Skip to content

[Model] Decouple glm4v #22751

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
Aug 13, 2025
Merged

[Model] Decouple glm4v #22751

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
2 changes: 1 addition & 1 deletion docs/models/supported_models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -615,7 +615,7 @@ These models primarily accept the [`LLM.generate`](./generative_models.md#llmgen
| `Gemma3nForConditionalGeneration` | Gemma 3n | T + I + A | `google/gemma-3n-E2B-it`, `google/gemma-3n-E4B-it`, etc. | | | ✅︎ |
| `GLM4VForCausalLM`<sup>^</sup> | GLM-4V | T + I | `zai-org/glm-4v-9b`, `zai-org/cogagent-9b-20241220`, etc. | ✅︎ | ✅︎ | ✅︎ |
| `Glm4vForConditionalGeneration` | GLM-4.1V-Thinking | T + I<sup>E+</sup> + V<sup>E+</sup> | `zai-org/GLM-4.1V-9B-Thinking`, etc. | ✅︎ | ✅︎ | ✅︎ |
| `Glm4vMoeForConditionalGeneration` | GLM-4.5V | T + I<sup>E+</sup> + V<sup>E+</sup> | `zai-org/GLM-4.5V`, etc. | ✅︎ | ✅︎ | ✅︎ |
| `Glm4vMoeForConditionalGeneration` | GLM-4.5V | T + I<sup>E+</sup> + V<sup>E+</sup> | `zai-org/GLM-4.5V`, etc. | | ✅︎ | ✅︎ |
| `GraniteSpeechForConditionalGeneration` | Granite Speech | T + A | `ibm-granite/granite-speech-3.3-8b` | ✅︎ | ✅︎ | ✅︎ |
| `H2OVLChatModel` | H2OVL | T + I<sup>E+</sup> | `h2oai/h2ovl-mississippi-800m`, `h2oai/h2ovl-mississippi-2b`, etc. | | ✅︎ | ✅︎ |
| `Idefics3ForConditionalGeneration` | Idefics3 | T + I | `HuggingFaceM4/Idefics3-8B-Llama3`, etc. | ✅︎ | | ✅︎ |
Expand Down
26 changes: 21 additions & 5 deletions vllm/model_executor/models/glm4_1v.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1227,10 +1227,7 @@ class Glm4vForConditionalGeneration(nn.Module, SupportsMultiModal,
"k_proj",
"v_proj",
],
"gate_up_proj": [
"gate_proj",
"up_proj",
],
"gate_up_proj": ["gate_up_proj"]
}

# To ensure correct weight loading and mapping.
Expand Down Expand Up @@ -1567,7 +1564,26 @@ def get_mm_mapping(self) -> MultiModelKeys:
Get the module prefix in multimodal models
"""
return MultiModelKeys.from_string_field(
language_model="language_model",
language_model="language_model.model",
connector="visual.merger.",
tower_model="visual.",
)


@MULTIMODAL_REGISTRY.register_processor(
Glm4vMultiModalProcessor,
info=Glm4vProcessingInfo,
dummy_inputs=Glm4vDummyInputsBuilder,
)
class Glm4vMoeForConditionalGeneration(Glm4vForConditionalGeneration):
packed_modules_mapping = {
"qkv_proj": [
"q_proj",
"k_proj",
"v_proj",
],
"gate_up_proj": [
"gate_proj",
"up_proj",
],
}
Comment on lines +1579 to +1589
Copy link
Contributor

Choose a reason for hiding this comment

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

high

The pull request description and documentation state that LoRA is not supported for Glm4vMoeForConditionalGeneration due to a bug. However, the class currently inherits SupportsLoRA from its parent, which will cause vLLM to attempt to apply LoRA adapters if provided, leading to a runtime error. To prevent this, I suggest adding an __init__ method that explicitly checks for and disallows LoRA configuration for this model. This will provide a clear error message to users and make the code's behavior consistent with the documentation.

Suggested change
packed_modules_mapping = {
"qkv_proj": [
"q_proj",
"k_proj",
"v_proj",
],
"gate_up_proj": [
"gate_proj",
"up_proj",
],
}
packed_modules_mapping = {
"qkv_proj": [
"q_proj",
"k_proj",
"v_proj",
],
"gate_up_proj": [
"gate_proj",
"up_proj",
],
}
def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
# LoRA is not supported for this model yet.
if vllm_config.lora_config:
raise NotImplementedError(
"LoRA is not currently supported for "
"Glm4vMoeForConditionalGeneration."
)
super().__init__(vllm_config=vllm_config, prefix=prefix)

2 changes: 1 addition & 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 @@ -208,7 +208,7 @@
"Gemma3nForConditionalGeneration": ("gemma3n_mm", "Gemma3nForConditionalGeneration"), # noqa: E501
"GLM4VForCausalLM": ("glm4v", "GLM4VForCausalLM"),
"Glm4vForConditionalGeneration": ("glm4_1v", "Glm4vForConditionalGeneration"), # noqa: E501
"Glm4vMoeForConditionalGeneration": ("glm4_1v", "Glm4vForConditionalGeneration"), # noqa: E501
"Glm4vMoeForConditionalGeneration": ("glm4_1v", "Glm4vMoeForConditionalGeneration"), # noqa: E501
"GraniteSpeechForConditionalGeneration": ("granite_speech", "GraniteSpeechForConditionalGeneration"), # noqa: E501
"H2OVLChatModel": ("h2ovl", "H2OVLChatModel"),
"InternVLChatModel": ("internvl", "InternVLChatModel"),
Expand Down