_config ...")
|
|
config = extra
|
|
config = config or {}
|
|
|
|
for key, attr, checker, caster, default in [
|
|
("SOURCE", "source", lambda v: isinstance(v, list), lambda v: v, None),
|
|
("MODEL", "model_path", lambda v: isinstance(v, str), lambda v: v, None),
|
|
(
|
|
"LISTEN_PORT",
|
|
"listen_port",
|
|
lambda v: isinstance(v, int) or (isinstance(v, str) and v.isdigit()),
|
|
lambda v: int(v),
|
|
None,
|
|
),
|
|
(
|
|
"INT8_CACHE",
|
|
"int8_cache",
|
|
lambda v: isinstance(v, str) and v.lower() in ["hbm", "dram", "no"],
|
|
lambda v: v.lower(),
|
|
"no",
|
|
),
|
|
("INT8_CACHE_NAME", "int8_cache_name", lambda v: isinstance(v, list), lambda v: v, None),
|
|
(
|
|
"OUTPUT_PREFIX",
|
|
"output_prefix",
|
|
lambda v: isinstance(v, str) and is_valid_path_prefix(v),
|
|
lambda v: v,
|
|
None,
|
|
),
|
|
]:
|
|
v = config.get(key, default)
|
|
if not checker(v):
|
|
v = default
|
|
else:
|
|
v = caster(v)
|
|
setattr(self, attr, v)
|
|
|
|
logger.info(
|
|
"Initializing elastic Netloader with config: "
|
|
"MODEL=%s, LISTEN_PORT=%s,"
|
|
"SOURCE=%s, INT8_CACHE=%s, INT8_CACHE_NAME=%s,"
|
|
"OUTPUT_PREFIX=%s)",
|
|
self.model_path,
|
|
self.listen_port,
|
|
self.source,
|
|
self.int8_cache,
|
|
self.int8_cache_name,
|
|
self.output_prefix,
|
|
)
|
|
|
|
@staticmethod
|
|
def _is_draft_model(model_config: ModelConfig) -> bool:
|
|
"""Check whether the model_config corresponds to a draft model for speculative decoding."""
|
|
return getattr(model_config, "runner_type", None) == "draft"
|
|
|
|
@staticmethod
|
|
def _sync_target_netloader_before_draft(vllm_config: VllmConfig) -> None:
|
|
if getattr(vllm_config, "speculative_config", None) is None:
|
|
return
|
|
if not torch.distributed.is_available() or not torch.distributed.is_initialized():
|
|
return
|
|
|
|
logger.info("Waiting for all target netloader ranks before loading draft model")
|
|
barrier_start = time.perf_counter()
|
|
torch.distributed.barrier()
|
|
logger.info(
|
|
"Target netloader barrier before draft model time: %s",
|
|
time.perf_counter() - barrier_start,
|
|
)
|
|
|
|
@staticmethod
|
|
def _get_static_forward_context(vllm_config: VllmConfig):
|
|
compilation_config = getattr(vllm_config, "compilation_config", None)
|
|
static_forward_context = getattr(compilation_config, "static_forward_context", None)
|
|
if static_forward_context is None or not hasattr(static_forward_context, "clear"):
|
|
return None
|
|
return static_forward_context
|
|
|
|
@staticmethod
|
|
def _clear_static_forward_context(vllm_config: VllmConfig) -> None:
|
|
"""Clear static layer registrations before rebuilding the model on fallback."""
|
|
candidates = [("vllm_config", vllm_config)]
|
|
if get_current_vllm_config is not None:
|
|
try:
|
|
candidates.append(("current_vllm_config", get_current_vllm_config()))
|
|
except Exception as e:
|
|
logger.debug("Failed to get current vLLM config while clearing static context: %s", e)
|
|
|
|
cleared_contexts = []
|
|
seen_context_ids = set()
|
|
for source, config in candidates:
|
|
static_forward_context = ModelNetLoaderElastic._get_static_forward_context(config)
|
|
if static_forward_context is None:
|
|
continue
|
|
|
|
context_id = id(static_forward_context)
|
|
if context_id in seen_context_ids:
|
|
continue
|
|
seen_context_ids.add(context_id)
|
|
|
|
try:
|
|
context_size = str(len(static_forward_context))
|
|
except TypeError:
|
|
context_size = "unknown"
|
|
static_forward_context.clear()
|
|
cleared_contexts.append(f"{source}:{context_size}")
|
|
|
|
if cleared_contexts:
|
|
logger.info("Cleared static_forward_context before fallback: %s", cleared_contexts)
|
|
|
|
def load_model(self, vllm_config: VllmConfig, model_config: ModelConfig, prefix: str = "") -> nn.Module:
|
|
"""
|
|
Loads the model using the specified configuration.
|
|
|
|
Parameters:
|
|
- vllm_config: Configuration for the VLLM.
|
|
- model_config: Configuration for the model.
|
|
- prefix: Module prefix for pipeline parallelism (e.g., "model.layers.0.").
|
|
|
|
Returns:
|
|
- The loaded model.
|
|
"""
|
|
|
|
device_config = vllm_config.device_config
|
|
parallel_config = vllm_config.parallel_config
|
|
|
|
need_process_weights_after_loading = False
|
|
|
|
if self.model_path is None:
|
|
self.model_path = model_config.model
|
|
logger.info("model_path is set to %s", self.model_path)
|
|
|
|
device_id = torch.distributed.get_rank()
|
|
is_draft = self._is_draft_model(model_config)
|
|
|
|
if is_draft:
|
|
logger.info("Loading draft model via netloader, model_path: %s", model_config.model)
|
|
else:
|
|
logger.info("Loading target model via netloader, model_path: %s", model_config.model)
|
|
|
|
if (
|
|
self.source is None
|
|
or not isinstance(self.source, list)
|
|
or device_id
|
|
not in [
|
|
one_device["device_id"]
|
|
for one_device in self.source
|
|
if isinstance(one_device, dict) and "device_id" in one_device
|
|
]
|
|
):
|
|
logger.warning("Did not get valid source info, use DefaultModelLoader")
|
|
model, need_process_weights_after_loading = self.revert_to_default(
|
|
model_config, vllm_config, device_config, prefix
|
|
)
|
|
|
|
else:
|
|
target_device = torch.device(device_config.device)
|
|
|
|
_quant_config = getattr(vllm_config, "quant_config", None)
|
|
_quant_config = deepcopy(_quant_config) if _quant_config is not None else None
|
|
model_config_backup = deepcopy(model_config)
|
|
|
|
with set_default_torch_dtype(model_config.dtype):
|
|
with target_device:
|
|
model = initialize_model(vllm_config=vllm_config, model_config=model_config, prefix=prefix)
|
|
|
|
start_elastic_load = time.perf_counter()
|
|
|
|
sources = self.source
|
|
if is_draft:
|
|
sources = [
|
|
{
|
|
"device_id": s["device_id"],
|
|
"sources": [
|
|
f"{parts[0]}:{int(parts[1]) + DRAFT_PORT_OFFSET}"
|
|
for addr in s.get("sources", [])
|
|
if isinstance(addr, str)
|
|
and len(parts := addr.rsplit(":", 1)) == 2
|
|
and parts[1].isdigit()
|
|
],
|
|
}
|
|
for s in self.source
|
|
if isinstance(s, dict) and "device_id" in s
|
|
]
|
|
|
|
model = elastic_load(
|
|
model=model,
|
|
device_id=device_id,
|
|
model_path=model_config.model,
|
|
sources=sources,
|
|
tp=parallel_config.tensor_parallel_size,
|
|
pp=parallel_config.pipeline_parallel_size,
|
|
group_name="netloader_draft" if is_draft else "netloader",
|
|
)
|
|
end_elastic_load = time.perf_counter()
|
|
logger.info("Elastic load time: %s, rank: %s", end_elastic_load - start_elastic_load, device_id)
|
|
need_process_weights_after_loading = True
|
|
|
|
if model is None:
|
|
logger.warning("Netloader elastic loading fails, use load format DefaultModelLoader")
|
|
|
|
if hasattr(vllm_config, "quant_config"):
|
|
vllm_config.quant_config = _quant_config
|
|
model_config = model_config_backup
|
|
|
|
del model
|
|
gc.collect()
|
|
if device_config.device_type == "npu":
|
|
logger.info("Empty NPU cache")
|
|
torch.npu.empty_cache()
|
|
elif device_config.device_type == "cuda":
|
|
logger.info("Empty CUDA cache")
|
|
torch.cuda.empty_cache()
|
|
|
|
# Clear registrations from the failed initialize_model
|
|
self._clear_static_forward_context(vllm_config)
|
|
|
|
model, need_process_weights_after_loading = self.revert_to_default(
|
|
model_config, vllm_config, device_config, prefix
|
|
)
|
|
|
|
start_elastic_server = time.perf_counter()
|
|
# start elastic server
|
|
if model is not None and (
|
|
(self.listen_port and self.listen_port in range(1024, 65535)) or (self.listen_port is None)
|
|
):
|
|
from vllm.utils.network_utils import get_ip
|
|
|
|
driver_ip = get_ip()
|
|
|
|
if driver_ip == "0.0.0.0":
|
|
logger.error("Driver IP is not set, skip to start Netloader server")
|
|
else:
|
|
if self.listen_port is None:
|
|
listen_port = find_free_port()
|
|
else:
|
|
listen_port = self.listen_port + device_id
|
|
if is_draft:
|
|
listen_port += DRAFT_PORT_OFFSET
|
|
self.listen_port = listen_port
|
|
|
|
group_name = "netloader_draft" if is_draft else "netloader"
|
|
|
|
logger.info(
|
|
"Start elastic Netloader server, rank: %s, listen port: %s:%s, group: %s",
|
|
device_id,
|
|
driver_ip,
|
|
listen_port,
|
|
group_name,
|
|
)
|
|
|
|
if self.output_prefix is not None and not is_draft:
|
|
try:
|
|
with open(self.output_prefix + str(device_id) + ".txt", "w") as file:
|
|
file.write(f"{driver_ip}:{listen_port}")
|
|
logger.info(
|
|
"Successfully wrote server address to file: %s", self.output_prefix + str(device_id)
|
|
)
|
|
except FileNotFoundError:
|
|
logger.error("File path %s does not exist.", self.output_prefix + str(device_id))
|
|
except PermissionError:
|
|
logger.error("No permission to write to file %s.", self.output_prefix + str(device_id))
|
|
except OSError as e:
|
|
logger.error(
|
|
"I/O error occurred while writing to file %s: %s", self.output_prefix + str(device_id), e
|
|
)
|
|
except Exception as e:
|
|
logger.error("Unknown error: %s", e)
|
|
|
|
try:
|
|
server_int8_cache = "hbm" if is_draft and self.int8_cache != "no" else self.int8_cache
|
|
elastic_server = ElasticServer(
|
|
driver_ip,
|
|
listen_port,
|
|
model,
|
|
device_id,
|
|
model_config.model,
|
|
parallel_config.tensor_parallel_size,
|
|
parallel_config.pipeline_parallel_size,
|
|
server_int8_cache,
|
|
self.int8_cache_name,
|
|
group_name=group_name,
|
|
)
|
|
elastic_server.start()
|
|
if is_draft:
|
|
self._draft_elastic_server = elastic_server
|
|
else:
|
|
self._target_elastic_server = elastic_server
|
|
except Exception as e:
|
|
logger.error("Failed to start Netloader server for rank: %s, details: %s", device_id, e)
|
|
else:
|
|
logger.info("Skip to start Netloader server")
|
|
|
|
end_elastic_server = time.perf_counter()
|
|
logger.info("Elastic server start time: %s, rank: %s", end_elastic_server - start_elastic_server, device_id)
|
|
|
|
if need_process_weights_after_loading:
|
|
process_weights_after_loading(model, model_config, torch.device(device_config.device))
|
|
|
|
if not is_draft:
|
|
self._sync_target_netloader_before_draft(vllm_config)
|
|
|
|
if model is None:
|
|
logger.error("NetLoader elastic loads model fails")
|
|
raise RuntimeError("NetLoader elastic loads model fails")
|
|
|
|
return model.eval()
|
|
|
|
def revert_to_default(self, model_config, vllm_config, device_config, prefix: str = "") -> tuple[nn.Module, bool]:
|
|
"""
|
|
Reverts to the default model loading logic when elastic loading fails or is not applicable.
|
|
|
|
This method resets the loader's extra config and load format to defaults,
|
|
then delegates model loading to a DefaultModelLoader.
|
|
If quantization is enabled, it will load the model and then run the
|
|
processing of weights (i.e. applying quantization adjustments) before returning.
|
|
|
|
Parameters:
|
|
- model_config: Configuration describing model architecture, quantization, etc.
|
|
- vllm_config: Configuration for vLLM (device, parallelism, dtype, etc).
|
|
- device_config: Configuration for the target device (device type, device id, etc).
|
|
- prefix: Module prefix for pipeline parallelism.
|
|
|
|
Returns:
|
|
- A tuple (model, need_process_weights_after_loading):
|
|
* model: The loaded `nn.Module` under default loading logic.
|
|
* need_process_weights_after_loading: A boolean flag indicating whether
|
|
weights post-processing (e.g. quantization adjustments) still needs to be applied.
|
|
"""
|
|
load_config = deepcopy(self.load_config)
|
|
load_config.model_loader_extra_config = {}
|
|
load_config.load_format = "auto"
|
|
default_model_loader = DefaultModelLoader(load_config)
|
|
|
|
if model_config.quantization is None:
|
|
model = default_model_loader.load_model(vllm_config=vllm_config, model_config=model_config, prefix=prefix)
|
|
need_process_weights_after_loading = False
|
|
else:
|
|
logger.warning("Quantization is set, netloader use DefaultModelLoader with process_weights_after_loading ")
|
|
need_process_weights_after_loading = True
|
|
target_device = torch.device(device_config.device)
|
|
with set_default_torch_dtype(model_config.dtype):
|
|
with target_device:
|
|
model = initialize_model(vllm_config=vllm_config, model_config=model_config, prefix=prefix)
|
|
default_model_loader.load_weights(model, model_config)
|
|
model = model.eval()
|
|
|
|
return model, need_process_weights_after_loading
|
|
|
|
def download_model(self, model_config: ModelConfig) -> None:
|
|
pass
|
|
|
|
def load_weights(self, model: nn.Module, model_config: ModelConfig) -> None:
|
|
pass
|