>> JAX FLAX nnx STRING_3D performance test using ViTTiny 6 layers on TPUv5e1 >> JAX STRING_3D performance test using ViTTiny 4 to 12 layers on TPUv8 >> JAX 2D_Positional_Encoding_Vision_Transformer >> Colab MLP Image Train and Inference Serial >> Colab MLP Train and Inference Parallel >> Colab MLP Train and Inference >> Colab EfficientSAM_example

>> Colab Image Segmentation Media-Segment-Depth-MLP drafting Segmentation from mediapipe holistic estimating Depth using BoostingMonocularDepth with an added MLP Model in JAX with mp_pose.Pose(static_image_mode=True, min_detection_confidence=0.2, model_complexity=2, enable_segmentation=True,) as pose: https://github.com/1kaiser/Media-Segment-Depth-MLP/blob/main/media/OUTPUT_FILE.mp4 OUTPUT_FILE.mp4 MediaPipe + EfficientSAM OUTPUT_FILE.1.mp4 with mp_pose.Pose(static_image_mode=False, min_detection_confidence=0.2, model_complexity=2, enable_segmentation=True,) as pose: https://github.com/1kaiser/Media-Segment-Depth-MLP/blob/main/media/OUTPUT_FILE%20(2).mp4 OUTPUT_FILE.2.mp4