vllm.model_executor.models.voyage ¶

class VoyageQwen3BidirectionalEmbedModel(nn.Module):
    """
    Qwen3Model + Voyage embedding head + bidirectional attention.

    Checkpoint conventions (HF):
      - MLP: gate_proj + up_proj (unfused)
      - Attn: q_proj + k_proj + v_proj (unfused)
      - Linear head: linear.weight
      - Weights prefixed with "model." (e.g., model.layers.0...)

    vLLM Qwen3Model expects:
      - mlp.gate_up_proj (fused)
      - self_attn.qkv_proj (fused)
      - No "model." prefix
    """

    def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
        super().__init__()
        self.config = vllm_config.model_config.hf_config
        self.model = Qwen3Model(
            vllm_config=vllm_config, prefix=maybe_prefix(prefix, "model")
        )

        # Embedding head (hidden_size -> num_labels, bias=False)
        self.linear = nn.Linear(
            self.config.hidden_size,
            self.config.num_labels,
            bias=False,
        )

        self.make_empty_intermediate_tensors = (
            self.model.make_empty_intermediate_tensors
        )

    def embed_input_ids(self, input_ids: torch.Tensor) -> torch.Tensor:
        return self.model.embed_input_ids(input_ids)

    def forward(
        self,
        input_ids: torch.Tensor | None,
        positions: torch.Tensor,
        intermediate_tensors: IntermediateTensors | None = None,
        inputs_embeds: torch.Tensor | None = None,
    ) -> torch.Tensor:
        out = self.model(input_ids, positions, intermediate_tensors, inputs_embeds)
        return self.linear(out)

    def load_weights(self, weights: Iterable[WeightItem]) -> set[str]:
        loader = AutoWeightsLoader(self)
        return loader.load_weights(weights)