Maestro: [ERROR] "The generation of the video has encountered an error, please check your terminal for m…

App: Maestro (Maestro.git)
Repo: https://github.com/Blizaine/Maestro.git
Generated: 2026-08-09T20:31:11.734Z
Pinokio: 8.0.40
Platform: win32 x64
Node: v22.21.1

Summary

System

{
  "pinokio": {
    "version": "8.0.40",
    "node": "v22.21.1",
    "platform": "win32",
    "arch": "x64"
  },
  "hardware": {
    "gpu": "nvidia",
    "gpu_model": "nvidia geforce rtx 3080 laptop gpu",
    "ram_gb": 16,
    "vram_gb": 16
  },
  "os": {
    "platform": "Windows",
    "distro": "Microsoft Windows 11 Home",
    "release": "10.0.26200",
    "codename": "25H2",
    "kernel": "10.0.26200",
    "arch": "x64",
    "build": "26200",
    "servicepack": "0.0",
    "uefi": true
  },
  "system": {
    "manufacturer": "LENOVO",
    "model": "82N6",
    "version": "Legion 7 16ACHg6",
    "virtual": false
  },
  "cpu": {
    "manufacturer": "AMD",
    "brand": "Ryzen 7 5800H with Radeon Graphics",
    "vendor": "AuthenticAMD",
    "family": "25",
    "model": "80",
    "stepping": "0",
    "revision": "20480",
    "speed": 3.2,
    "speedMin": 3.2,
    "speedMax": 3.2,
    "cores": 16,
    "physicalCores": 8,
    "processors": 1,
    "performanceCores": 16,
    "efficiencyCores": 0,
    "virtualization": true,
    "cache": {
      "l1d": 256,
      "l1i": 256,
      "l2": 4194304,
      "l3": 16777216
    }
  },
  "memory": {
    "total": 17041780736,
    "free": 7538057216,
    "used": 9503723520,
    "active": 9503723520,
    "available": 7537991680,
    "buffers": 0,
    "cached": 0,
    "slab": 0,
    "buffcache": 0,
    "swaptotal": 21002977280,
    "swapused": 1366294528,
    "swapfree": 19636682752
  },
  "gpus": [
    {
      "model": "nvidia geforce rtx 3080 laptop gpu"
    }
  ],
  "graphics": {
    "controllers": [
      {
        "vendor": "NVIDIA",
        "model": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "bus": "PCI",
        "vram": 16384,
        "vramDynamic": false,
        "driverVersion": "572.83"
      }
    ],
    "displays": [
      {
        "model": "Lenovo DisplayHDR",
        "main": true,
        "builtin": false,
        "connection": "DP embedded",
        "currentResX": 1707,
        "currentResY": 1067,
        "resolutionX": 2560,
        "resolutionY": 1600,
        "pixelDepth": 32,
        "currentRefreshRate": 165
      }
    ]
  }
}

Logs

logs/api/start.js/1786307398147

Source: api / start.js
Lines: 132 total, last 132 included

[api shell.run]
Microsoft Windows [Version 10.0.26200.8875]
(c) Microsoft Corporation. All rights reserved.

C:\pinokio\api\Maestro.git\app>conda_hook & conda deactivate & conda deactivate & conda deactivate & conda activate base & C:\pinokio\api\Maestro.git\app\env\Scripts\activate C:\pinokio\api\Maestro.git\app\env & bluefairy-activate.cmd && python launch.py 
[OK] bluefairy active: uv/npm/bun protection enabled
[Maestro] Installing download stall protection...
[safe_download] tqdm download hook installed
[Maestro] Importing WanGP engine...
[GGUF][llama.cpp CUDA] kernels available.
[GuideLoader] Guide not found: prompt_enhancer/dramabox_speech_rules.md
[GuideLoader] Guide not found: prompt_enhancer/dramabox_dialogue_rules.md
[Quanto][INT8] Injected int8 kernels ACTIVE (backend=triton).
[Maestro] WanGP loaded: 193 models available
[Workspace] Active workspace: default (outputs)
[Maestro] Gradio classic UI mounted at /classic
[Maestro] React UI serving from C:\pinokio\api\Maestro.git\ui\dist

==================================================
  Maestro UI:    http://127.0.0.1:42003/
  Classic UI:    http://127.0.0.1:42003/classic/
  API docs:      http://127.0.0.1:42003/docs
==================================================

