issue on colab, was working fine 50 days ago, today when I run the same notebook again, I get this error.

[irixium]

RuntimeError Traceback (most recent call last) in <cell line: 0>() 19 # Load some sample pictures and learn how to recognize them. 20 lmm_image = face_recognition.load_image_file("/content/persons/Anshuman/IMG20230430023329.jpg") ---> 21 lmm_face_encoding = face_recognition.face_encodings(lmm_image)[0] 22 23

1 frames /usr/local/lib/python3.11/dist-packages/face_recognition/api.py in (.0) 212 """ 213 raw_landmarks = _raw_face_landmarks(face_image, known_face_locations, model) --> 214 return [np.array(face_encoder.compute_face_descriptor(face_image, raw_landmark_set, num_jitters)) for raw_landmark_set in raw_landmarks] 215 216

RuntimeError: Error while calling cudaOccupancyMaxPotentialBlockSize(&num_blocks,&num_threads,K) in file /root/.cache/uv/sdists-v7/pypi/dlib/19.24.2/4_6PNx5_2RaZQs-tY1P75/src/dlib/cuda/cuda_utils.h:186. code: 222, reason: the provided PTX was compiled with an unsupported toolchain.

I am running on T4 Colab. The exact same notebook worked fine without error 50 days ago, but it gives this error now. What can I do to fix

[diogoviannaaraujo]

getting the same problem here

[KjelleJ]

The same problem for me, created an issue on Colab. Could be a Colab problem? Is face_recognition maintained?

[Mat198]

Installing the binaries of dlib worked for me:

pip install dlib-bin==19.24.6 face-recognition==1.3.0 Pillow