[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:
Harry Mellor
2026-09-24 14:14:10 +00:00
committed by GitHub
co-authored by Claude Opus 5.5
parent 484c211fe8
commit d881812500
218 changed files with 869 additions and 813 deletions
+6 -6
View File
@@ -16,7 +16,7 @@ from einops import rearrange
from transformers import (
BaseImageProcessor,
BatchFeature,
PretrainedConfig,
PreTrainedConfig,
)
from vllm.compilation.decorators import support_torch_compile
@@ -405,7 +405,7 @@ class MolmoAttention(nn.Module):
def __init__(
self,
config: PretrainedConfig,
config: PreTrainedConfig,
cache_config: CacheConfig | None = None,
quant_config: QuantizationConfig | None = None,
prefix: str = "",
@@ -515,7 +515,7 @@ class LanguageModelMLP(nn.Module):
def __init__(
self,
config: PretrainedConfig,
config: PreTrainedConfig,
input_dim: int | None = None,
quant_config: QuantizationConfig | None = None,
prefix: str = "",
@@ -557,7 +557,7 @@ class ImageProjectorMLP(nn.Module):
def __init__(
self,
config: PretrainedConfig,
config: PreTrainedConfig,
input_dim: int | None = None,
quant_config: QuantizationConfig | None = None,
prefix: str = "",
@@ -598,7 +598,7 @@ class ImageProjectorMLP(nn.Module):
class MolmoDecoderLayer(nn.Module):
def __init__(
self,
config: PretrainedConfig,
config: PreTrainedConfig,
cache_config: CacheConfig | None = None,
quant_config: QuantizationConfig | None = None,
prefix: str = "",
@@ -681,7 +681,7 @@ class MolmoVisionBackbone(nn.Module, SupportsQuant):
def __init__(
self,
config: PretrainedConfig,
config: PreTrainedConfig,
vision_config: VisionBackboneConfig,
quant_config: QuantizationConfig | None = None,
prefix: str = "",