20231088/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[Frontend] [Core] perf: Automatically detect vLLM-tensorized model, update tensorizer to version 2.9.0 (#4208)

Browse Source
This commit is contained in:
Sanger Steel 2024-05-13 17:57:07 -04:00 committed by GitHub
parent 0fca3cdcf2
commit 8bc68e198c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
10 changed files with 259 additions and 523 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
.buildkite/test-pipeline.yaml
View File

@ -60,11 +60,13 @@ steps:
mirror_hardwares: [amd]
commands:
# install aws cli for llava_example.py
- pip install awscli
# install tensorizer for tensorize_vllm_model.py
- pip install awscli tensorizer
- python3 offline_inference.py
- python3 offline_inference_with_prefix.py
- python3 llm_engine_example.py
- python3 llava_example.py
- python3 tensorize_vllm_model.py --model facebook/opt-125m serialize --serialized-directory /tmp/ --suffix v1 && python3 tensorize_vllm_model.py --model facebook/opt-125m deserialize --path-to-tensors /tmp/vllm/facebook/opt-125m/v1/model.tensors
- label: Kernels Test %N
command: pytest -v -s kernels --shard-id=$$BUILDKITE_PARALLEL_JOB --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT

200
examples/tensorize_vllm_model.py
View File

@ -1,23 +1,20 @@
import argparse
import dataclasses
import json
import os
import time
import uuid
from functools import partial
from typing import Type
import torch
import torch.nn as nn
from tensorizer import (DecryptionParams, EncryptionParams, TensorDeserializer,
TensorSerializer, stream_io)
from tensorizer.utils import convert_bytes, get_mem_usage, no_init_or_tensor
from transformers import AutoConfig, PretrainedConfig
from tensorizer import stream_io
from vllm.distributed import initialize_model_parallel
from vllm import LLM
from vllm.distributed import (init_distributed_environment,
initialize_model_parallel)
from vllm.engine.arg_utils import EngineArgs
from vllm.engine.llm_engine import LLMEngine
from vllm.model_executor.model_loader.tensorizer import TensorizerArgs
from vllm.model_executor.models import ModelRegistry
from vllm.model_executor.model_loader.tensorizer import (TensorizerArgs,
TensorizerConfig,
serialize_vllm_model)
# yapf conflicts with isort for this docstring
# yapf: disable
@ -27,25 +24,25 @@ deserialize vLLM models. These models can be loaded using tensorizer
to the GPU extremely quickly over an HTTP/HTTPS endpoint, an S3 endpoint,
or locally. Tensor encryption and decryption is also supported, although
libsodium must be installed to use it. Install vllm with tensorizer support
using `pip install vllm[tensorizer]`.
using `pip install vllm[tensorizer]`. To learn more about tensorizer, visit
https://github.com/coreweave/tensorizer
To serialize a model, install vLLM from source, then run something
like this from the root level of this repository:
python -m examples.tensorize_vllm_model \
--model EleutherAI/gpt-j-6B \
--dtype float16 \
--model facebook/opt-125m \
serialize \
--serialized-directory s3://my-bucket/ \
--suffix vllm
--serialized-directory s3://my-bucket \
--suffix v1
Which downloads the model from HuggingFace, loads it into vLLM, serializes it,
and saves it to your S3 bucket. A local directory can also be used. This
assumes your S3 credentials are specified as environment variables
in the form of `S3_ACCESS_KEY_ID`, `S3_SECRET_ACCESS_KEY`, and `S3_ENDPOINT`.
To provide S3 credentials directly, you can provide `--s3-access-key-id` and
`--s3-secret-access-key`, as well as `--s3-endpoint` as CLI args to this
script.
in the form of `S3_ACCESS_KEY_ID`, `S3_SECRET_ACCESS_KEY`, and
`S3_ENDPOINT_URL`. To provide S3 credentials directly, you can provide
`--s3-access-key-id` and `--s3-secret-access-key`, as well as `--s3-endpoint`
as CLI args to this script.
You can also encrypt the model weights with a randomly-generated key by
providing a `--keyfile` argument.
@ -57,7 +54,7 @@ python -m examples.tensorize_vllm_model \
--model EleutherAI/gpt-j-6B \
--dtype float16 \
deserialize \
--path-to-tensors s3://my-bucket/vllm/EleutherAI/gpt-j-6B/vllm/model.tensors
--path-to-tensors s3://my-bucket/vllm/EleutherAI/gpt-j-6B/v1/model.tensors
Which downloads the model tensors from your S3 bucket and deserializes them.
@ -71,26 +68,30 @@ Or for deserializing:
`python -m examples.tensorize_vllm_model deserialize --help`.
Once a model is serialized, it can be used to load the model when running the
OpenAI inference client at `vllm/entrypoints/openai/api_server.py` by providing
the `--tensorizer-uri` CLI argument that is functionally the same as the
`--path-to-tensors` argument in this script, along with `--vllm-tensorized`, to
signify that the model to be deserialized is a vLLM model, rather than a
HuggingFace `PreTrainedModel`, which can also be deserialized using tensorizer
in the same inference server, albeit without the speed optimizations. To
deserialize an encrypted file, the `--encryption-keyfile` argument can be used
to provide the path to the keyfile used to encrypt the model weights. For
information on all the arguments that can be used to configure tensorizer's
deserialization, check out the tensorizer options argument group in the
`vllm/entrypoints/openai/api_server.py` script with `--help`.
Tensorizer can also be invoked with the `LLM` class directly to load models:
Once a model is serialized, tensorizer can be invoked with the `LLM` class
directly to load models:
llm = LLM(model="facebook/opt-125m",
load_format="tensorizer",
tensorizer_uri=path_to_opt_tensors,
num_readers=3,
vllm_tensorized=True)
model_loader_extra_config=TensorizerConfig(
