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

[Misc] Split model loader (#17712)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-05-07 12:42:26 +08:00
committed by GitHub
parent 8d84d836d1
commit 822de7fb94
21 changed files with 1681 additions and 1566 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/model_executor/models/mllama4.py
View File

@@ -37,7 +37,7 @@ from vllm.model_executor.layers.linear import (ColumnParallelLinear,
RowParallelLinear)
from vllm.model_executor.layers.quantization import QuantizationConfig
from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.model_loader.loader import _initialize_model
from vllm.model_executor.model_loader.utils import initialize_model
from vllm.model_executor.model_loader.weight_utils import default_weight_loader
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
@@ -670,7 +670,7 @@ class Llama4ForConditionalGeneration(nn.Module, SupportsMultiModal,
self.config,
None,
prefix=maybe_prefix(prefix, "multi_modal_projector"))
self.language_model = _initialize_model(
self.language_model = initialize_model(
vllm_config=vllm_config.with_hf_config(config.text_config,
["LlamaForCausalLM"]),
prefix=maybe_prefix(prefix, "language_model"),