issue: notes + whisper +cuda failure #5058

Closed
opened 2025-11-11 16:11:05 -06:00 by GiteaMirror · 0 comments
Owner

Originally created by @CedricT1 on GitHub (May 6, 2025).

Check Existing Issues

  • I have searched the existing issues and discussions.
  • I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

0.6.6

Ollama Version (if applicable)

0.6.8

Operating System

Ubuntu 24.04

Browser (if applicable)

Firefox

Confirmation

  • I have read and followed all instructions in README.md.
  • I am using the latest version of both Open WebUI and Ollama.
  • I have included the browser console logs.
  • I have included the Docker container logs.
  • I have listed steps to reproduce the bug in detail.

Expected Behavior

Hello,

I have a problem, but only with the Docker image ghcr.io/open-webui/open-webui:cuda. The ghcr.io/open-webui/open-webui:main image does not have this issue.

Actual Behavior

open-webui.log

When I save a note, the microphone whisper does not work (I have tried both base and turbo). It tells me:

open-webui  | RuntimeError: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

This does not surprise me since I have two NVIDIA Quadro P5000 cards on Pascal architecture; they are old cards.

cedric@docker3:~/open-webui$ nvidia-smi
Tue May  6 09:34:02 2025       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.230.02             Driver Version: 535.230.02   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  Quadro P5000                   Off | 00000000:04:00.0 Off |                  Off |
| 26%   25C    P8               5W / 180W |   2482MiB / 16384MiB |      0%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
|   1  Quadro P5000                   Off | 00000000:91:00.0 Off |                  Off |
| 26%   32C    P8              12W / 180W |      4MiB / 16384MiB |      0%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A      2384      C   python                                      106MiB |
|    0   N/A  N/A     32275      C   /usr/local/bin/python                      2372MiB |
+---------------------------------------------------------------------------------------+
cedric@docker3:~/open-webui$ 

However, when in the audio settings I specify to use OpenAI (I have entered my OpenAI key), the problem persists, meaning it does not use the OpenAI API for whisper.

On the instance without CUDA, there is no problem, but open-webui is quite slow, especially in RAG.

Thank you for your work.

Best regards,

P.S. I have attached the log file.

Steps to Reproduce

  1. use image ghcr.io/open-webui/open-webui:cuda with a pascal architecture
  2. add a new note
  3. record a voice note
  4. the message off failure appar.

Logs & Screenshots

open-webui  | 2025-05-06 07:06:11.287 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22 HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:11.600 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22/update HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:22.037 | INFO     | open_webui.routers.files:upload_file:91 - file.content_type: audio/webm; codecs=opus - {}
open-webui  | 2025-05-06 07:06:22.059 | INFO     | open_webui.routers.audio:transcribe:499 - transcribe: /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm - {}
open-webui  | 2025-05-06 07:06:22.390 | INFO     | open_webui.routers.audio:convert_audio_to_wav:98 - Converted /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm to /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm - {}
open-webui  | 2025-05-06 07:06:24.243 | INFO     | open_webui.routers.retrieval:save_docs_to_vector_db:885 - save_docs_to_vector_db: document 2025 09:06.webm file-a7974075-0137-4c01-8963-4e4aba31bfe1 - {}
open-webui  | 2025-05-06 07:06:24.261 | INFO     | open_webui.routers.retrieval:save_docs_to_vector_db:968 - adding to collection file-a7974075-0137-4c01-8963-4e4aba31bfe1 - {}
Batches:   0%|          | 0/1 [00:00<?, ?it/s]
open-webui  | 2025-05-06 07:06:24.282 | ERROR    | open_webui.routers.retrieval:save_docs_to_vector_db:1009 - CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  |  - {}
open-webui  | Traceback (most recent call last):
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap
open-webui  |     self._bootstrap_inner()
open-webui  |     │    └ <function Thread._bootstrap_inner at 0x7188e1d74860>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner
open-webui  |     self.run()
open-webui  |     │    └ <function WorkerThread.run at 0x71875e94da80>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run
open-webui  |     result = context.run(func, *args)
open-webui  |              │       │   │      └ ()
open-webui  |              │       │   └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c...
open-webui  |              │       └ <method 'run' of '_contextvars.Context' objects>
open-webui  |              └ <_contextvars.Context object at 0x71875e967fc0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/files.py", line 141, in upload_file
open-webui  |     process_file(
open-webui  |     └ <function process_file at 0x71879ee04cc0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file
open-webui  |     result = save_docs_to_vector_db(
open-webui  |              └ <function save_docs_to_vector_db at 0x71879ef23ec0>
open-webui  | 
open-webui  | > File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db
open-webui  |     embeddings = embedding_function(
open-webui  |                  └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda>
open-webui  |     return lambda query, prefix=None, user=None: embedding_function.encode(
open-webui  |                   │                              │                  └ <function SentenceTransformer.encode at 0x71876059b2e0>
open-webui  |                   │                              └ SentenceTransformer(
open-webui  |                   │                                  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                   │                                 ...
open-webui  |                   └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."]
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode
open-webui  |     out_features = self.forward(features, **kwargs)
open-webui  |                    │    │       │           └ {}
open-webui  |                    │    │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                    │    │                   1010, 29536, 11733,  1...
open-webui  |                    │    └ <function SentenceTransformer.forward at 0x71876059b380>
open-webui  |                    └ SentenceTransformer(
open-webui  |                        (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                       ...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward
open-webui  |     input = module(input, **module_kwargs)
open-webui  |             │      │        └ {}
open-webui  |             │      └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |             │                  1010, 29536, 11733,  1...
open-webui  |             └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  ...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  ...
open-webui  |            └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward
open-webui  |     outputs = self.auto_model(**trans_features, **kwargs, return_dict=True)
open-webui  |               │                 │                 └ {}
open-webui  |               │                 └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |               │                             1010, 29536, 11733,  1...
open-webui  |               └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                  (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertModel.forward of BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, pa...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward
open-webui  |     embedding_output = self.embeddings(
open-webui  |                        └ BertModel(
open-webui  |                            (embeddings): BertEmbeddings(
open-webui  |                              (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                              (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position_embeddings): Embedding(512, 384)
open-webui  |                (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertEmbeddings.forward of BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward
open-webui  |     inputs_embeds = self.word_embeddings(input_ids)
open-webui  |                     │                    └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                     │                                1010, 29536, 11733,  1010,  3671, 13...
open-webui  |                     └ BertEmbeddings(
open-webui  |                         (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                         (position_embeddings): Embedding(512, 384)
open-webui  |                         (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Embedding(30522, 384, padding_idx=0)
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward
open-webui  |     return F.embedding(
open-webui  |            │ └ <function embedding at 0x7187c9696700>
open-webui  |            └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding
open-webui  |     return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
open-webui  |            │     │         │       │      │            │                   └ False
open-webui  |            │     │         │       │      │            └ False
open-webui  |            │     │         │       │      └ 0
open-webui  |            │     │         │       └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │     │         │                   1010, 29536, 11733,  1010,  3671, 13...
open-webui  |            │     │         └ <unprintable Parameter object>
open-webui  |            │     └ <built-in method embedding of type object at 0x7188d0ad8f80>
open-webui  |            └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'>
open-webui  | 
open-webui  | RuntimeError: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  | 
open-webui  | 2025-05-06 07:06:24.314 | ERROR    | open_webui.routers.retrieval:process_file:1187 - CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  |  - {}
open-webui  | Traceback (most recent call last):
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap
open-webui  |     self._bootstrap_inner()
open-webui  |     │    └ <function Thread._bootstrap_inner at 0x7188e1d74860>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner
open-webui  |     self.run()
open-webui  |     │    └ <function WorkerThread.run at 0x71875e94da80>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run
open-webui  |     result = context.run(func, *args)
open-webui  |              │       │   │      └ ()
open-webui  |              │       │   └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c...
open-webui  |              │       └ <method 'run' of '_contextvars.Context' objects>
open-webui  |              └ <_contextvars.Context object at 0x71875e967fc0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/files.py", line 141, in upload_file
open-webui  |     process_file(
open-webui  |     └ <function process_file at 0x71879ee04cc0>
open-webui  | 
open-webui  | > File "/app/backend/open_webui/routers/retrieval.py", line 1177, in process_file
open-webui  |     raise e
open-webui  |           └ RuntimeError('CUDA error: no kernel image is available for execution on the device\nCUDA kernel errors might be asynchronousl...
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file
open-webui  |     result = save_docs_to_vector_db(
open-webui  |              └ <function save_docs_to_vector_db at 0x71879ef23ec0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1010, in save_docs_to_vector_db
open-webui  |     raise e
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db
open-webui  |     embeddings = embedding_function(
open-webui  |                  └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda>
open-webui  |     return lambda query, prefix=None, user=None: embedding_function.encode(
open-webui  |                   │                              │                  └ <function SentenceTransformer.encode at 0x71876059b2e0>
open-webui  |                   │                              └ SentenceTransformer(
open-webui  |                   │                                  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                   │                                 ...
open-webui  |                   └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."]
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode
open-webui  |     out_features = self.forward(features, **kwargs)
open-webui  |                    │    │       │           └ {}
open-webui  |                    │    │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                    │    │                   1010, 29536, 11733,  1...
open-webui  |                    │    └ <function SentenceTransformer.forward at 0x71876059b380>
open-webui  |                    └ SentenceTransformer(
open-webui  |                        (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                       ...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward
open-webui  |     input = module(input, **module_kwargs)
open-webui  |             │      │        └ {}
open-webui  |             │      └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |             │                  1010, 29536, 11733,  1...
open-webui  |             └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  ...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  ...
open-webui  |            └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward
open-webui  |     outputs = self.auto_model(**trans_features, **kwargs, return_dict=True)
open-webui  |               │                 │                 └ {}
open-webui  |               │                 └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |               │                             1010, 29536, 11733,  1...
open-webui  |               └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                  (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertModel.forward of BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, pa...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward
open-webui  |     embedding_output = self.embeddings(
open-webui  |                        └ BertModel(
open-webui  |                            (embeddings): BertEmbeddings(
open-webui  |                              (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                              (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position_embeddings): Embedding(512, 384)
open-webui  |                (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertEmbeddings.forward of BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward
open-webui  |     inputs_embeds = self.word_embeddings(input_ids)
open-webui  |                     │                    └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                     │                                1010, 29536, 11733,  1010,  3671, 13...
open-webui  |                     └ BertEmbeddings(
open-webui  |                         (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                         (position_embeddings): Embedding(512, 384)
open-webui  |                         (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Embedding(30522, 384, padding_idx=0)
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward
open-webui  |     return F.embedding(
open-webui  |            │ └ <function embedding at 0x7187c9696700>
open-webui  |            └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding
open-webui  |     return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
open-webui  |            │     │         │       │      │            │                   └ False
open-webui  |            │     │         │       │      │            └ False
open-webui  |            │     │         │       │      └ 0
open-webui  |            │     │         │       └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │     │         │                   1010, 29536, 11733,  1010,  3671, 13...
open-webui  |            │     │         └ <unprintable Parameter object>
open-webui  |            │     └ <built-in method embedding of type object at 0x7188d0ad8f80>
open-webui  |            └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'>
open-webui  | 
open-webui  | RuntimeError: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  | 
open-webui  | 2025-05-06 07:06:24.333 | ERROR    | open_webui.routers.files:upload_file:159 - 400: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  |  - {}
open-webui  | Traceback (most recent call last):
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1177, in process_file
open-webui  |     raise e
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file
open-webui  |     result = save_docs_to_vector_db(
open-webui  |              └ <function save_docs_to_vector_db at 0x71879ef23ec0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1010, in save_docs_to_vector_db
open-webui  |     raise e
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db
open-webui  |     embeddings = embedding_function(
open-webui  |                  └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda>
open-webui  |     return lambda query, prefix=None, user=None: embedding_function.encode(
open-webui  |                   │                              │                  └ <function SentenceTransformer.encode at 0x71876059b2e0>
open-webui  |                   │                              └ SentenceTransformer(
open-webui  |                   │                                  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                   │                                 ...
open-webui  |                   └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."]
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode
open-webui  |     out_features = self.forward(features, **kwargs)
open-webui  |                    │    │       │           └ {}
open-webui  |                    │    │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                    │    │                   1010, 29536, 11733,  1...
open-webui  |                    │    └ <function SentenceTransformer.forward at 0x71876059b380>
open-webui  |                    └ SentenceTransformer(
open-webui  |                        (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
open-webui  |                       ...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward
open-webui  |     input = module(input, **module_kwargs)
open-webui  |             │      │        └ {}
open-webui  |             │      └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |             │                  1010, 29536, 11733,  1...
open-webui  |             └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  ...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ ({'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  ...
open-webui  |            └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward
open-webui  |     outputs = self.auto_model(**trans_features, **kwargs, return_dict=True)
open-webui  |               │                 │                 └ {}
open-webui  |               │                 └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |               │                             1010, 29536, 11733,  1...
open-webui  |               └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                  (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertModel.forward of BertModel(
open-webui  |                (embeddings): BertEmbeddings(
open-webui  |                  (word_embeddings): Embedding(30522, 384, pa...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward
open-webui  |     embedding_output = self.embeddings(
open-webui  |                        └ BertModel(
open-webui  |                            (embeddings): BertEmbeddings(
open-webui  |                              (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                              (position_embeddin...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │           │                   1010, 29536, 11733,  1...
open-webui  |            │    │           └ ()
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position_embeddings): Embedding(512, 384)
open-webui  |                (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {'input_ids': tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │             │                   1010, 29536, 11733,  1...
open-webui  |            │             └ ()
open-webui  |            └ <bound method BertEmbeddings.forward of BertEmbeddings(
open-webui  |                (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                (position...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward
open-webui  |     inputs_embeds = self.word_embeddings(input_ids)
open-webui  |                     │                    └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |                     │                                1010, 29536, 11733,  1010,  3671, 13...
open-webui  |                     └ BertEmbeddings(
open-webui  |                         (word_embeddings): Embedding(30522, 384, padding_idx=0)
open-webui  |                         (position_embeddings): Embedding(512, 384)
open-webui  |                         (tok...
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
open-webui  |     return self._call_impl(*args, **kwargs)
open-webui  |            │    │           │       └ {}
open-webui  |            │    │           └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │    │                       1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            │    └ <function Module._call_impl at 0x7187cd9609a0>
open-webui  |            └ Embedding(30522, 384, padding_idx=0)
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
open-webui  |     return forward_call(*args, **kwargs)
open-webui  |            │             │       └ {}
open-webui  |            │             └ (tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │                         1010, 29536, 11733,  1010,  3671, 1...
open-webui  |            └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward
open-webui  |     return F.embedding(
open-webui  |            │ └ <function embedding at 0x7187c9696700>
open-webui  |            └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding
open-webui  |     return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
open-webui  |            │     │         │       │      │            │                   └ False
open-webui  |            │     │         │       │      │            └ False
open-webui  |            │     │         │       │      └ 0
open-webui  |            │     │         │       └ tensor([[  101,  9686, 15816,  1040,  1005,  4372,  2890, 24063, 28578,  4765,
open-webui  |            │     │         │                   1010, 29536, 11733,  1010,  3671, 13...
open-webui  |            │     │         └ <unprintable Parameter object>
open-webui  |            │     └ <built-in method embedding of type object at 0x7188d0ad8f80>
open-webui  |            └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'>
open-webui  | 
open-webui  | RuntimeError: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  | 
open-webui  | 
open-webui  | 
open-webui  | During handling of the above exception, another exception occurred:
open-webui  | 
open-webui  | 
open-webui  | Traceback (most recent call last):
open-webui  | 
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap
open-webui  |     self._bootstrap_inner()
open-webui  |     │    └ <function Thread._bootstrap_inner at 0x7188e1d74860>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner
open-webui  |     self.run()
open-webui  |     │    └ <function WorkerThread.run at 0x71875e94da80>
open-webui  |     └ <WorkerThread(AnyIO worker thread, started 124825683097280)>
open-webui  |   File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run
open-webui  |     result = context.run(func, *args)
open-webui  |              │       │   │      └ ()
open-webui  |              │       │   └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c...
open-webui  |              │       └ <method 'run' of '_contextvars.Context' objects>
open-webui  |              └ <_contextvars.Context object at 0x71875e967fc0>
open-webui  | 
open-webui  | > File "/app/backend/open_webui/routers/files.py", line 141, in upload_file
open-webui  |     process_file(
open-webui  |     └ <function process_file at 0x71879ee04cc0>
open-webui  | 
open-webui  |   File "/app/backend/open_webui/routers/retrieval.py", line 1194, in process_file
open-webui  |     raise HTTPException(
open-webui  |           └ <class 'fastapi.exceptions.HTTPException'>
open-webui  | 
open-webui  | fastapi.exceptions.HTTPException: 400: CUDA error: no kernel image is available for execution on the device
open-webui  | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
open-webui  | For debugging consider passing CUDA_LAUNCH_BLOCKING=1
open-webui  | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
open-webui  | 
open-webui  | 2025-05-06 07:06:24.356 | ERROR    | open_webui.routers.files:upload_file:160 - Error processing file: a7974075-0137-4c01-8963-4e4aba31bfe1 - {}
open-webui  | 2025-05-06 07:06:24.360 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/files/ HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:24.612 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22/update HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:26.149 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/08d886ce-5e7b-4ac5-b6f0-e7bc89eed521 HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:26.287 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/0664ef73-79a1-473f-b376-88e854805d61 HTTP/1.1" 200 - {}
open-webui  | 2025-05-06 07:06:27.412 | INFO     | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/08d886ce-5e7b-4ac5-b6f0-e7bc89eed521 HTTP/1.1" 200 - {}
^C
cedric@docker3:~/open-webui$ 