tensorizer_uri = path_to_tensors,
num_readers=3,
)
)
A serialized model can be used during model loading for the vLLM OpenAI
inference server. `model_loader_extra_config` is exposed as the CLI arg
`--model-loader-extra-config`, and accepts a JSON string literal of the
TensorizerConfig arguments desired.
In order to see all of the available arguments usable to configure
loading with tensorizer that are given to `TensorizerConfig`, run:
`python -m examples.tensorize_vllm_model deserialize --help`
under the `tensorizer options` section. These can also be used for
deserialization in this example script, although `--tensorizer-uri` and
`--path-to-tensors` are functionally the same in this case.
"""
@ -158,95 +159,35 @@ def parse_args():
help=("Path to a binary key to use to decrypt the model weights,"
" if the model was serialized with encryption"))
TensorizerArgs.add_cli_args(deserialize_parser)
return parser.parse_args()
def make_model_contiguous(model):
# Ensure tensors are saved in memory contiguously
for param in model.parameters():
param.data = param.data.contiguous()
def _get_vllm_model_architecture(config: PretrainedConfig) -> Type[nn.Module]:
architectures = getattr(config, "architectures", [])
for arch in architectures:
model_cls = ModelRegistry.load_model_cls(arch)
if model_cls is not None:
return model_cls
raise ValueError(
f"Model architectures {architectures} are not supported for now. "
f"Supported architectures: {ModelRegistry.get_supported_archs()}")
def serialize():
eng_args_dict = {f.name: getattr(args, f.name) for f in
dataclasses.fields(EngineArgs)}
engine_args = EngineArgs.from_cli_args(argparse.Namespace(**eng_args_dict))
engine = LLMEngine.from_engine_args(engine_args)
model = (engine.model_executor.driver_worker.
model_runner.model)
encryption_params = EncryptionParams.random() if keyfile else None
if keyfile:
with _write_stream(keyfile) as stream:
stream.write(encryption_params.key)
with _write_stream(model_path) as stream:
serializer = TensorSerializer(stream, encryption=encryption_params)
serializer.write_module(model)
serializer.close()
print("Serialization complete. Model tensors saved to", model_path)
if keyfile:
print("Key saved to", keyfile)
def deserialize():
config = AutoConfig.from_pretrained(model_ref)
with no_init_or_tensor():
model_class = _get_vllm_model_architecture(config)
model = model_class(config)
before_mem = get_mem_usage()
start = time.time()
if keyfile:
with _read_stream(keyfile) as stream:
key = stream.read()
decryption_params = DecryptionParams.from_key(key)
tensorizer_args.deserializer_params['encryption'] = \
decryption_params
with (_read_stream(model_path)) as stream, TensorDeserializer(
stream, **tensorizer_args.deserializer_params) as deserializer:
deserializer.load_into_module(model)
end = time.time()
# Brag about how fast we are.
total_bytes_str = convert_bytes(deserializer.total_tensor_bytes)
duration = end - start
per_second = convert_bytes(deserializer.total_tensor_bytes / duration)
after_mem = get_mem_usage()
print(
f"Deserialized {total_bytes_str} in {end - start:0.2f}s, {per_second}/s"
llm = LLM(model=args.model,
load_format="tensorizer",
model_loader_extra_config=tensorizer_config
)
print(f"Memory usage before: {before_mem}")
print(f"Memory usage after: {after_mem}")
return llm
return model
args = parse_args()
s3_access_key_id = (args.s3_access_key_id or os.environ.get("S3_ACCESS_KEY_ID")
or None)
s3_secret_access_key = (args.s3_secret_access_key
or os.environ.get("S3_SECRET_ACCESS_KEY") or None)
s3_access_key_id = (getattr(args, 's3_access_key_id', None)
or os.environ.get("S3_ACCESS_KEY_ID", None))
s3_secret_access_key = (getattr(args, 's3_secret_access_key', None)
or os.environ.get("S3_SECRET_ACCESS_KEY", None))
s3_endpoint = (getattr(args, 's3_endpoint', None)
or os.environ.get("S3_ENDPOINT_URL", None))
s3_endpoint = (args.s3_endpoint or os.environ.get("S3_ENDPOINT_URL") or None)
credentials = {
"s3_access_key_id": s3_access_key_id,
"s3_secret_access_key": s3_secret_access_key,
"s3_endpoint": s3_endpoint
}
_read_stream, _write_stream = (partial(
stream_io.open_stream,
@ -263,20 +204,41 @@ model_name = model_ref.split("/")[1]
os.environ["MASTER_ADDR"] = "127.0.0.1"
os.environ["MASTER_PORT"] = "8080"
torch.distributed.init_process_group(world_size=1, rank=0)
init_distributed_environment(world_size=1, rank=0, local_rank=0)
initialize_model_parallel()
keyfile = args.keyfile if args.keyfile else None
if args.model_loader_extra_config:
config = json.loads(args.model_loader_extra_config)
tensorizer_args = TensorizerConfig(**config)._construct_tensorizer_args()
tensorizer_args.tensorizer_uri = args.path_to_tensors
else:
tensorizer_args = None
if args.command == "serialize":
eng_args_dict = {f.name: getattr(args, f.name) for f in
dataclasses.fields(EngineArgs)}
engine_args = EngineArgs.from_cli_args(argparse.Namespace(**eng_args_dict))
engine = LLMEngine.from_engine_args(engine_args)
input_dir = args.serialized_directory.rstrip('/')
suffix = args.suffix if args.suffix else uuid.uuid4().hex
base_path = f"{input_dir}/vllm/{model_ref}/{suffix}"
model_path = f"{base_path}/model.tensors"
serialize()
tensorizer_config = TensorizerConfig(
tensorizer_uri=model_path,
**credentials)
serialize_vllm_model(engine, tensorizer_config, keyfile)
elif args.command == "deserialize":
tensorizer_args = TensorizerArgs.from_cli_args(args)
model_path = args.path_to_tensors
if not tensorizer_args:
tensorizer_config = TensorizerConfig(
tensorizer_uri=args.path_to_tensors,
encryption_keyfile = keyfile,
**credentials
)
deserialize()
else:
raise ValueError("Either serialize or deserialize must be specified.")

