Error when running benchmark

[Derjyn]

Getting this error in the console when running a benchmark (and performance shows error/error/error):

Startup time: 19.9s (initial startup: 0.1s, prepare environment: 2.6s, import torch: 4.0s, import gradio: 1.6s, setup paths: 1.0s, initialize shared: 0.2s, other imports: 0.9s, opts onchange: 0.3s, setup codeformer: 0.1s, list SD models: 0.3s, load scripts: 3.6s, scripts before_ui_callback: 3.1s, create ui: 1.3s, gradio launch: 0.2s, app_started_callback: 0.4s).
  0%|                                                                                        | 0/20 [00:04<?, ?it/s]
SD-System-Info benchmark error: 1 'KDiffusionSampler' object has no attribute 'trajectory_enable'
  0%|                                                                                        | 0/50 [00:00<?, ?it/s]
SD-System-Info benchmark error: 1 'KDiffusionSampler' object has no attribute 'trajectory_enable'
  0%|                                                                                        | 0/50 [00:00<?, ?it/s]
SD-System-Info benchmark error: 2 'KDiffusionSampler' object has no attribute 'trajectory_enable'
  0%|                                                                                        | 0/50 [00:00<?, ?it/s]
SD-System-Info benchmark error: 4 'KDiffusionSampler' object has no attribute 'trajectory_enable'

[vladmandic]

provide information - i don't know what base app you're using, backend, version, nothing. system info shows it all.

[Derjyn]

More console inspection when testing this returned this (hopefully) more helpful error:

SD-System-Info benchmark error: 1 Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument index in method wrapper_CUDA__index_select)

Also, here is a more complete system info dump:

{
  "date": "Wed Feb  7 18:46:32 2024",
  "timestamp": "18:46:34",
  "uptime": "Wed Feb  7 09:42:27 2024",
  "version": {},
  "torch": "2.0.1+cu118 autocast  half",
  "gpu": {
    "device": "NVIDIA GeForce RTX 3070 (1) (compute_37) (8, 6)",
    "cuda": "11.8",
    "cudnn": 8700,
    "driver": "546.01"
  },
  "state": {
    "started": "Wed Feb  7 18:45:29 2024",
    "step": "0 / 40",
    "jobs": "0 / 1",
    "flags": "",
    "job": "benchmark",
    "text-info": null
  },
  "memory": {
    "ram": {
      "free": 27.49,
      "used": 4.45,
      "total": 31.94
    },
    "gpu": {
      "free": 5.76,
      "used": 2.24,
      "total": 8
    },
    "gpu-active": {
      "current": 0.86,
      "peak": 3.43
    },
    "gpu-allocated": {
      "current": 0.86,
      "peak": 3.43
    },
    "gpu-reserved": {
      "current": 1.09,
      "peak": 4.05
    },
    "gpu-inactive": {
      "current": 0.22,
      "peak": 1.03
    },
    "events": {
      "retries": 0,
      "oom": 0
    },
    "utilization": 0
  },
  "optimizations": [
    "none"
  ],
  "libs": {
    "xformers": "0.0.20",
    "diffusers": "0.25.0",
    "transformers": "4.30.2"
  },
  "repos": {
    "Stable Diffusion": "[cf1d67a] 2023-03-25",
    "Stable Diffusion XL": "[45c443b] 2023-07-26",
    "CodeFormer": "[c5b4593] 2022-09-09",
    "BLIP": "[48211a1] 2022-06-07",
    "k_diffusion": "[ab527a9] 2023-08-12"
  },
  "device": {
    "active": "cuda",
    "dtype": "torch.float16",
    "vae": "torch.float16",
    "unet": "torch.float16"
  },
  "model": {
    "configured": {
      "base": "Kerberos\\KerberosX.safetensors",
      "refiner": "",
      "vae": "Kerberos_VAE.ckpt"
    },
    "loaded": {
      "base": "..\\..\\Assets\\Models\\Kerberos\\KerberosX.safetensors",
      "refiner": "",
      "vae": "..\\..\\Assets\\VAE\\Kerberos_VAE.ckpt"
    }
  },
  "schedulers": [
    "DDIM",
    "DPM adaptive",
    "DPM fast",
    "DPM++ 2M",
    "DPM++ 2M Karras",
    "DPM++ 2M SDE",
    "DPM++ 2M SDE Exponential",
    "DPM++ 2M SDE Heun",
    "DPM++ 2M SDE Heun Exponential",
    "DPM++ 2M SDE Heun Karras",
    "DPM++ 2M SDE Karras",
    "DPM++ 2S a",
    "DPM++ 2S a Karras",
    "DPM++ 3M SDE",
    "DPM++ 3M SDE Exponential",
    "DPM++ 3M SDE Karras",
    "DPM++ SDE",
    "DPM++ SDE Karras",
    "DPM2",
    "DPM2 Karras",
    "DPM2 a",
    "DPM2 a Karras",
    "Euler",
    "Euler a",
    "Heun",
    "LMS",
    "LMS Karras",
    "PLMS",
    "Restart",
    "UniPC"
  ],
  "extensions": [
    "LDSR (enabled builtin)",
    "Lora (enabled builtin)",
    "ScuNET (enabled builtin)",
    "SwinIR (enabled builtin)",
    "animatediff-cli-prompt-travel (enabled)",
    "canvas-zoom-and-pan (enabled builtin)",
    "ebsynth_utility (enabled)",
    "extra-options-section (enabled builtin)",
    "hypertile (enabled builtin)",
    "multidiffusion-upscaler-for-automatic1111 (enabled)",
    "prompt-bracket-checker (enabled builtin)",
    "sd-dynamic-thresholding (enabled)",
    "sd-extension-system-info (enabled)",
    "sd-webui-animatediff (enabled)",
    "sd-webui-bayesian-merger (enabled)",
    "sd-webui-cads (enabled)",
    "sd-webui-controlnet (enabled)",
    "sd-webui-cutoff (enabled)",
    "sd-webui-diffusion-cg (enabled)",
    "sd-webui-memory-release (enabled)",
    "sd-webui-model-converter (enabled)",
    "sd-webui-mosaic-outpaint (enabled)",
    "sd-webui-mov2mov (enabled)",
    "sd-webui-negpip (enabled)",
    "sd-webui-regional-prompter (enabled)",
    "sd-webui-resharpen (enabled)",
    "sd-webui-segment-anything (enabled)",
    "sd-webui-semantic-guidance (enabled)",
    "sd_webui_SAG (enabled)",
    "sdweb-merge-block-weighted-gui (enabled)",
    "stable-diffusion-webui-model-toolkit (enabled)",
    "uddetailer (enabled)",
    "zccz (enabled)"
  ],
  "platform": {
    "arch": "AMD64",
    "cpu": "Intel64 Family 6 Model 94 Stepping 3, GenuineIntel",
    "system": "Windows",
    "release": "Windows-10-10.0.19045-SP0",
    "python": "3.11.1"
  },
  "crossattention": "sdp - scaled dot product",
  "backend": "",
  "pipeline": ""
}

[vladmandic]

error Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument index in method wrapper_CUDA__index_select) is coming from underlying app and is a general problem there, not specific to system info benchmark.

you never said which app and which version of that app are you using???

[Derjyn]

Automatic1111 SD Web UI, 1.7.0

[vladmandic]

both of the errors are in core app, nothing to do with system info:

• `'KDiffusionSampler' object has no attribute 'trajectory_enable'
• Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0!

benchmark triggers simple txt2img api call, it doesn't set any special settings or interact with app engine.

i really cannot tell you more than that.