[Chore] Separate out vllm.utils.mem_utils (#27143)

Signed-off-by: iAmir97 <Amir.balwel@embeddedllm.com>
Signed-off-by: iAmir97 <71513472+iAmir97@users.noreply.github.com>
Co-authored-by: iAmir97 <Amir.balwel@embeddedllm.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
This commit is contained in:
iAmir97
2025-10-18 10:06:59 +00:00
committed by GitHub
co-authored by iAmir97 gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
parent 83004020fd
commit 1d165d6d85
19 changed files with 268 additions and 252 deletions
+1 -1
View File
@@ -6,7 +6,7 @@ import torch
from vllm import LLM, SamplingParams
from vllm.device_allocator.cumem import CuMemAllocator
from vllm.utils import GiB_bytes
from vllm.utils.mem_constants import GiB_bytes
from ..utils import create_new_process_for_each_test
+1 -1
View File
@@ -11,7 +11,7 @@ from tests.kernels.utils import opcheck
from vllm import _custom_ops as ops
from vllm.attention.layer import Attention, MultiHeadAttention
from vllm.platforms import current_platform
from vllm.utils import get_max_shared_memory_bytes
from vllm.utils.mem_utils import get_max_shared_memory_bytes
if not current_platform.is_rocm():
from xformers import ops as xops
+1 -1
View File
@@ -7,7 +7,7 @@ from unittest.mock import patch
import pytest
from vllm import LLM
from vllm.utils import GiB_bytes
from vllm.utils.mem_constants import GiB_bytes
from vllm.v1.core.kv_cache_utils import (
generate_scheduler_kv_cache_config,
get_kv_cache_configs,
+1 -1
View File
@@ -46,10 +46,10 @@ from vllm.platforms import current_platform
from vllm.transformers_utils.tokenizer import get_tokenizer
from vllm.utils import (
FlexibleArgumentParser,
GB_bytes,
cuda_device_count_stateless,
get_open_port,
)
from vllm.utils.mem_constants import GB_bytes
if current_platform.is_rocm():
from amdsmi import (
+1 -2
View File
@@ -23,7 +23,6 @@ from vllm.transformers_utils.detokenizer_utils import convert_ids_list_to_tokens
from vllm.utils import (
FlexibleArgumentParser,
MemorySnapshot,
bind_kv_cache,
common_broadcastable_dtype,
current_stream,
@@ -33,13 +32,13 @@ from vllm.utils import (
join_host_port,
make_zmq_path,
make_zmq_socket,
memory_profiling,
sha256,
split_host_port,
split_zmq_path,
unique_filepath,
)
from vllm.utils.mem_utils import MemorySnapshot, memory_profiling
from ..utils import create_new_process_for_each_test, flat_product
+2 -1
View File
@@ -14,7 +14,8 @@ from vllm.multimodal.inputs import (
PlaceholderRange,
)
from vllm.sampling_params import SamplingParams
from vllm.utils import GiB_bytes, sha256, sha256_cbor
from vllm.utils import sha256, sha256_cbor
from vllm.utils.mem_constants import GiB_bytes
from vllm.v1.core.kv_cache_manager import KVCacheManager
from vllm.v1.core.kv_cache_utils import (
BlockHash,
+1 -1
View File
@@ -13,7 +13,7 @@ from vllm.config import (
)
from vllm.pooling_params import PoolingParams
from vllm.sampling_params import SamplingParams
from vllm.utils import GiB_bytes
from vllm.utils.mem_constants import GiB_bytes
from vllm.v1.core.kv_cache_utils import estimate_max_model_len, get_kv_cache_configs
from vllm.v1.core.sched.output import CachedRequestData, NewRequestData, SchedulerOutput
from vllm.v1.worker.tpu_model_runner import (
+2 -1
View File
@@ -21,7 +21,8 @@ from vllm.distributed.parallel_state import (
from vllm.model_executor.layers.mamba.mamba_mixer2 import MambaMixer2
from vllm.platforms import current_platform
from vllm.sampling_params import SamplingParams
from vllm.utils import GiB_bytes, update_environment_variables
from vllm.utils import update_environment_variables
from vllm.utils.mem_constants import GiB_bytes
from vllm.v1.core.kv_cache_utils import estimate_max_model_len, get_kv_cache_configs
from vllm.v1.core.sched.output import CachedRequestData, NewRequestData, SchedulerOutput
from vllm.v1.kv_cache_interface import (
@@ -11,7 +11,7 @@ import pytest
import torch
from vllm.engine.arg_utils import EngineArgs
from vllm.utils import MemorySnapshot
from vllm.utils.mem_utils import MemorySnapshot
from vllm.v1.worker.gpu_worker import Worker, init_worker_distributed_environment
# Global queue to track operation order across processes