Skip to content

Support for LlamaForSequenceClassification #20807

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 1 commit into from
Jul 13, 2025
Merged

Support for LlamaForSequenceClassification #20807

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
1 change: 1 addition & 0 deletions tests/models/registry.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -326,6 +326,7 @@ def check_available_online(
hf_overrides={"architectures": ["GemmaForSequenceClassification"], # noqa: E501
"classifier_from_token": ["Yes"], # noqa: E501
"method": "no_post_processing"}), # noqa: E501
"LlamaForSequenceClassification": _HfExamplesInfo("Skywork/Skywork-Reward-V2-Llama-3.2-1B"), # noqa: E501
"ModernBertForSequenceClassification": _HfExamplesInfo("Alibaba-NLP/gte-reranker-modernbert-base", v0_only=True), # noqa: E501
"RobertaForSequenceClassification": _HfExamplesInfo("cross-encoder/quora-roberta-base", v0_only=True), # noqa: E501
"XLMRobertaForSequenceClassification": _HfExamplesInfo("BAAI/bge-reranker-v2-m3", v0_only=True), # noqa: E501
Expand Down
4 changes: 4 additions & 0 deletions vllm/model_executor/models/llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,6 +49,7 @@
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import IntermediateTensors

from .adapters import as_seq_cls_model
from .interfaces import SupportsLoRA, SupportsPP
from .utils import (AutoWeightsLoader, PPMissingLayer, extract_layer_index,
is_pp_missing_parameter,
Expand Down Expand Up @@ -645,3 +646,6 @@ def permute(w: torch.Tensor, n_heads: int):
name = name.replace(item, mapping[item])

return name, loaded_weight


LlamaForSequenceClassification = as_seq_cls_model(LlamaForCausalLM)
3 changes: 2 additions & 1 deletion vllm/model_executor/models/registry.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,7 +182,8 @@
"GemmaForSequenceClassification": ("gemma", "GemmaForSequenceClassification"), # noqa: E501
"Qwen2ForSequenceClassification": ("qwen2", "Qwen2ForSequenceClassification"), # noqa: E501
"Qwen3ForSequenceClassification": ("qwen3", "Qwen3ForSequenceClassification"), # noqa: E501
"JinaVLForRanking": ("jina_vl", "JinaVLForSequenceClassification"), # noqa: E501
"LlamaForSequenceClassification": ("llama", "LlamaForSequenceClassification"), # noqa: E501
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please keep this with the other text models. JinaVLForRanking is specially put at the end because it's multimodal, maybe add a code comment like in the other sections

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you. I rearranged as you suggested.

"JinaVLForRanking": ("jina_vl", "JinaVLForSequenceClassification"), # noqa: E501,
}

_MULTIMODAL_MODELS = {
Expand Down