convert-ckpt-to-ggml.py : Not a directory: 'asr-whisper-medium-commonvoice-it.ckpt/encoder.json'

[raphael10-collab]

As described here: ggml-org/whisper.cpp#2731
using this converter : https://github.com/ggerganov/ggml/blob/master/examples/gpt-2/convert-ckpt-to-ggml.py

I get this error:

(.venv) raphy@raohy:~/whisper.cpp/models$ python3 convert-ckpt-to-ggml.py asr-whisper-medium-commonvoice-it.ckpt asr-whisper-commnvoice-it.bin
2025-01-13 12:56:24.374279: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2025-01-13 12:56:24.377338: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2025-01-13 12:56:24.388135: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1736769384.403006   81627 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1736769384.410188   81627 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2025-01-13 12:56:24.432522: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
Traceback (most recent call last):
  File "/home/raphy/whisper.cpp/models/convert-ckpt-to-ggml.py", line 68, in <module>
    with open(dir_model + "/encoder.json", "r", encoding="utf-8") as f:
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
NotADirectoryError: [Errno 20] Not a directory: 'asr-whisper-medium-commonvoice-it.ckpt/encoder.json'
(.venv) raphy@raohy:~/whisper.cpp/models$ 

Same error when trying to convert another ckpt model :

(.venv) raphy@raohy:~/whisper.cpp/models$ python3 convert-ckpt-to-ggml.py asr-crdnn-commonvoice-it.ckpt asr-crdnn-commonvoice-it.bin .
2025-01-13 16:49:18.924811: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2025-01-13 16:49:18.927827: I external/local_xla/xla/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2025-01-13 16:49:18.937363: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:477] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1736783358.952481   90633 cuda_dnn.cc:8310] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1736783358.957398   90633 cuda_blas.cc:1418] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2025-01-13 16:49:18.972919: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
Traceback (most recent call last):
  File "/home/raphy/whisper.cpp/models/convert-ckpt-to-ggml.py", line 68, in <module>
    with open(dir_model + "/encoder.json", "r", encoding="utf-8") as f:
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
NotADirectoryError: [Errno 20] Not a directory: 'asr-crdnn-commonvoice-it.ckpt/encoder.json'

OS: Ubuntu 24.04
Python 3.12.3

How to make it work?