You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since yesterday I get an nginx error (502 bad gateway), when trying to transcribe my audio file on my Ubuntu 22.04.3 LTS server. Before that everything worked fine. On my local machine (macbook with m3) everything works fine as well with laravel sail and docker.
The laravel application runs on:
Laravel 11.36.1
PHP 8.4
codewithkyrian/whisper.php 1.1.0
I set memory_limit to -1 without success. The audio file is only 30 seconds long.
I also noticed that the process stucks on line $whisper = Whisper::fromPretrained( (which i didn't change for a couple of days):
Since yesterday I get an nginx error (502 bad gateway), when trying to transcribe my audio file on my Ubuntu 22.04.3 LTS server. Before that everything worked fine. On my local machine (macbook with m3) everything works fine as well with laravel sail and docker.
The laravel application runs on:
I set memory_limit to -1 without success. The audio file is only 30 seconds long.
I also noticed that the process stucks on line
$whisper = Whisper::fromPretrained(
(which i didn't change for a couple of days):Any ideas ?
Sadly there are no errors in the nginx logs.
Below is the whole log output on the ubuntu server:
[2024-12-20 18:45:10] local.INFO: whisper_init_with_params_no_state: use gpu = 0
[2024-12-20 18:45:10] local.INFO: whisper_init_with_params_no_state: flash attn = 0
[2024-12-20 18:45:10] local.INFO: whisper_init_with_params_no_state: gpu_device = 0
[2024-12-20 18:45:10] local.INFO: whisper_init_with_params_no_state: dtw = 0
[2024-12-20 18:45:10] local.INFO: whisper_init_with_params_no_state: backends = 1
[2024-12-20 18:45:10] local.INFO: whisper_model_load: loading model
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_vocab = 51864
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_audio_ctx = 1500
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_audio_state = 768
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_audio_head = 12
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_audio_layer = 12
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_text_ctx = 448
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_text_state = 768
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_text_head = 12
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_text_layer = 12
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_mels = 80
[2024-12-20 18:45:10] local.INFO: whisper_model_load: ftype = 1
[2024-12-20 18:45:10] local.INFO: whisper_model_load: qntvr = 0
[2024-12-20 18:45:10] local.INFO: whisper_model_load: type = 3 (small)
[2024-12-20 18:45:10] local.INFO: whisper_model_load: adding 1607 extra tokens
[2024-12-20 18:45:10] local.INFO: whisper_model_load: n_langs = 99
The text was updated successfully, but these errors were encountered: