biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Update deprecated type hinting in models (#18132)

Browse Source 
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
Download Patch File Download Diff File Expand all files Collapse all files

19
vllm/model_executor/models/qwen2.py
View File

@@ -23,7 +23,8 @@
# See the License for the specific language governing permissions and
# limitations under the License.
"""Inference-only Qwen2 model compatible with HuggingFace weights."""
from typing import Any, Iterable, Optional, Set, Tuple, Union
from collections.abc import Iterable
from typing import Any, Optional, Union
import torch
from torch import nn
@@ -108,7 +109,7 @@ class Qwen2Attention(nn.Module):
rope_theta: float = 10000,
cache_config: Optional[CacheConfig] = None,
quant_config: Optional[QuantizationConfig] = None,
rope_scaling: Optional[Tuple] = None,
rope_scaling: Optional[tuple] = None,
prefix: str = "",
attn_type: str = AttentionType.DECODER,
dual_chunk_attention_config: Optional[dict[str, Any]] = None,
@@ -245,7 +246,7 @@ class Qwen2DecoderLayer(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
@@ -367,8 +368,8 @@ class Qwen2Model(nn.Module):
hidden_states, _ = self.norm(hidden_states, residual)
return hidden_states
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)
("qkv_proj", "q_proj", "q"),
@@ -378,7 +379,7 @@ class Qwen2Model(nn.Module):
("gate_up_proj", "up_proj", 1),
]
params_dict = dict(self.named_parameters(remove_duplicate=False))
loaded_params: Set[str] = set()
loaded_params: set[str] = set()
for name, loaded_weight in weights:
if "rotary_emb.inv_freq" in name:
continue
@@ -490,8 +491,8 @@ class Qwen2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
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]:
loader = AutoWeightsLoader(
self,
skip_prefixes=(["lm_head."]
@@ -559,7 +560,7 @@ class Qwen2EmbeddingModel(nn.Module, SupportsLoRA, SupportsPP):
) -> Optional[PoolerOutput]:
return self._pooler(hidden_states, pooling_metadata)
def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]):
weights = self.hf_to_vllm_mapper.apply(weights)
weights = ((name, data) for name, data in weights
if not name.startswith("lm_head."))