[Misc] Use config definitions from Transformers library (#21913)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
@@ -29,7 +29,7 @@ from typing import Any, Optional, Union
|
||||
|
||||
import torch
|
||||
from torch import nn
|
||||
from transformers import PretrainedConfig
|
||||
from transformers import Dots1Config
|
||||
|
||||
from vllm.attention import Attention
|
||||
from vllm.compilation.decorators import support_torch_compile
|
||||
@@ -99,7 +99,7 @@ class Dots1MoE(nn.Module):
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
config: PretrainedConfig,
|
||||
config: Dots1Config,
|
||||
quant_config: Optional[QuantizationConfig] = None,
|
||||
prefix: str = "",
|
||||
):
|
||||
@@ -174,7 +174,7 @@ class Dots1Attention(nn.Module):
|
||||
hidden_size: int,
|
||||
num_heads: int,
|
||||
num_kv_heads: int,
|
||||
config: PretrainedConfig,
|
||||
config: Dots1Config,
|
||||
rope_theta: float = 10000,
|
||||
rope_scaling: Optional[dict[str, Any]] = None,
|
||||
max_position_embeddings: int = 8192,
|
||||
@@ -260,7 +260,7 @@ class Dots1DecoderLayer(nn.Module):
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
config: PretrainedConfig,
|
||||
config: Dots1Config,
|
||||
prefix: str,
|
||||
model_config: ModelConfig,
|
||||
cache_config: Optional[CacheConfig] = None,
|
||||
|
||||
Reference in New Issue
Block a user