2
requirements-dev.txt
View File

@ -14,7 +14,7 @@ types-setuptools
# testing
pytest
tensorizer==2.9.0
tensorizer>=2.9.0
pytest-forked
pytest-asyncio
pytest-rerunfailures

2
setup.py
View File

@ -426,7 +426,7 @@ setup(
install_requires=get_requirements(),
ext_modules=ext_modules,
extras_require={
"tensorizer": ["tensorizer==2.9.0"],
"tensorizer": ["tensorizer>=2.9.0"],
},
cmdclass={"build_ext": cmake_build_ext} if not _is_neuron() else {},
package_data=package_data,

245
tests/tensorizer_loader/tensorize_vllm_model_for_testing.py
View File

@ -1,245 +0,0 @@
import argparse
import dataclasses
import os
import time
import uuid
from functools import partial
from typing import Type
import torch.nn as nn
from tensorizer import (DecryptionParams, EncryptionParams, TensorDeserializer,
TensorSerializer, stream_io)
from tensorizer.utils import convert_bytes, get_mem_usage, no_init_or_tensor
from transformers import AutoConfig, PretrainedConfig
from vllm.distributed import (init_distributed_environment,
initialize_model_parallel)
from vllm.engine.arg_utils import EngineArgs
from vllm.engine.llm_engine import LLMEngine
from vllm.model_executor.model_loader.tensorizer import TensorizerArgs
from vllm.model_executor.models import ModelRegistry
# yapf conflicts with isort for this docstring
# yapf: disable
"""
tensorize_vllm_model.py is a script that can be used to serialize and
deserialize vLLM models. These models can be loaded using tensorizer directly
to the GPU extremely quickly. Tensor encryption and decryption is also
supported, although libsodium must be installed to use it. Install
vllm with tensorizer support using `pip install vllm[tensorizer]`.
To serialize a model, you can run something like this:
python tensorize_vllm_model.py \
--model EleutherAI/gpt-j-6B \
--dtype float16 \
serialize \
--serialized-directory s3://my-bucket/ \
--suffix vllm
Which downloads the model from HuggingFace, loads it into vLLM, serializes it,
and saves it to your S3 bucket. A local directory can also be used.
You can also encrypt the model weights with a randomly-generated key by
providing a `--keyfile` argument.
To deserialize a model, you can run something like this:
python tensorize_vllm_model.py \
--model EleutherAI/gpt-j-6B \
--dtype float16 \
deserialize \
--path-to-tensors s3://my-bucket/vllm/EleutherAI/gpt-j-6B/vllm/model.tensors
Which downloads the model tensors from your S3 bucket and deserializes them.
To provide S3 credentials, you can provide `--s3-access-key-id` and
`--s3-secret-access-key`, as well as `--s3-endpoint` as CLI args to this script,
the OpenAI entrypoint, as arguments for LLM(), or as environment variables
in the form of `S3_ACCESS_KEY_ID`, `S3_SECRET_ACCESS_KEY`, and `S3_ENDPOINT`.
You can also provide a `--keyfile` argument to decrypt the model weights if
they were serialized with encryption.
For more information on the available arguments, run
`python tensorize_vllm_model.py --help`.
"""
def parse_args():
parser = argparse.ArgumentParser(
description="An example script that can be used to serialize and "
"deserialize vLLM models. These models "
"can be loaded using tensorizer directly to the GPU "
"extremely quickly. Tensor encryption and decryption is "
"also supported, although libsodium must be installed to "
"use it.")
parser = TensorizerArgs.add_cli_args(EngineArgs.add_cli_args(parser))
subparsers = parser.add_subparsers(dest='command')
serialize_parser = subparsers.add_parser(
'serialize', help="Serialize a model to `--serialized-directory`")
serialize_parser.add_argument(
"--suffix",
type=str,
required=False,
help=(
"The suffix to append to the serialized model directory, which is "
"used to construct the location of the serialized model tensors, "
"e.g. if `--serialized-directory` is `s3://my-bucket/` and "
"`--suffix` is `v1`, the serialized model tensors will be "
"saved to "
"`s3://my-bucket/vllm/EleutherAI/gpt-j-6B/v1/model.tensors`. "
"If none is provided, a random UUID will be used."))
serialize_parser.add_argument(
"--serialized-directory",
type=str,
required=True)
serialize_parser.add_argument(
"--keyfile",
type=str,
required=False,
