Comfy installed but Error occurred when executing CLIPTextEncode

I have installed comfy using:

The UI loads no issues.

But when run a model it gives this error:

Error occurred when executing CLIPTextEncode:  CUDA error: operation not supported CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1. Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.   File "/stable-diffusion/execution.py", line 151, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) File "/stable-diffusion/execution.py", line 81, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) File "/stable-diffusion/execution.py", line 74, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) File "/stable-diffusion/nodes.py", line 58, in encode cond, pooled = clip.encode_from_tokens(tokens, return_pooled=True) File "/stable-diffusion/comfy/sd.py", line 135, in encode_from_tokens self.load_model() File "/stable-diffusion/comfy/sd.py", line 155, in load_model model_management.load_model_gpu(self.patcher) File "/stable-diffusion/comfy/model_management.py", line 467, in load_model_gpu return load_models_gpu([model]) File "/stable-diffusion/comfy/model_management.py", line 461, in load_models_gpu cur_loaded_model = loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights) File "/stable-diffusion/comfy/model_management.py", line 305, in model_load raise e File "/stable-diffusion/comfy/model_management.py", line 301, in model_load self.real_model = self.model.patch_model(device_to=patch_model_to, patch_weights=load_weights) File "/stable-diffusion/comfy/model_patcher.py", line 271, in patch_model self.model.to(device_to) File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1173, in to return self._apply(convert) File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply module._apply(fn) File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply module._apply(fn) File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply module._apply(fn) [Previous line repeated 2 more times] File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 804, in _apply param_applied = fn(param) File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1159, in convert return t.to(

full logs:

Mounted .cache
mkdir: created directory '/output/comfy'
Mounted comfy
mkdir: created directory '/data/config/comfy/input'
Mounted input
Total VRAM 24576 MB, total RAM 48135 MB
pytorch version: 2.3.0
Set vram state to: NORMAL_VRAM
Device: cuda:0 GRID P40-24Q : cudaMallocAsync
VAE dtype: torch.float32
Using pytorch cross attention
****** User settings have been changed to be stored on the server instead of browser storage. ******
****** For multi-user setups add the --multi-user CLI argument to enable multiple user profiles. ******
Adding extra search path checkpoints /data/models/Stable-diffusion
Adding extra search path configs /data/models/Stable-diffusion
Adding extra search path vae /data/models/VAE
Adding extra search path loras /data/models/Lora
Adding extra search path upscale_models /data/models/RealESRGAN

Adding extra search path hypernetworks /data/models/hypernetworks
Adding extra search path controlnet /data/models/ControlNet
Adding extra search path gligen /data/models/GLIGEN
Adding extra search path clip /data/models/CLIPEncoder
Adding extra search path embeddings /data/embeddings
Adding extra search path custom_nodes /data/config/comfy/custom_nodes

Import times for custom nodes:
   0.0 seconds: /stable-diffusion/custom_nodes/websocket_image_save.py

Starting server

To see the GUI go to: http://0.0.0.0:7860
got prompt
model_type EPS
Using pytorch attention in VAE
Using pytorch attention in VAE
/opt/conda/lib/python3.10/site-packages/transformers/tokenization_utils_base.py:1601: FutureWarning: `clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This behavior will be depracted in transformers v4.45, and will be then set to `False` by default. For more details check this issue: https://github.com/huggingface/transformers/issues/31884
  warnings.warn(
Requested to load SD1ClipModel
Loading 1 new model
!!! Exception during processing!!! CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Traceback (most recent call last):
  File "/stable-diffusion/execution.py", line 151, in recursive_execute
    output_data, output_ui = get_output_data(obj, input_data_all)
  File "/stable-diffusion/execution.py", line 81, in get_output_data
    return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
  File "/stable-diffusion/execution.py", line 74, in map_node_over_list
    results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
  File "/stable-diffusion/nodes.py", line 58, in encode
    cond, pooled = clip.encode_from_tokens(tokens, return_pooled=True)
  File "/stable-diffusion/comfy/sd.py", line 135, in encode_from_tokens
    self.load_model()
  File "/stable-diffusion/comfy/sd.py", line 155, in load_model
    model_management.load_model_gpu(self.patcher)
  File "/stable-diffusion/comfy/model_management.py", line 467, in load_model_gpu
    return load_models_gpu([model])
  File "/stable-diffusion/comfy/model_management.py", line 461, in load_models_gpu
    cur_loaded_model = loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights)
  File "/stable-diffusion/comfy/model_management.py", line 305, in model_load
    raise e
  File "/stable-diffusion/comfy/model_management.py", line 301, in model_load
    self.real_model = self.model.patch_model(device_to=patch_model_to, patch_weights=load_weights)
  File "/stable-diffusion/comfy/model_patcher.py", line 271, in patch_model
    self.model.to(device_to)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1173, in to
    return self._apply(convert)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  [Previous line repeated 2 more times]
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 804, in _apply
    param_applied = fn(param)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1159, in convert
    return t.to(
RuntimeError: CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.


