Skip to content

Support LoRA for Mistral3 #17428

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Apr 30, 2025
Merged

Support LoRA for Mistral3 #17428

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2a3d0be
Support LoRA for Mistral3
mgoin Apr 30, 2025
170c2e6
Update docs
mgoin Apr 30, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion docs/source/models/supported_models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -990,7 +990,7 @@ See [this page](#generative-models) for more information on how to use generativ
* Mistral3
* T + I<sup>+</sup>
* `mistralai/Mistral-Small-3.1-24B-Instruct-2503`, etc.
*
* ✅︎
* ✅︎
* ✅︎
- * `MllamaForConditionalGeneration`
Expand Down
17 changes: 14 additions & 3 deletions vllm/model_executor/models/mistral3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@
from vllm.model_executor.layers.linear import (ColumnParallelLinear,
RowParallelLinear)
from vllm.model_executor.layers.quantization import QuantizationConfig
from vllm.model_executor.models.module_mapping import MultiModelKeys
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.multimodal import MULTIMODAL_REGISTRY
from vllm.multimodal.inputs import (MultiModalDataDict, MultiModalFieldConfig,
Expand All @@ -31,7 +32,8 @@
from vllm.multimodal.profiling import BaseDummyInputsBuilder
from vllm.sequence import IntermediateTensors

from .interfaces import MultiModalEmbeddings, SupportsMultiModal, SupportsPP
from .interfaces import (MultiModalEmbeddings, SupportsLoRA,
SupportsMultiModal, SupportsPP)
from .pixtral import PixtralHFEncoderInfo, PixtralHFVisionModel
from .utils import (AutoWeightsLoader, flatten_bn, init_vllm_registered_model,
maybe_prefix, merge_multimodal_embeddings)
Expand Down Expand Up @@ -382,8 +384,8 @@ def init_vision_tower_for_llava(
_build_mistral3_processor,
info=_build_mistral3_info,
dummy_inputs=Mistral3DummyInputsBuilder)
class Mistral3ForConditionalGeneration(nn.Module, SupportsMultiModal,
SupportsPP):
class Mistral3ForConditionalGeneration(nn.Module, SupportsLoRA,
SupportsMultiModal, SupportsPP):

packed_modules_mapping = {
"qkv_proj": ["q_proj", "k_proj", "v_proj"],
Expand Down Expand Up @@ -594,3 +596,12 @@ def load_weights(self, weights: Iterable[Tuple[str,
torch.Tensor]]) -> Set[str]:
loader = AutoWeightsLoader(self)
return loader.load_weights(weights)

def get_mm_mapping(self) -> MultiModelKeys:
"""
Get the module prefix in multimodal models
"""
return MultiModelKeys.from_string_field(
language_model="language_model",
connector="multi_modal_projector",
tower_model="vision_tower")