[Chore] Use Transformers v5 names and drop redundant processor use_fast (#58550)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5.5
parent
484c211fe8
commit
d881812500
@@ -1,7 +1,7 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
import torch
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
from vllm.config import (
|
||||
VllmConfig,
|
||||
@@ -24,7 +24,7 @@ class GatedDeltaNetAttention(PluggableLayer, MambaBase):
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
config: PretrainedConfig,
|
||||
config: PreTrainedConfig,
|
||||
vllm_config: VllmConfig,
|
||||
prefix: str = "",
|
||||
) -> None:
|
||||
|
||||
@@ -4,7 +4,7 @@ import copy
|
||||
|
||||
import torch
|
||||
import torch.nn.functional as F
|
||||
from transformers.configuration_utils import PretrainedConfig
|
||||
from transformers.configuration_utils import PreTrainedConfig
|
||||
|
||||
from vllm.config import (
|
||||
VllmConfig,
|
||||
@@ -37,7 +37,7 @@ from vllm.triton_utils import tl, triton
|
||||
from vllm.v1.attention.backends.linear_attn import LinearAttentionMetadata
|
||||
|
||||
|
||||
def _build_rope_parameters(config: PretrainedConfig) -> dict | None:
|
||||
def _build_rope_parameters(config: PreTrainedConfig) -> dict | None:
|
||||
rope_parameters = copy.deepcopy(getattr(config, "rope_parameters", None)) or {}
|
||||
if "rope_theta" not in rope_parameters and hasattr(config, "rope_theta"):
|
||||
rope_parameters["rope_theta"] = config.rope_theta
|
||||
@@ -385,7 +385,7 @@ class BailingMoELinearAttention(LinearAttention):
|
||||
# --8<-- [end:bailing_moe_linear_attention]
|
||||
def __init__(
|
||||
self,
|
||||
config: PretrainedConfig,
|
||||
config: PreTrainedConfig,
|
||||
vllm_config: VllmConfig,
|
||||
prefix: str = "linear_attn",
|
||||
):
|
||||
|
||||
@@ -1,7 +1,7 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
import torch
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
from vllm.config import (
|
||||
VllmConfig,
|
||||
@@ -24,7 +24,7 @@ class LinearAttention(PluggableLayer, MambaBase):
|
||||
"""Base class for Linear attention layer."""
|
||||
|
||||
def __init__(
|
||||
self, config: PretrainedConfig, vllm_config: VllmConfig, prefix: str = ""
|
||||
self, config: PreTrainedConfig, vllm_config: VllmConfig, prefix: str = ""
|
||||
):
|
||||
super().__init__()
|
||||
self.layer_idx = extract_layer_index(prefix)
|
||||
|
||||
@@ -7,7 +7,7 @@ from typing import TypeVar
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
from vllm.config import ModelConfig
|
||||
from vllm.logger import init_logger
|
||||
@@ -17,7 +17,7 @@ logger = init_logger(__name__)
|
||||
|
||||
|
||||
def get_act_fn(
|
||||
config: PretrainedConfig,
|
||||
config: PreTrainedConfig,
|
||||
static_num_labels: bool = True,
|
||||
) -> "PoolerActivation":
|
||||
# get classification act_fn
|
||||
|
||||
@@ -6,7 +6,7 @@ from typing import TYPE_CHECKING, Any, Union
|
||||
import torch
|
||||
from safetensors.torch import _TYPES as _SAFETENSORS_TO_TORCH_DTYPE
|
||||
from torch.nn import Parameter
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
import vllm.model_executor.layers.fused_moe # noqa
|
||||
from vllm import _custom_ops as ops
|
||||
@@ -370,7 +370,7 @@ class AutoAWQConfig(QuantizationConfig):
|
||||
def maybe_update_config(
|
||||
self,
|
||||
model_name: str,
|
||||
hf_config: PretrainedConfig | None = None,
|
||||
hf_config: PreTrainedConfig | None = None,
|
||||
revision: str | None = None,
|
||||
):
|
||||
if self.modules_to_not_convert:
|
||||
|
||||
@@ -6,7 +6,7 @@ from typing import Any
|
||||
|
||||
import torch
|
||||
from safetensors.torch import _TYPES as _SAFETENSORS_TO_TORCH_DTYPE
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
import vllm.model_executor.layers.fused_moe # noqa
|
||||
from vllm.logger import init_logger
|
||||
@@ -273,7 +273,7 @@ class AutoGPTQConfig(QuantizationConfig):
|
||||
def maybe_update_config(
|
||||
self,
|
||||
model_name: str,
|
||||
hf_config: PretrainedConfig | None = None,
|
||||
hf_config: PreTrainedConfig | None = None,
|
||||
revision: str | None = None,
|
||||
):
|
||||
if self.modules_in_block_to_quantize:
|
||||
|
||||
@@ -8,7 +8,7 @@ from typing import TYPE_CHECKING, Any
|
||||
import regex as re
|
||||
import torch
|
||||
from torch import nn
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import PreTrainedConfig
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from vllm.model_executor.layers.quantization import QuantizationMethods
|
||||
@@ -265,7 +265,7 @@ class QuantizationConfig(ABC):
|
||||
def maybe_update_config( # noqa: B027
|
||||
self,
|
||||
model_name: str,
|
||||
hf_config: PretrainedConfig | None = None,
|
||||
hf_config: PreTrainedConfig | None = None,
|
||||
revision: str | None = None,
|
||||
):
|
||||
"""Interface to update values after config initialization.
|
||||
|
||||
Reference in New Issue
Block a user