thanks to naver ❤

MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e28f91b488554653e2b46ddae9c78c1143e0bcb2e27d3e26cdb0b717f1568eb2
+size 2754910614

README.md

@@ -0,0 +1,57 @@
+---
+tags:
+- image-to-3d
+- pytorch_model_hub_mixin
+- model_hub_mixin
+library_name: mast3r
+repo_url: https://github.com/naver/mast3r
+---
+
+
+## Grounding Image Matching in 3D with MASt3R
+
+```bibtex
+@misc{mast3r_arxiv24,
+      title={Grounding Image Matching in 3D with MASt3R}, 
+      author={Vincent Leroy and Yohann Cabon and Jerome Revaud},
+      year={2024},
+      eprint={2406.09756},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
+}
+
+@inproceedings{dust3r_cvpr24,
+      title={DUSt3R: Geometric 3D Vision Made Easy}, 
+      author={Shuzhe Wang and Vincent Leroy and Yohann Cabon and Boris Chidlovskii and Jerome Revaud},
+      booktitle = {CVPR},
+      year = {2024}
+}
+```
+
+# License
+The code is distributed under the CC BY-NC-SA 4.0 License. See [LICENSE](https://github.com/naver/mast3r/blob/main/LICENSE) for more information.  
+For the checkpoints, make sure to agree to the license of all the public training datasets and base checkpoints we used, in addition to CC-BY-NC-SA 4.0.  
+The mapfree dataset license in particular is very restrictive. For more information, check [CHECKPOINTS_NOTICE](https://github.com/naver/mast3r/blob/main/CHECKPOINTS_NOTICE).
+
+# Model info
+
+Gihub page: https://github.com/naver/mast3r/
+
+| Modelname   | Training resolutions | Head | Encoder | Decoder |
+|-------------|----------------------|------|---------|---------|
+| MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_nonmetric | 512x384, 512x336, 512x288, 512x256, 512x160 | CatMLP+DPT | ViT-L | ViT-B |
+
+# How to use
+
+First, [install mast3r](https://github.com/naver/mast3r?tab=readme-ov-file#installation).
+To load the model:
+
+```python
+from mast3r.model import AsymmetricMASt3R
+import torch
+
+model = AsymmetricMASt3R.from_pretrained("naver/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_nonmetric")
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+```

config.json

@@ -0,0 +1,34 @@
+{
+  "conf_mode": [
+    "exp",
+    1,
+    Infinity
+  ],
+  "dec_depth": 12,
+  "dec_embed_dim": 768,
+  "dec_num_heads": 12,
+  "depth_mode": [
+    "exp",
+    -Infinity,
+    Infinity
+  ],
+  "desc_conf_mode": [
+    "exp",
+    0,
+    Infinity
+  ],
+  "desc_mode": "norm",
+  "enc_depth": 24,
+  "enc_embed_dim": 1024,
+  "enc_num_heads": 16,
+  "head_type": "catmlp+dpt",
+  "img_size": [
+    512,
+    512
+  ],
+  "landscape_only": false,
+  "output_mode": "pts3d+desc24",
+  "patch_embed_cls": "PatchEmbedDust3R",
+  "pos_embed": "RoPE100",
+  "two_confs": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a615eb05fa9db654050aa655945ee5696e7c6c1b7f93f1ee8c37249010f6feb
+size 2754661648