Prompt executed in 20.59 seconds
got prompt
Requested to load SD1ClipModel
Loading 1 new model
!!! Exception during processing!!! CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Traceback (most recent call last):
  File "/stable-diffusion/execution.py", line 151, in recursive_execute
    output_data, output_ui = get_output_data(obj, input_data_all)
  File "/stable-diffusion/execution.py", line 81, in get_output_data
    return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
  File "/stable-diffusion/execution.py", line 74, in map_node_over_list
    results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
  File "/stable-diffusion/nodes.py", line 58, in encode
    cond, pooled = clip.encode_from_tokens(tokens, return_pooled=True)
  File "/stable-diffusion/comfy/sd.py", line 135, in encode_from_tokens
    self.load_model()
  File "/stable-diffusion/comfy/sd.py", line 155, in load_model
    model_management.load_model_gpu(self.patcher)
  File "/stable-diffusion/comfy/model_management.py", line 467, in load_model_gpu
    return load_models_gpu([model])
  File "/stable-diffusion/comfy/model_management.py", line 461, in load_models_gpu
    cur_loaded_model = loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights)
  File "/stable-diffusion/comfy/model_management.py", line 305, in model_load
    raise e
  File "/stable-diffusion/comfy/model_management.py", line 301, in model_load
    self.real_model = self.model.patch_model(device_to=patch_model_to, patch_weights=load_weights)
  File "/stable-diffusion/comfy/model_patcher.py", line 271, in patch_model
    self.model.to(device_to)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1173, in to
    return self._apply(convert)
  File "/stable-diffusion/execution.py", line 74, in map_node_over_list
    results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
  File "/stable-diffusion/nodes.py", line 58, in encode
    cond, pooled = clip.encode_from_tokens(tokens, return_pooled=True)
  File "/stable-diffusion/comfy/sd.py", line 135, in encode_from_tokens
    self.load_model()
  File "/stable-diffusion/comfy/sd.py", line 155, in load_model
    model_management.load_model_gpu(self.patcher)
  File "/stable-diffusion/comfy/model_management.py", line 467, in load_model_gpu
    return load_models_gpu([model])
  File "/stable-diffusion/comfy/model_management.py", line 461, in load_models_gpu
    cur_loaded_model = loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights)
  File "/stable-diffusion/comfy/model_management.py", line 305, in model_load
    raise e
  File "/stable-diffusion/comfy/model_management.py", line 301, in model_load
    self.real_model = self.model.patch_model(device_to=patch_model_to, patch_weights=load_weights)
  File "/stable-diffusion/comfy/model_patcher.py", line 271, in patch_model
    self.model.to(device_to)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1173, in to
    return self._apply(convert)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 779, in _apply
    module._apply(fn)
  [Previous line repeated 2 more times]
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 804, in _apply
    param_applied = fn(param)
  File "/opt/conda/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1159, in convert
    return t.to(
RuntimeError: CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.


Prompt executed in 0.08 seconds

my docker compose file:

stable-diffusion-base-download:
    build: ./stable-diffusion-webui-docker/services/download/
    image: stable-diffusion-base
    container_name: stable-diffusion-base
    environment:
      - PUID=${PUID:-1000}
      - PGID=${PGID:-1000}
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /etc/timezone:/etc/timezone:ro
      - ./stable-diffusion-webui-docker/data:/data

  automatic1111-stable-diffusion-webui:
    build: ./stable-diffusion-webui-docker/services/AUTOMATIC1111/
    image: automatic1111
    container_name: automatic1111
    environment:
      - PUID=${PUID:-1000}
      - PGID=${PGID:-1000}
      - CLI_ARGS=--allow-code --medvram --xformers --enable-insecure-extension-access --api
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /etc/timezone:/etc/timezone:ro
      - ./stable-diffusion-webui-docker/data:/data
      - ./stable-diffusion-webui-docker/output:/output
    stop_signal: SIGKILL
    tty: true
    deploy:
      resources:
        reservations:
          devices:
              - driver: nvidia
                device_ids: ['0']
                capabilities: [compute, utility]
    restart: unless-stopped
    networks:
      - traefik
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.stable-diffusion.rule=Host(`ai-images.local.example.com`)"
      - "traefik.http.routers.stable-diffusion.entrypoints=https"
      - "traefik.http.routers.stable-diffusion.tls=true"
      - "traefik.http.routers.stable-diffusion.tls.certresolver=cloudflare"
      - "traefik.http.services.stable-diffusion.loadbalancer.server.port=7860"
      - "traefik.http.routers.stable-diffusion.middlewares=default-headers@file"

  comfy-webui:
    build: ./stable-diffusion-webui-docker/services/comfy/
    image: comfy-webui
    container_name: comfy-webui
    environment:
      - PUID=${PUID:-1000}
      - PGID=${PGID:-1000}
      - CLI_ARGS=
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /etc/timezone:/etc/timezone:ro
      - ./stable-diffusion-webui-docker/data:/data
      - ./stable-diffusion-webui-docker/output:/output
    stop_signal: SIGKILL
    tty: true
    deploy:
      resources:
        reservations:
          devices:
              - driver: nvidia
                device_ids: ['0']
                capabilities: [compute, utility]
    restart: unless-stopped
    networks:
      - traefik
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.comfy.rule=Host(`comfy.local.example.com`)"
      - "traefik.http.routers.comfy.entrypoints=https"
      - "traefik.http.routers.comfy.tls=true"
      - "traefik.http.routers.comfy.tls.certresolver=cloudflare"
      - "traefik.http.services.comfy.loadbalancer.server.port=7860"
      - "traefik.http.routers.comfy.middlewares=default-headers@file"
      - ```

Similar issue here. Have you ever found a solution to this issue?

For anyone else that comes across this, I edited the docker-compose.yml to add --disable-cuda-malloc to the CLI_ARGS environment variable for the "comfy" service. I suppose I could've also used an override to achieve the same effect, but this is the route I took and it worked for me.

Navigation Menu

Repository navigation

Comfy installed but Error occurred when executing CLIPTextEncode #726

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Comfy installed but Error occurred when executing CLIPTextEncode #726

Description

Activity

fahadshery commented on Aug 23, 2024

fahadshery commented on Aug 23, 2024

lonewaffle commented on May 5, 2025

lonewaffle commented on May 5, 2025

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions