SAM 2.1 tiny as a Core AI .aimodel (float16, encode + decode), with README, licence and export script
a0b4f4d verified | cutout-sam2-tiny: SAM 2.1 hiera tiny, Copyright (c) Meta Platforms, Inc. and | |
| affiliates. Code: facebookresearch/sam2 at commit | |
| 2b90b9f5ceec907a1c18123530e92e794ad901a4. Weights: facebook/sam2.1-hiera-tiny | |
| at revision de431c4043854a71d8101e17995dfe596bf101a5 (sam2.1_hiera_tiny.pt). | |
| Both Apache License 2.0; see LICENSE. | |
| Changed for Core AI, in export_sam2.py: | |
| - The image predictor's own steps are traced as two entry points: `encode` | |
| (normalisation, image encoder, no-memory embedding) and `decode` (prompt | |
| encoder for one point, mask decoder, multimask output). | |
| - Weights and arithmetic are float16, except the prompt encoder, which stays | |
| float32; every input and output is float32. | |
| - The 256x256 mask logits are returned unclamped (the PyTorch predictor clamps | |
| them to +/-32; the host thresholds them). | |
| The trained weights are otherwise the authors', unmodified. | |