INFO:     Started server process [26632]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:42003 (Press CTRL+C to quit)
INFO:     127.0.0.1:62072 - "GET /api/v1/presets HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "GET /api/v1/loras/ltx2_22B_distilled_1_1 HTTP/1.1" 200 OK
INFO:     127.0.0.1:57699 - "GET /api/v1/loras/ltx2_22B_distilled_1_1/details HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "GET /api/v1/model-visibility HTTP/1.1" 200 OK
INFO:     127.0.0.1:57699 - "GET /api/v1/workspaces HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "GET /api/v1/system-config HTTP/1.1" 200 OK
INFO:     127.0.0.1:54666 - "GET /api/v1/services-config HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "GET /api/v1/llm/models HTTP/1.1" 200 OK
INFO:     127.0.0.1:62610 - "GET /api/v1/system/preflight HTTP/1.1" 200 OK
INFO:     127.0.0.1:54666 - "GET /api/v1/director/pipelines HTTP/1.1" 200 OK
INFO:     127.0.0.1:62610 - "GET /api/v1/system-detect HTTP/1.1" 200 OK
INFO:     127.0.0.1:62072 - "GET /api/v1/models HTTP/1.1" 200 OK
INFO:     127.0.0.1:54666 - "GET /api/v1/defaults/minimax_h3_ref2va HTTP/1.1" 200 OK
INFO:     127.0.0.1:62072 - "GET /api/v1/loras/minimax_h3_ref2va HTTP/1.1" 200 OK
INFO:     127.0.0.1:62610 - "GET /api/v1/model-options/minimax_h3_ref2va HTTP/1.1" 200 OK
INFO:     127.0.0.1:57699 - "GET /api/v1/loras/minimax_h3_ref2va/details HTTP/1.1" 200 OK
INFO:     127.0.0.1:54666 - "GET /api/v1/loras/minimax_h3_ref2va HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "GET /api/v1/loras/installed HTTP/1.1" 200 OK
INFO:     127.0.0.1:53299 - "POST /api/v1/loras/check-updates HTTP/1.1" 200 OK
INFO:     127.0.0.1:58198 - "POST /api/v1/upload HTTP/1.1" 200 OK
INFO:     127.0.0.1:58198 - "GET /api/v1/uploads/b751c9c0.jpg HTTP/1.1" 200 OK
INFO:     127.0.0.1:58198 - "POST /api/v1/generate HTTP/1.1" 200 OK
[VRAM] Job 96a74f4b: coefficient 0.80 → 0.589 (cap 12.8GB → 9.4GB)
[VRAM]   +0.022 bonus for 0.51× compute vs baseline (1280×704 × 124 frames, lighter than baseline)
[VRAM]   - MiniMax H3 transformer residency capped at 9.4 GB to preserve 6.6 GB of packed-sequence workspace
[load] Task 1/1 ready, ID: 1, model: minimax_h3_ref2va
[Gen 96a74f4b] save_path locked to: outputs

[Task 1/1] make it smoke up ...
Loading Model 'ckpts\minimax_h3_ref2va_pruned_fp8_scaled.safetensors' ...
Loading Text Encoder 'ckpts\minimax_h3\qwen3vl-32B-MiniMax-H3-Q2_K.gguf' ...
Loading model H3 Omni — Pruned...
[MiniMax H3] Loaded pruned 20B curve transformer (contiguous QKV, fused projection).
C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\utils\weight_norm.py:143: FutureWarning: `torch.nn.utils.weight_norm` is deprecated in favor of `torch.nn.utils.parametrizations.weight_norm`.
  WeightNorm.apply(module, name, dim)
Pytorch compilation is not supported for this Model
************ Memory Management for the GPU Poor (mmgp 3.7.6) by DeepBeepMeep ************
Hooked to model 'transformer' (MiniMaxH3Transformer)
Async loading plan for model 'transformer' : base size of 58.04 MB will be preloaded with a 369.18 MB async circular shuttle
Hooked to model 'text_encoder' (Qwen3VLTextModel)
Async loading plan for model 'text_encoder' : base size of 243.43 MB will be preloaded with a 153.13 MB async circular shuttle
Hooked to model 'vision_encoder' (Qwen3VLVisionModel)
Hooked to model 'vae' (AutoencoderKLMiniMaxH3)
Async loading plan for model 'vae' : 270.82 MB will be preloaded (base size of 14.69 MB + 5.2% of recurrent layers data) with a 190.02 MB async shuttle
Hooked to model 'audio_vae' (AutoencoderKLMiniMaxH3Audio)
[LoRA] No LoRAs activated for this generation (model_type=minimax_h3_ref2va)
  Model loaded[MiniMax H3 Ref2VA] Added explicit reference relationships to an untagged prompt.
  Encoding PromptTraceback (most recent call last):
  File "C:\pinokio\api\Maestro.git\app\wgp.py", line 8678, in generate_video
    samples = call_with_sticky_interrupt(
  File "C:\pinokio\api\Maestro.git\app\services\job_lifecycle.py", line 175, in call_with_sticky_interrupt
    return callable_(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\models\minimax_h3\minimax_h3_main.py", line 814, in generate
    prompt_embeds, text_tags = self.conditioner.forward_ref2va(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\models\minimax_h3\conditioner.py", line 293, in forward_ref2va
    image_embeds, image_deepstack = self.qwen.visual(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\mmgp\offload.py", line 3270, in check_change_module
    return previous_method(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\models\ideogram4\qwen3_vl_transformers.py", line 997, in forward
    hidden_states = blk(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\mmgp\offload.py", line 3248, in check_load_into_GPU_needed_other
    return previous_method(*args, **kwargs) # other
  File "C:\pinokio\api\Maestro.git\app\models\ideogram4\qwen3_vl_transformers.py", line 370, in forward
    self.norm1(hidden_states),
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\mmgp\offload.py", line 3248, in check_load_into_GPU_needed_other
    return previous_method(*args, **kwargs) # other
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\modules\normalization.py", line 217, in forward
    return F.layer_norm(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\functional.py", line 2901, in layer_norm
    return handle_torch_function(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\overrides.py", line 1721, in handle_torch_function
    result = mode.__torch_function__(public_api, types, args, kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\utils\_device.py", line 104, in __torch_function__
    return func(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\functional.py", line 2901, in layer_norm
    return handle_torch_function(
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\overrides.py", line 1743, in handle_torch_function
    result = torch_func_method(public_api, types, args, kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\_tensor.py", line 1668, in __torch_function__
    ret = func(*args, **kwargs)
  File "C:\pinokio\api\Maestro.git\app\env\lib\site-packages\torch\nn\functional.py", line 2910, in layer_norm
    return torch.layer_norm(
RuntimeError: expected scalar type Half but found Float

  [ERROR] "The generation of the video has encountered an error, please check your terminal for more information. 'expected scalar type Half but found Float'"

==================================================
Queue completed: 0/1 tasks in 10.2s