invoke-ai/InvokeAI · error · TypeError

Expected ModelPatchRaw for LoRA '{lora.lora.key}', got {type

Error message

Expected ModelPatchRaw for LoRA '{lora.lora.key}', got {type(lora_info.model).__name__}. The LoRA model may be corrupted or incompatible.

What it means

_lora_iterator throws this TypeError when a LoRA listed on the Mistral encoder input loads as something other than a ModelPatchRaw. ModelPatchRaw is the internal representation LoRA application expects; any other object means the model file is corrupted or incompatible with LoRA patching. The message identifies the offending LoRA key and the actual loaded type.

Source

Thrown at invokeai/app/invocations/flux2_dev_text_encoder.py:248

                f"layers {DEV_EXTRACTION_LAYERS} and requires at least {max(DEV_EXTRACTION_LAYERS)}. "
                "This is not a supported FLUX.2 [dev] text encoder."
            )
        extraction_layers = DEV_EXTRACTION_LAYERS

        # Concatenate the selected layers along the hidden dim: (B, seq, 3 * hidden_size).
        # This is byte-identical to stack(dim=1).permute(0,2,1,3).reshape(...) but avoids
        # the two intermediate full copies that stack + permute-reshape would allocate.
        prompt_embeds = torch.cat([outputs.hidden_states[i] for i in extraction_layers], dim=-1)
        prompt_embeds = prompt_embeds.to(dtype=text_encoder.dtype, device=device)

        return prompt_embeds

    def _lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[ModelPatchRaw, float]]:
        """Iterate over LoRAs to apply to the Mistral encoder."""
        for lora in self.mistral_encoder.loras:
            lora_info = context.models.load(lora.lora)
            if not isinstance(lora_info.model, ModelPatchRaw):
                raise TypeError(
                    f"Expected ModelPatchRaw for LoRA '{lora.lora.key}', got {type(lora_info.model).__name__}. "
                    "The LoRA model may be corrupted or incompatible."
                )
            yield (lora_info.model, lora.weight)
            del lora_info

View on GitHub (pinned to 0b6a024f2f)

Solutions

Re-download the LoRA file referenced by the reported key and re-import it into the model manager.
Confirm the wired model is actually a LoRA compatible with the Mistral encoder, not another model type.
Remove the problematic LoRA from the encoder's LoRA list and retry.

Defensive patterns

Strategy: type-guard

Validate before calling

lora_info = context.models.load(lora.lora)
if not isinstance(lora_info.model, ModelPatchRaw):
    raise TypeError(f"LoRA {lora.lora.key} loaded as {type(lora_info.model).__name__}, expected ModelPatchRaw")

Type guard

def is_lora_patch(model) -> bool:
    return isinstance(model, ModelPatchRaw)

Try / catch

try:
    output = text_encoder_invocation.invoke(context)
except TypeError as e:
    if "Expected ModelPatchRaw for LoRA" in str(e):
        remove_bad_loras(context)
        output = text_encoder_invocation.invoke(context)
    else:
        raise

Prevention

Validate LoRA files import correctly in the Model Manager before use.
Only wire LoRA models into LoRA inputs.
Re-download LoRAs that fail to load as ModelPatchRaw.

When it happens

Trigger: invoke() -> _encode_prompt -> _lora_iterator iterates mistral_encoder.loras, and context.models.load(lora.lora).model is not an instance of ModelPatchRaw.

Common situations: A corrupted or wrong-format LoRA file is attached to the FLUX.2 [dev] text encoder's LoRA list, or a non-LoRA model was mistakenly wired into the LoRA input.

Related errors

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/36d551bb67164471. Report an issue: GitHub.