help=("Encrypt the model weights with a randomly-generated binary key,"
" and save the key at this path"))
deserialize_parser = subparsers.add_parser(
'deserialize',
help=("Deserialize a model from `--path-to-tensors`"
" to verify it can be loaded and used."))
deserialize_parser.add_argument(
"--path-to-tensors",
type=str,
required=True,
help="The local path or S3 URI to the model tensors to deserialize. ")
deserialize_parser.add_argument(
"--keyfile",
type=str,
required=False,
help=("Path to a binary key to use to decrypt the model weights,"
" if the model was serialized with encryption"))
return parser.parse_args()
def make_model_contiguous(model):
# Ensure tensors are saved in memory contiguously
for param in model.parameters():
param.data = param.data.contiguous()
def _get_vllm_model_architecture(config: PretrainedConfig) -> Type[nn.Module]:
architectures = getattr(config, "architectures", [])
for arch in architectures:
model_cls = ModelRegistry.load_model_cls(arch)
if model_cls is not None:
return model_cls
raise ValueError(
f"Model architectures {architectures} are not supported for now. "
f"Supported architectures: {ModelRegistry.get_supported_archs()}")
def serialize():
eng_args_dict = {f.name: getattr(args, f.name) for f in
dataclasses.fields(EngineArgs)}
engine_args = EngineArgs.from_cli_args(argparse.Namespace(**eng_args_dict))
engine = LLMEngine.from_engine_args(engine_args)
model = (engine.model_executor.driver_worker.
model_runner.model)
encryption_params = EncryptionParams.random() if keyfile else None
if keyfile:
with _write_stream(keyfile) as stream:
stream.write(encryption_params.key)
with _write_stream(model_path) as stream:
serializer = TensorSerializer(stream, encryption=encryption_params)
serializer.write_module(model)
serializer.close()
print("Serialization complete. Model tensors saved to", model_path)
if keyfile:
print("Key saved to", keyfile)
def deserialize():
config = AutoConfig.from_pretrained(model_ref)
with no_init_or_tensor():
model_class = _get_vllm_model_architecture(config)
model = model_class(config)
before_mem = get_mem_usage()
start = time.time()
if keyfile:
with _read_stream(keyfile) as stream:
key = stream.read()
decryption_params = DecryptionParams.from_key(key)
tensorizer_args.deserializer_params['encryption'] = \
decryption_params
with (_read_stream(model_path)) as stream, TensorDeserializer(
stream, **tensorizer_args.deserializer_params) as deserializer:
deserializer.load_into_module(model)
end = time.time()
# Brag about how fast we are.
total_bytes_str = convert_bytes(deserializer.total_tensor_bytes)
duration = end - start
per_second = convert_bytes(deserializer.total_tensor_bytes / duration)
after_mem = get_mem_usage()
print(
f"Deserialized {total_bytes_str} in {end - start:0.2f}s, {per_second}/s"
)
print(f"Memory usage before: {before_mem}")
print(f"Memory usage after: {after_mem}")
return model
args = parse_args()
s3_access_key_id = (args.s3_access_key_id or os.environ.get("S3_ACCESS_KEY_ID")
or None)
s3_secret_access_key = (args.s3_secret_access_key
or os.environ.get("S3_SECRET_ACCESS_KEY") or None)
s3_endpoint = (args.s3_endpoint or os.environ.get("S3_ENDPOINT_URL") or None)
_read_stream, _write_stream = (partial(
stream_io.open_stream,
mode=mode,
s3_access_key_id=s3_access_key_id,
s3_secret_access_key=s3_secret_access_key,
s3_endpoint=s3_endpoint,
) for mode in ("rb", "wb+"))
model_ref = args.model
model_name = model_ref.split("/")[1]
os.environ["MASTER_ADDR"] = "127.0.0.1"
os.environ["MASTER_PORT"] = "8080"
init_distributed_environment(world_size=1, rank=0, local_rank=0)
initialize_model_parallel()
keyfile = args.keyfile if args.keyfile else None
if args.command == "serialize":
input_dir = args.serialized_directory.rstrip('/')
suffix = args.suffix if args.suffix else uuid.uuid4().hex
base_path = f"{input_dir}/vllm/{model_ref}/{suffix}"
model_path = f"{base_path}/model.tensors"
serialize()
elif args.command == "deserialize":
tensorizer_args = TensorizerArgs.from_cli_args(args)
model_path = args.path_to_tensors
deserialize()
else:
raise ValueError("Either serialize or deserialize must be specified.")

