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

11
vllm/model_executor/models/olmo.py
View File

@@ -22,7 +22,8 @@
# See the License for the specific language governing permissions and
# limitations under the License.
"""Inference-only OLMo model compatible with HuggingFace weights."""
from typing import Iterable, Optional, Set, Tuple, Union
from collections.abc import Iterable
from typing import Optional, Union
import torch
from torch import nn
@@ -209,7 +210,7 @@ class OlmoDecoderLayer(nn.Module):
self,
positions: torch.Tensor,
hidden_states: torch.Tensor,
) -> Tuple[torch.Tensor, Optional[Tuple[torch.Tensor, torch.Tensor]]]:
) -> tuple[torch.Tensor, Optional[tuple[torch.Tensor, torch.Tensor]]]:
# Attention block.
residual = hidden_states
hidden_states = self.input_layernorm(hidden_states)
@@ -338,8 +339,8 @@ class OlmoForCausalLM(nn.Module, 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]:
stacked_params_mapping = [
# (param_name, shard_name, shard_id)
("qkv_proj", "q_proj", "q"),
@@ -349,7 +350,7 @@ class OlmoForCausalLM(nn.Module, SupportsPP):
("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