System Info

• transformers version: 4.54.1
• Platform: Windows-11-10.0.26100-SP0
• Python version: 3.12.10
• Huggingface_hub version: 0.34.3
• Safetensors version: 0.5.3
• Accelerate version: 1.1.0
• Accelerate config: - compute_environment: LOCAL_MACHINE
  - distributed_type: NO
  - mixed_precision: bf16
  - use_cpu: False
  - debug: False
  - num_processes: 1
  - machine_rank: 0
  - num_machines: 1
  - gpu_ids: all
  - rdzv_backend: static
  - same_network: True
  - main_training_function: main
  - enable_cpu_affinity: True
  - downcast_bf16: no
  - tpu_use_cluster: False
  - tpu_use_sudo: False
  - tpu_env: []
  - dynamo_config: {'dynamo_backend': 'INDUCTOR'}
• DeepSpeed version: not installed
• PyTorch version (accelerator?): 2.7.1+cu128 (CUDA)
• Tensorflow version (GPU?): not installed (NA)
• Flax version (CPU?/GPU?/TPU?): not installed (NA)
• Jax version: not installed
• JaxLib version: not installed
• Using distributed or parallel set-up in script?:
• Using GPU in script?:
• GPU type: NVIDIA GeForce RTX 4090

Who can help?

@ArthurZucker
@ArthurZucker and @itazap
@SunMarc @MekkCyber

Information

• The official example scripts
• My own modified scripts

Tasks

• An officially supported task in the examples folder (such as GLUE/SQuAD, ...)
• My own task or dataset (give details below)

Reproduction

 # --- VAE / VAE ---
    vae = AutoencoderKLWan.from_pretrained(
        "Wan-AI/Wan2.1-VACE-14B-diffusers",
        subfolder="vae",
        torch_dtype=torch.float32,
    )

    # --- Text Encoder / Кодировщик текста ---
    text_encoder = UMT5EncoderModel.from_pretrained(
        "city96/umt5-xxl-encoder-gguf",
        gguf_file="umt5-xxl-encoder-Q8_0.gguf",
        torch_dtype=torch.float16,
    )

    # --- Transformer / Трансформер ---
    transformer = WanVACETransformer3DModel.from_single_file(
        "https://huggingface.co/QuantStack/Wan2.1_T2V_14B_FusionX_VACE-GGUF/blob/main/Wan2.1_T2V_14B_FusionX_VACE-Q6_K.gguf",
        quantization_config=GGUFQuantizationConfig(
            compute_dtype=torch.float16
        ),
        torch_dtype=torch.float16,
    )

    # --- Pipeline assembly / Сборка пайплайна ---
    pipe = WanVACEPipeline.from_pretrained(
        "Wan-AI/Wan2.1-VACE-14B-diffusers",
        vae=vae,
        text_encoder=text_encoder,
        transformer=transformer,
        torch_dtype=torch.float16
    )

    # --- Scheduler / Планировщик ---
    pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config, flow_shift=flow_shift)

Expected behavior

[WARNING|configuration_utils.py:622] 2025-08-10 11:46:45,092 >> You are using a model of type t5 to instantiate a model of type umt5. This is not supported for all configurations of models and can yield errors.

I want to cinvert into umt5, not simple t5