189
tests/tensorizer_loader/test_tensorizer.py
View File

@ -10,12 +10,19 @@ import ray
import torch
from vllm import SamplingParams
from vllm.model_executor.model_loader.tensorizer import (
EncryptionParams, TensorizerConfig, TensorSerializer,
is_vllm_serialized_tensorizer, load_with_tensorizer, open_stream)
# yapf: disable
from vllm.model_executor.model_loader.tensorizer import (TensorizerConfig,
TensorSerializer,
is_vllm_tensorized,
load_with_tensorizer,
open_stream,
serialize_vllm_model)
from ..utils import ServerRunner
# yapf conflicts with isort for this docstring
prompts = [
"Hello, my name is",
"The president of the United States is",
@ -40,7 +47,7 @@ def is_curl_installed():
@pytest.fixture(autouse=True)
def tensorizer_config():
config = TensorizerConfig(tensorizer_uri="vllm", vllm_tensorized=True)
config = TensorizerConfig(tensorizer_uri="vllm")
return config
@ -59,47 +66,6 @@ def test_load_with_tensorizer(mock_agent, tensorizer_config):
assert result == mock_agent_instance.deserialize.return_value
def test_is_vllm_model_with_vllm_in_uri(tensorizer_config):
tensorizer_config.vllm_tensorized = True
result = is_vllm_serialized_tensorizer(tensorizer_config)
assert result is True
def test_is_vllm_model_without_vllm_in_uri(tensorizer_config):
tensorizer_config.vllm_tensorized = False
result = is_vllm_serialized_tensorizer(tensorizer_config)
assert result is False
def test_deserialized_vllm_model_has_same_outputs(vllm_runner, tmp_path):
vllm_model = vllm_runner(model_ref)
model_path = tmp_path / (model_ref + ".tensors")
outputs = vllm_model.generate(prompts, sampling_params)
model = (vllm_model.model.llm_engine.model_executor.driver_worker.
model_runner.model)
with open_stream(model_path, "wb+") as stream:
serializer = TensorSerializer(stream)
serializer.write_module(model)
del vllm_model, model
gc.collect()
torch.cuda.empty_cache()
loaded_vllm_model = vllm_runner(
model_ref,
load_format="tensorizer",
model_loader_extra_config=TensorizerConfig(tensorizer_uri=model_path,
num_readers=1,
vllm_tensorized=True),
)
deserialized_outputs = loaded_vllm_model.generate(prompts, sampling_params)
# Assumes SamplingParams being seeded ensures the outputs are deterministic
assert outputs == deserialized_outputs
@pytest.mark.skipif(not is_curl_installed(), reason="cURL is not installed")
def test_can_deserialize_s3(vllm_runner):
model_ref = "EleutherAI/pythia-1.4b"
@ -110,7 +76,6 @@ def test_can_deserialize_s3(vllm_runner):
model_loader_extra_config=TensorizerConfig(
tensorizer_uri=tensorized_path,
num_readers=1,
vllm_tensorized=False,
s3_endpoint="object.ord1.coreweave.com",
))
@ -126,29 +91,26 @@ def test_deserialized_encrypted_vllm_model_has_same_outputs(
model_path = tmp_path / (model_ref + ".tensors")
key_path = tmp_path / (model_ref + ".key")
outputs = vllm_model.generate(prompts, sampling_params)
model = (vllm_model.model.llm_engine.model_executor.driver_worker.
model_runner.model)
encryption_params = EncryptionParams.random()
with open_stream(model_path, "wb+") as stream:
serializer = TensorSerializer(stream, encryption=encryption_params)
serializer.write_module(model)
with open_stream(key_path, "wb+") as stream:
stream.write(encryption_params.key)
del vllm_model, model
config_for_serializing = TensorizerConfig(tensorizer_uri=model_path)
serialize_vllm_model(vllm_model.model.llm_engine,
config_for_serializing,
encryption_key_path=key_path)
del vllm_model
gc.collect()
torch.cuda.empty_cache()
loaded_vllm_model = vllm_runner(model_ref,
load_format="tensorizer",
model_loader_extra_config=TensorizerConfig(
tensorizer_uri=model_path,
encryption_keyfile=key_path,
num_readers=1,
vllm_tensorized=True))
config_for_deserializing = TensorizerConfig(tensorizer_uri=model_path,
encryption_keyfile=key_path)
loaded_vllm_model = vllm_runner(
model_ref,
load_format="tensorizer",
model_loader_extra_config=config_for_deserializing)
deserialized_outputs = loaded_vllm_model.generate(prompts, sampling_params)
# Assumes SamplingParams being seeded ensures the outputs are deterministic
assert outputs == deserialized_outputs
@ -169,7 +131,7 @@ def test_deserialized_hf_model_has_same_outputs(hf_runner, vllm_runner,
model_loader_extra_config=TensorizerConfig(
tensorizer_uri=model_path,
num_readers=1,
vllm_tensorized=False))
))
deserialized_outputs = loaded_hf_model.generate_greedy(
prompts, max_tokens=max_tokens)
@ -190,12 +152,11 @@ def test_vllm_model_can_load_with_lora(vllm_runner, tmp_path):
# Serialize model before deserializing and binding LoRA adapters
vllm_model = vllm_runner(model_ref, )
model_path = tmp_path / (model_ref + ".tensors")
model = (vllm_model.model.llm_engine.model_executor.driver_worker.
model_runner.model)
with open_stream(model_path, "wb+") as stream:
serializer = TensorSerializer(stream)
serializer.write_module(model)
del vllm_model, model
serialize_vllm_model(vllm_model.model.llm_engine,
TensorizerConfig(tensorizer_uri=model_path))
del vllm_model
gc.collect()
torch.cuda.empty_cache()
loaded_vllm_model = vllm_runner(
@ -204,7 +165,6 @@ def test_vllm_model_can_load_with_lora(vllm_runner, tmp_path):
model_loader_extra_config=TensorizerConfig(
tensorizer_uri=model_path,
num_readers=1,
vllm_tensorized=True,
),
enable_lora=True,
max_loras=1,
@ -220,58 +180,28 @@ def test_vllm_model_can_load_with_lora(vllm_runner, tmp_path):
def test_load_without_tensorizer_load_format(vllm_runner):
with pytest.raises(ValueError):
vllm_runner(model_ref,
model_loader_extra_config=TensorizerConfig(
tensorizer_uri="test", vllm_tensorized=False))
vllm_runner(
model_ref,
model_loader_extra_config=TensorizerConfig(tensorizer_uri="test"))
@pytest.mark.skipif(not is_curl_installed(), reason="cURL is not installed")
def test_tensorize_vllm_model(tmp_path):
# Test serialize command
serialize_args = [
"python3", tensorize_model_for_testing_script, "--model", model_ref,
"--dtype", "float16", "serialize", "--serialized-directory", tmp_path,
"--suffix", "tests"
]
result = subprocess.run(serialize_args, capture_output=True, text=True)
print(result.stdout) # Print the output of the serialize command
assert result.returncode == 0, (f"Serialize command failed with output:"
f"\n{result.stdout}\n{result.stderr}")
path_to_tensors = f"{tmp_path}/vllm/{model_ref}/tests/model.tensors"
# Test deserialize command
deserialize_args = [
"python3", tensorize_model_for_testing_script, "--model", model_ref,
"--dtype", "float16", "deserialize", "--path-to-tensors",
path_to_tensors
]
result = subprocess.run(deserialize_args, capture_output=True, text=True)
assert result.returncode == 0, (f"Deserialize command failed with output:"
f"\n{result.stdout}\n{result.stderr}")
@pytest.mark.skipif(not is_curl_installed(), reason="cURL is not installed")
def test_openai_apiserver_with_tensorizer(tmp_path):
def test_openai_apiserver_with_tensorizer(vllm_runner, tmp_path):
## Serialize model
serialize_args = [
"python3", tensorize_model_for_testing_script, "--model", model_ref,
"--dtype", "float16", "serialize", "--serialized-directory", tmp_path,
"--suffix", "tests"
]
result = subprocess.run(serialize_args, capture_output=True, text=True)
print(result.stdout) # Print the output of the serialize command
vllm_model = vllm_runner(model_ref, )
model_path = tmp_path / (model_ref + ".tensors")
assert result.returncode == 0, (f"Serialize command failed with output:"
f"\n{result.stdout}\n{result.stderr}")
serialize_vllm_model(vllm_model.model.llm_engine,
TensorizerConfig(tensorizer_uri=model_path))
path_to_tensors = f"{tmp_path}/vllm/{model_ref}/tests/model.tensors"
model_loader_extra_config = {
"tensorizer_uri": path_to_tensors,
"vllm_tensorized": True
"tensorizer_uri": str(model_path),
}
del vllm_model
gc.collect()
torch.cuda.empty_cache()
## Start OpenAI API server
openai_args = [
"--model", model_ref, "--dtype", "float16", "--load-format",
@ -304,10 +234,10 @@ def test_openai_apiserver_with_tensorizer(tmp_path):
def test_raise_value_error_on_invalid_load_format(vllm_runner):
with pytest.raises(ValueError):
vllm_runner(model_ref,
load_format="safetensors",
model_loader_extra_config=TensorizerConfig(
tensorizer_uri="test", vllm_tensorized=False))
vllm_runner(
model_ref,
load_format="safetensors",
model_loader_extra_config=TensorizerConfig(tensorizer_uri="test"))
def test_tensorizer_with_tp(vllm_runner):
@ -321,8 +251,29 @@ def test_tensorizer_with_tp(vllm_runner):
model_loader_extra_config=TensorizerConfig(
tensorizer_uri=tensorized_path,
num_readers=1,
vllm_tensorized=False,
s3_endpoint="object.ord1.coreweave.com",
),
tensor_parallel_size=2,
)
def test_vllm_tensorized_model_has_same_outputs(vllm_runner, tmp_path):
model_ref = "facebook/opt-125m"
model_path = tmp_path / (model_ref + ".tensors")
config = TensorizerConfig(tensorizer_uri=str(model_path))
vllm_model = vllm_runner(model_ref)
outputs = vllm_model.generate(prompts, sampling_params)
serialize_vllm_model(vllm_model.model.llm_engine, config)
assert is_vllm_tensorized(config)
del vllm_model
gc.collect()
torch.cuda.empty_cache()
loaded_vllm_model = vllm_runner(model_ref,
load_format="tensorizer",
model_loader_extra_config=config)
deserialized_outputs = loaded_vllm_model.generate(prompts, sampling_params)
assert outputs == deserialized_outputs