Additional Information

No response

Originally created by @CedricT1 on GitHub (May 6, 2025). ### Check Existing Issues - [x] I have searched the existing issues and discussions. - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version 0.6.6 ### Ollama Version (if applicable) 0.6.8 ### Operating System Ubuntu 24.04 ### Browser (if applicable) Firefox ### Confirmation - [x] I have read and followed all instructions in `README.md`. - [x] I am using the latest version of **both** Open WebUI and Ollama. - [x] I have included the browser console logs. - [x] I have included the Docker container logs. - [x] I have listed steps to reproduce the bug in detail. ### Expected Behavior Hello, I have a problem, but only with the Docker image ghcr.io/open-webui/open-webui:cuda. The ghcr.io/open-webui/open-webui:main image does not have this issue. ### Actual Behavior [open-webui.log](https://github.com/user-attachments/files/20054366/open-webui.log) When I save a note, the microphone whisper does not work (I have tried both base and turbo). It tells me: ``` open-webui | RuntimeError: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. ``` This does not surprise me since I have two NVIDIA Quadro P5000 cards on Pascal architecture; they are old cards. ``` cedric@docker3:~/open-webui$ nvidia-smi Tue May 6 09:34:02 2025 +---------------------------------------------------------------------------------------+ | NVIDIA-SMI 535.230.02 Driver Version: 535.230.02 CUDA Version: 12.2 | |-----------------------------------------+----------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+======================+======================| | 0 Quadro P5000 Off | 00000000:04:00.0 Off | Off | | 26% 25C P8 5W / 180W | 2482MiB / 16384MiB | 0% Default | | | | N/A | +-----------------------------------------+----------------------+----------------------+ | 1 Quadro P5000 Off | 00000000:91:00.0 Off | Off | | 26% 32C P8 12W / 180W | 4MiB / 16384MiB | 0% Default | | | | N/A | +-----------------------------------------+----------------------+----------------------+ +---------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=======================================================================================| | 0 N/A N/A 2384 C python 106MiB | | 0 N/A N/A 32275 C /usr/local/bin/python 2372MiB | +---------------------------------------------------------------------------------------+ cedric@docker3:~/open-webui$ ``` However, when in the audio settings I specify to use OpenAI (I have entered my OpenAI key), the problem persists, meaning it does not use the OpenAI API for whisper. On the instance without CUDA, there is no problem, but open-webui is quite slow, especially in RAG. Thank you for your work. Best regards, P.S. I have attached the log file. ### Steps to Reproduce 1. use image ghcr.io/open-webui/open-webui:cuda with a pascal architecture 2. add a new note 3. record a voice note 4. the message off failure appar. ### Logs & Screenshots ``` open-webui | 2025-05-06 07:06:11.287 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22 HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:11.600 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22/update HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:22.037 | INFO | open_webui.routers.files:upload_file:91 - file.content_type: audio/webm; codecs=opus - {} open-webui | 2025-05-06 07:06:22.059 | INFO | open_webui.routers.audio:transcribe:499 - transcribe: /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm - {} open-webui | 2025-05-06 07:06:22.390 | INFO | open_webui.routers.audio:convert_audio_to_wav:98 - Converted /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm to /app/backend/data/uploads/a7974075-0137-4c01-8963-4e4aba31bfe1_2025 09:06.webm - {} open-webui | 2025-05-06 07:06:24.243 | INFO | open_webui.routers.retrieval:save_docs_to_vector_db:885 - save_docs_to_vector_db: document 2025 09:06.webm file-a7974075-0137-4c01-8963-4e4aba31bfe1 - {} open-webui | 2025-05-06 07:06:24.261 | INFO | open_webui.routers.retrieval:save_docs_to_vector_db:968 - adding to collection file-a7974075-0137-4c01-8963-4e4aba31bfe1 - {} Batches: 0%| | 0/1 [00:00<?, ?it/s] open-webui | 2025-05-06 07:06:24.282 | ERROR | open_webui.routers.retrieval:save_docs_to_vector_db:1009 - CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | - {} open-webui | Traceback (most recent call last): open-webui | open-webui | File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap open-webui | self._bootstrap_inner() open-webui | │ └ <function Thread._bootstrap_inner at 0x7188e1d74860> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner open-webui | self.run() open-webui | │ └ <function WorkerThread.run at 0x71875e94da80> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run open-webui | result = context.run(func, *args) open-webui | │ │ │ └ () open-webui | │ │ └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c... open-webui | │ └ <method 'run' of '_contextvars.Context' objects> open-webui | └ <_contextvars.Context object at 0x71875e967fc0> open-webui | open-webui | File "/app/backend/open_webui/routers/files.py", line 141, in upload_file open-webui | process_file( open-webui | └ <function process_file at 0x71879ee04cc0> open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file open-webui | result = save_docs_to_vector_db( open-webui | └ <function save_docs_to_vector_db at 0x71879ef23ec0> open-webui | open-webui | > File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db open-webui | embeddings = embedding_function( open-webui | └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00> open-webui | open-webui | File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda> open-webui | return lambda query, prefix=None, user=None: embedding_function.encode( open-webui | │ │ └ <function SentenceTransformer.encode at 0x71876059b2e0> open-webui | │ └ SentenceTransformer( open-webui | │ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | │ ... open-webui | └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."] open-webui | open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode open-webui | out_features = self.forward(features, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ <function SentenceTransformer.forward at 0x71876059b380> open-webui | └ SentenceTransformer( open-webui | (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | ... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward open-webui | input = module(input, **module_kwargs) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, ... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, ... open-webui | └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward open-webui | outputs = self.auto_model(**trans_features, **kwargs, return_dict=True) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertModel.forward of BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, pa... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward open-webui | embedding_output = self.embeddings( open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertEmbeddings.forward of BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward open-webui | inputs_embeds = self.word_embeddings(input_ids) open-webui | │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Embedding(30522, 384, padding_idx=0) open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward open-webui | return F.embedding( open-webui | │ └ <function embedding at 0x7187c9696700> open-webui | └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding open-webui | return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse) open-webui | │ │ │ │ │ │ └ False open-webui | │ │ │ │ │ └ False open-webui | │ │ │ │ └ 0 open-webui | │ │ │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | │ │ └ <unprintable Parameter object> open-webui | │ └ <built-in method embedding of type object at 0x7188d0ad8f80> open-webui | └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'> open-webui | open-webui | RuntimeError: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | open-webui | 2025-05-06 07:06:24.314 | ERROR | open_webui.routers.retrieval:process_file:1187 - CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | - {} open-webui | Traceback (most recent call last): open-webui | open-webui | File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap open-webui | self._bootstrap_inner() open-webui | │ └ <function Thread._bootstrap_inner at 0x7188e1d74860> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner open-webui | self.run() open-webui | │ └ <function WorkerThread.run at 0x71875e94da80> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run open-webui | result = context.run(func, *args) open-webui | │ │ │ └ () open-webui | │ │ └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c... open-webui | │ └ <method 'run' of '_contextvars.Context' objects> open-webui | └ <_contextvars.Context object at 0x71875e967fc0> open-webui | open-webui | File "/app/backend/open_webui/routers/files.py", line 141, in upload_file open-webui | process_file( open-webui | └ <function process_file at 0x71879ee04cc0> open-webui | open-webui | > File "/app/backend/open_webui/routers/retrieval.py", line 1177, in process_file open-webui | raise e open-webui | └ RuntimeError('CUDA error: no kernel image is available for execution on the device\nCUDA kernel errors might be asynchronousl... open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file open-webui | result = save_docs_to_vector_db( open-webui | └ <function save_docs_to_vector_db at 0x71879ef23ec0> open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1010, in save_docs_to_vector_db open-webui | raise e open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db open-webui | embeddings = embedding_function( open-webui | └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00> open-webui | open-webui | File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda> open-webui | return lambda query, prefix=None, user=None: embedding_function.encode( open-webui | │ │ └ <function SentenceTransformer.encode at 0x71876059b2e0> open-webui | │ └ SentenceTransformer( open-webui | │ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | │ ... open-webui | └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."] open-webui | open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode open-webui | out_features = self.forward(features, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ <function SentenceTransformer.forward at 0x71876059b380> open-webui | └ SentenceTransformer( open-webui | (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | ... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward open-webui | input = module(input, **module_kwargs) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, ... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, ... open-webui | └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward open-webui | outputs = self.auto_model(**trans_features, **kwargs, return_dict=True) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertModel.forward of BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, pa... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward open-webui | embedding_output = self.embeddings( open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertEmbeddings.forward of BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward open-webui | inputs_embeds = self.word_embeddings(input_ids) open-webui | │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Embedding(30522, 384, padding_idx=0) open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward open-webui | return F.embedding( open-webui | │ └ <function embedding at 0x7187c9696700> open-webui | └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding open-webui | return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse) open-webui | │ │ │ │ │ │ └ False open-webui | │ │ │ │ │ └ False open-webui | │ │ │ │ └ 0 open-webui | │ │ │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | │ │ └ <unprintable Parameter object> open-webui | │ └ <built-in method embedding of type object at 0x7188d0ad8f80> open-webui | └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'> open-webui | open-webui | RuntimeError: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | open-webui | 2025-05-06 07:06:24.333 | ERROR | open_webui.routers.files:upload_file:159 - 400: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | - {} open-webui | Traceback (most recent call last): open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1177, in process_file open-webui | raise e open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1149, in process_file open-webui | result = save_docs_to_vector_db( open-webui | └ <function save_docs_to_vector_db at 0x71879ef23ec0> open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1010, in save_docs_to_vector_db open-webui | raise e open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 986, in save_docs_to_vector_db open-webui | embeddings = embedding_function( open-webui | └ <function get_embedding_function.<locals>.<lambda> at 0x71872e198e00> open-webui | open-webui | File "/app/backend/open_webui/retrieval/utils.py", line 391, in <lambda> open-webui | return lambda query, prefix=None, user=None: embedding_function.encode( open-webui | │ │ └ <function SentenceTransformer.encode at 0x71876059b2e0> open-webui | │ └ SentenceTransformer( open-webui | │ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | │ ... open-webui | └ ["Essai d'enregistrement, voilà, normalement on verra s'il détecte la voix, et puis c'est tout."] open-webui | open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 685, in encode open-webui | out_features = self.forward(features, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ <function SentenceTransformer.forward at 0x71876059b380> open-webui | └ SentenceTransformer( open-webui | (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | ... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 758, in forward open-webui | input = module(input, **module_kwargs) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, ... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ ({'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, ... open-webui | └ <bound method Transformer.forward of Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: Ber... open-webui | File "/usr/local/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 442, in forward open-webui | outputs = self.auto_model(**trans_features, **kwargs, return_dict=True) open-webui | │ │ └ {} open-webui | │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1... open-webui | └ Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertModel.forward of BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, pa... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 1080, in forward open-webui | embedding_output = self.embeddings( open-webui | └ BertModel( open-webui | (embeddings): BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddin... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1... open-webui | │ │ └ () open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {'input_ids': tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1... open-webui | │ └ () open-webui | └ <bound method BertEmbeddings.forward of BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position... open-webui | File "/usr/local/lib/python3.11/site-packages/transformers/models/bert/modeling_bert.py", line 211, in forward open-webui | inputs_embeds = self.word_embeddings(input_ids) open-webui | │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | └ BertEmbeddings( open-webui | (word_embeddings): Embedding(30522, 384, padding_idx=0) open-webui | (position_embeddings): Embedding(512, 384) open-webui | (tok... open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl open-webui | return self._call_impl(*args, **kwargs) open-webui | │ │ │ └ {} open-webui | │ │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | │ └ <function Module._call_impl at 0x7187cd9609a0> open-webui | └ Embedding(30522, 384, padding_idx=0) open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl open-webui | return forward_call(*args, **kwargs) open-webui | │ │ └ {} open-webui | │ └ (tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ 1010, 29536, 11733, 1010, 3671, 1... open-webui | └ <bound method Embedding.forward of Embedding(30522, 384, padding_idx=0)> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/modules/sparse.py", line 190, in forward open-webui | return F.embedding( open-webui | │ └ <function embedding at 0x7187c9696700> open-webui | └ <module 'torch.nn.functional' from '/usr/local/lib/python3.11/site-packages/torch/nn/functional.py'> open-webui | File "/usr/local/lib/python3.11/site-packages/torch/nn/functional.py", line 2551, in embedding open-webui | return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse) open-webui | │ │ │ │ │ │ └ False open-webui | │ │ │ │ │ └ False open-webui | │ │ │ │ └ 0 open-webui | │ │ │ └ tensor([[ 101, 9686, 15816, 1040, 1005, 4372, 2890, 24063, 28578, 4765, open-webui | │ │ │ 1010, 29536, 11733, 1010, 3671, 13... open-webui | │ │ └ <unprintable Parameter object> open-webui | │ └ <built-in method embedding of type object at 0x7188d0ad8f80> open-webui | └ <module 'torch' from '/usr/local/lib/python3.11/site-packages/torch/__init__.py'> open-webui | open-webui | RuntimeError: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | open-webui | open-webui | open-webui | During handling of the above exception, another exception occurred: open-webui | open-webui | open-webui | Traceback (most recent call last): open-webui | open-webui | File "/usr/local/lib/python3.11/threading.py", line 1002, in _bootstrap open-webui | self._bootstrap_inner() open-webui | │ └ <function Thread._bootstrap_inner at 0x7188e1d74860> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/threading.py", line 1045, in _bootstrap_inner open-webui | self.run() open-webui | │ └ <function WorkerThread.run at 0x71875e94da80> open-webui | └ <WorkerThread(AnyIO worker thread, started 124825683097280)> open-webui | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 967, in run open-webui | result = context.run(func, *args) open-webui | │ │ │ └ () open-webui | │ │ └ functools.partial(<function upload_file at 0x7187ac0d4400>, user=UserModel(id='9978dec7-0b8d-46e6-abd7-3d8098fee68b', name='c... open-webui | │ └ <method 'run' of '_contextvars.Context' objects> open-webui | └ <_contextvars.Context object at 0x71875e967fc0> open-webui | open-webui | > File "/app/backend/open_webui/routers/files.py", line 141, in upload_file open-webui | process_file( open-webui | └ <function process_file at 0x71879ee04cc0> open-webui | open-webui | File "/app/backend/open_webui/routers/retrieval.py", line 1194, in process_file open-webui | raise HTTPException( open-webui | └ <class 'fastapi.exceptions.HTTPException'> open-webui | open-webui | fastapi.exceptions.HTTPException: 400: CUDA error: no kernel image is available for execution on the device open-webui | CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. open-webui | For debugging consider passing CUDA_LAUNCH_BLOCKING=1 open-webui | Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. open-webui | open-webui | 2025-05-06 07:06:24.356 | ERROR | open_webui.routers.files:upload_file:160 - Error processing file: a7974075-0137-4c01-8963-4e4aba31bfe1 - {} open-webui | 2025-05-06 07:06:24.360 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/files/ HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:24.612 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "POST /api/v1/notes/c2a3bb41-31db-4062-abca-301c1d1b3c22/update HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:26.149 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/08d886ce-5e7b-4ac5-b6f0-e7bc89eed521 HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:26.287 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/0664ef73-79a1-473f-b376-88e854805d61 HTTP/1.1" 200 - {} open-webui | 2025-05-06 07:06:27.412 | INFO | uvicorn.protocols.http.httptools_impl:send:476 - 83.173.222.26:0 - "GET /api/v1/chats/08d886ce-5e7b-4ac5-b6f0-e7bc89eed521 HTTP/1.1" 200 - {} ^C cedric@docker3:~/open-webui$ ``` ### Additional Information _No response_
GiteaMirror added the bug label 2025-11-11 16:11:05 -06:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#5058