Comment by K0balt 8 hours ago Anyone know how many parameters? 2 comments K0balt Reply UnboundedContex 4 hours ago Approx ~131 million total. Made up of estimated:102M - modified BERT-base-Chinese text encoder26M - 3D U-Net-style vision/anatomy encoder2.8M - projection layers, anatomy-specific projections, query tokens and attention layerWritten to run on something like an H100 though- as the CT Scan data is quite large. anon373839 6 hours ago The model weights are only ~5GB, so this is small.
UnboundedContex 4 hours ago Approx ~131 million total. Made up of estimated:102M - modified BERT-base-Chinese text encoder26M - 3D U-Net-style vision/anatomy encoder2.8M - projection layers, anatomy-specific projections, query tokens and attention layerWritten to run on something like an H100 though- as the CT Scan data is quite large.
Approx ~131 million total. Made up of estimated:
102M - modified BERT-base-Chinese text encoder
26M - 3D U-Net-style vision/anatomy encoder
2.8M - projection layers, anatomy-specific projections, query tokens and attention layer
Written to run on something like an H100 though- as the CT Scan data is quite large.
The model weights are only ~5GB, so this is small.