4
vllm/engine/arg_utils.py
View File

@ -167,8 +167,8 @@ class EngineArgs:
'* "dummy" will initialize the weights with random values, '
'which is mainly for profiling.\n'
'* "tensorizer" will load the weights using tensorizer from '
'CoreWeave which assumes tensorizer_uri is set to the location of '
'the serialized weights.')
'CoreWeave. See the Tensorize vLLM Model script in the Examples'
'section for more information.\n')
parser.add_argument(
'--dtype',
type=str,

2
vllm/envs.py
View File

@ -145,7 +145,7 @@ environment_variables: Dict[str, Callable[[], Any]] = {
# S3 access information, used for tensorizer to load model from S3
"S3_ACCESS_KEY_ID":
lambda: os.environ.get("S3_ACCESS_KEY", None),
lambda: os.environ.get("S3_ACCESS_KEY_ID", None),
"S3_SECRET_ACCESS_KEY":
lambda: os.environ.get("S3_SECRET_ACCESS_KEY", None),
"S3_ENDPOINT_URL":

28
vllm/model_executor/model_loader/loader.py
View File

@ -17,7 +17,7 @@ from vllm.logger import init_logger
from vllm.model_executor.layers.quantization.base_config import (
QuantizationConfig)
from vllm.model_executor.model_loader.tensorizer import (
TensorizerConfig, is_vllm_serialized_tensorizer, load_with_tensorizer,
TensorizerConfig, is_vllm_tensorized, load_with_tensorizer,
tensorizer_weights_iterator)
from vllm.model_executor.model_loader.utils import (get_model_architecture,
set_default_torch_dtype)
@ -291,7 +291,7 @@ class TensorizerLoader(BaseModelLoader):
tensorizer_args = self.tensorizer_config._construct_tensorizer_args()
return tensorizer_weights_iterator(tensorizer_args)
def _load_model_unserialized(
def _load_model_serialized_cpu(
self,
model_config: ModelConfig,
device_config: DeviceConfig,
@ -299,11 +299,12 @@ class TensorizerLoader(BaseModelLoader):
vision_language_config: Optional[VisionLanguageConfig],
cache_config: CacheConfig,
) -> nn.Module:
"""Load an unserialized model with tensorizer.
"""Load a serialized model with tensorizer to the CPU.
Unserialized here means "not serialized with tensorizer". This
should still be faster than default HuggingFace loading, but will
be slower than loading a tensorizer-serialized model.
This is only necessary when the model isn't vLLM-tensorized (see
examples/tensorize_vllm_model.py) This should still be faster than
default HuggingFace loading, but will be slower than loading a
vLLM-tensorized model.
"""
with set_default_torch_dtype(model_config.dtype):
with torch.device(device_config.device):
@ -324,8 +325,9 @@ class TensorizerLoader(BaseModelLoader):
) -> nn.Module:
"""Load a serialized model with tensorizer.
See the examples/tensorize_vllm_model.py example "
script for serializing vLLM models."""
Expects a vLLM-tensorized model. See the
examples/tensorize_vllm_model.py example script
for serializing vLLM models."""
with set_default_torch_dtype(model_config.dtype):
with torch.device(device_config.device):
model_class = get_model_architecture(model_config)[0]
@ -353,15 +355,15 @@ class TensorizerLoader(BaseModelLoader):
cache_config: CacheConfig) -> nn.Module:
self._verify_config(model_config, parallel_config)
if is_vllm_serialized_tensorizer(self.tensorizer_config):
if is_vllm_tensorized(self.tensorizer_config):
return self._load_model_serialized(model_config, device_config,
lora_config,
vision_language_config,
cache_config)
return self._load_model_unserialized(model_config, device_config,
lora_config,
vision_language_config,
cache_config)
return self._load_model_serialized_cpu(model_config, device_config,
lora_config,
vision_language_config,
cache_config)
def get_model_loader(load_config: LoadConfig) -> BaseModelLoader:

106
vllm/model_executor/model_loader/tensorizer.py
View File

@ -5,6 +5,7 @@ import os
import time
import typing
from dataclasses import dataclass
from functools import partial
from typing import Generator, Optional, Tuple, Type, Union
import torch
@ -13,6 +14,7 @@ from transformers import PretrainedConfig
import vllm.envs as envs
from vllm.config import ModelConfig, ParallelConfig
from vllm.engine.llm_engine import LLMEngine
from vllm.logger import init_logger
from vllm.model_executor.layers.quantization.base_config import (
QuantizationConfig)
@ -27,6 +29,11 @@ try:
from tensorizer.stream_io import open_stream
from tensorizer.utils import (convert_bytes, get_mem_usage,
no_init_or_tensor)
_read_stream, _write_stream = (partial(
open_stream,
mode=mode,
) for mode in ("rb", "wb+"))
except ImportError as e:
tensorizer_error_msg = str(e)
@ -43,7 +50,7 @@ logger = init_logger(__name__)
class TensorizerConfig:
tensorizer_uri: Union[io.BufferedIOBase, io.RawIOBase, typing.BinaryIO,
str, bytes, os.PathLike, int]
vllm_tensorized: bool
vllm_tensorized: Optional[bool] = False
verify_hash: Optional[bool] = False
num_readers: Optional[int] = None
encryption_keyfile: Optional[str] = None
@ -93,17 +100,11 @@ def load_with_tensorizer(tensorizer_config: TensorizerConfig,
return tensorizer.deserialize()
def is_vllm_serialized_tensorizer(tensorizer_config: TensorizerConfig) -> bool:
if tensorizer_config is None:
return False
return tensorizer_config.vllm_tensorized
@dataclass
class TensorizerArgs:
tensorizer_uri: Union[io.BufferedIOBase, io.RawIOBase, typing.BinaryIO,
str, bytes, os.PathLike, int]
vllm_tensorized: bool
vllm_tensorized: Optional[bool] = False
verify_hash: Optional[bool] = False
num_readers: Optional[int] = None
encryption_keyfile: Optional[str] = None
@ -121,7 +122,9 @@ class TensorizerArgs:
vLLM model. This is used to determine the behavior of the
TensorDeserializer when loading tensors from a serialized model.
It is far faster to deserialize a vLLM model as it utilizes
tensorizer's optimized GPU loading.
tensorizer's optimized GPU loading. Note that this is now
deprecated, as serialized vLLM models are now automatically
inferred as vLLM models.
verify_hash: If True, the hashes of each tensor will be verified against
the hashes stored in the metadata. A `HashMismatchError` will be
raised if any of the hashes do not match.
@ -158,6 +161,7 @@ class TensorizerArgs:
"encryption": self.encryption_keyfile,
"num_readers": self.num_readers
}
if self.encryption_keyfile:
with open_stream(
self.encryption_keyfile,
@ -177,7 +181,14 @@ class TensorizerArgs:
'tensorizer options',
description=('Options for configuring the behavior of the'
' tensorizer deserializer when '
'--load-format=tensorizer'))
'load_format=tensorizer is specified when '
'initializing an LLMEngine, either via the CLI '
'when running the vLLM OpenAI inference server '
'with a JSON string passed to '
'--model-loader-extra-config or as arguments given '
'to TensorizerConfig when passed to '
'model_loader_extra_config in the constructor '
'for LLMEngine.'))
group.add_argument(
"--tensorizer-uri",
@ -222,13 +233,6 @@ class TensorizerArgs:
help="The endpoint for the S3 bucket. Can also be set via the "
"S3_ENDPOINT_URL environment variable.",
)
group.add_argument(
"--vllm-tensorized",
action="store_true",
help="If enabled, indicates that the serialized model is a vLLM "
"model. This is used to determine the behavior of the "
"TensorDeserializer when loading tensors from a "
"serialized model.")
return parser
@ -322,10 +326,9 @@ class TensorizerAgent:
"""
before_mem = get_mem_usage()
start = time.perf_counter()
with open_stream(
self.tensorizer_args.tensorizer_uri,
mode="rb",
**self.tensorizer_args.stream_params,
with _read_stream(
self.tensorizer_config.tensorizer_uri,
**self.tensorizer_args.stream_params
) as stream, TensorDeserializer(
stream,
dtype=self.tensorizer_config.dtype,
@ -345,6 +348,7 @@ class TensorizerAgent:
self._check_tensors_on_meta_device()
self._resize_lora_embeddings()
del self.model.vllm_tensorized_marker
return self.model.eval()
@ -366,3 +370,63 @@ def tensorizer_weights_iterator(
for name, param in state.items():
yield name, param
del state
def is_vllm_tensorized(tensorizer_config: "TensorizerConfig") -> bool:
"""
Infer if the model is a vLLM model by checking the weights for
a vLLM tensorized marker.
Args:
tensorizer_config: The TensorizerConfig object containing the
tensorizer_uri to the serialized model.
Returns:
bool: True if the model is a vLLM model, False otherwise.
"""
tensorizer_args = tensorizer_config._construct_tensorizer_args()
deserializer = TensorDeserializer(open_stream(
tensorizer_args.tensorizer_uri, **tensorizer_args.stream_params),
**tensorizer_args.deserializer_params,
lazy_load=True)
if tensorizer_config.vllm_tensorized:
logger.warning(
"Please note that newly serialized vLLM models are automatically "
"inferred as vLLM models, so setting vllm_tensorized=True is "
"only necessary for models serialized prior to this change.")
return True
if (".vllm_tensorized_marker" in deserializer):
return True
return False
def get_pretensorized_vllm_model(engine: "LLMEngine") -> nn.Module:
model = (engine.model_executor.driver_worker.model_runner.model)
model.register_parameter(
"vllm_tensorized_marker",
nn.Parameter(torch.tensor((1, ), device="meta"), requires_grad=False))
return model
def serialize_vllm_model(engine: "LLMEngine",
tensorizer_config : TensorizerConfig,
encryption_key_path: Optional[str] = None) \
-> nn.Module:
model = get_pretensorized_vllm_model(engine)
tensorizer_args = tensorizer_config._construct_tensorizer_args()
encryption_params = None
if encryption_key_path is not None:
encryption_params = EncryptionParams.random()
with _write_stream(encryption_key_path,
**tensorizer_args.stream_params) as stream:
stream.write(encryption_params.key)
with _write_stream(tensorizer_args.tensorizer_uri,
**tensorizer_args.stream_params) as stream:
serializer = TensorSerializer(stream, encryption=encryption_params)
serializer.write_module(model)
serializer.close()
logger.info("Successfully serialized model to %s",
str(tensorizer_args.tensorizer_uri))
return model