Update deprecated type hinting in models (#18132)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-05-15 06:06:50 +01:00
committed by GitHub
parent 83f74c698f
commit 26d0419309
130 changed files with 971 additions and 901 deletions

View File

@@ -6,7 +6,8 @@
# LICENSE: https://huggingface.co/Qwen/Qwen-7B/blob/main/LICENSE
"""Inference-only QWen model compatible with HuggingFace weights."""
import json
from typing import Any, Dict, Iterable, Optional, Set, Tuple, Union
from collections.abc import Iterable
from typing import Any, Optional, Union
import torch
from torch import nn
@@ -76,7 +77,7 @@ class QWenAttention(nn.Module):
num_heads: int,
max_position_embeddings: int,
rope_theta: float = 10000,
rope_scaling: Optional[Dict[str, Any]] = None,
rope_scaling: Optional[dict[str, Any]] = None,
cache_config: Optional[CacheConfig] = None,
quant_config: Optional[QuantizationConfig] = None,
prefix: str = "",
@@ -166,7 +167,7 @@ class QWenBlock(nn.Module):
positions: torch.Tensor,
hidden_states: torch.Tensor,
residual: Optional[torch.Tensor],
) -> Tuple[torch.Tensor, torch.Tensor]:
) -> tuple[torch.Tensor, torch.Tensor]:
# Self Attention
if residual is None:
residual = hidden_states
@@ -284,15 +285,15 @@ class QWenBaseModel(nn.Module):
sampling_metadata)
return logits
def load_weights(self, weights: Iterable[Tuple[str,
torch.Tensor]]) -> Set[str]:
def load_weights(self, weights: Iterable[tuple[str,
torch.Tensor]]) -> set[str]:
stacked_params_mapping = [
# (param_name, shard_name, shard_id)
("gate_up_proj", "w2", 0),
("gate_up_proj", "w1", 1),
]
params_dict = dict(self.named_parameters())
loaded_params: Set[str] = set()
loaded_params: set[str] = set()
for name, loaded_weight in weights:
if "rotary_emb.inv_freq" in name:
continue