Release MeshAxiom-Base Exact weights

.gitattributes

@@ -0,0 +1 @@
+weights/*.pt filter=lfs diff=lfs merge=lfs -text

LICENSE.md

@@ -0,0 +1,9 @@
+# MeshAxiom licensing
+
+The MeshAxiom code and Base model weights are publicly available. The authors'
+standalone code/weight license has not yet been assigned. This release note does
+not assign an MIT, Apache-2.0 or other general license.
+
+The three EMA checkpoints contain frozen DINOv3 backbone tensors. Their use and
+redistribution remain subject to `external/dinov3/LICENSE.md`. Preserve that license
+and all applicable third-party notices.

README.md

@@ -0,0 +1,74 @@
+---
+library_name: meshaxiom
+pipeline_tag: image-to-3d
+tags:
+  - meshaxiom
+  - image-to-mesh
+  - mesh
+  - 3d
+  - exact
+language:
+  - en
+  - zh
+---
+
+# MeshAxiom-Base
+
+**MeshAxiom: An Agent-Native 3D Foundation Stack for Programmable Worlds**
+
+[Project](https://alaya-lab.github.io/MeshAxiom/) · [Code](https://github.com/AlayaLab/MeshAxiom)
+
+Public Base weights for **Exact** image-to-native-mesh inference. This repository
+contains vertex EMA 100K, topology EMA 10K and topology VAE EMA 30K. Install the
+MeshAxiom source package for the Python API, CLI and agent tools. No optimizer,
+training data or alternative inference preset is included.
+
+```python
+from huggingface_hub import HfApi
+from meshaxiom import MeshAxiomPipeline
+
+# Resolve once, then reuse this immutable revision for reproducible runs.
+revision = HfApi().model_info("AlayaLab/MeshAxiom-Base").sha
+
+with MeshAxiomPipeline.from_pretrained(
+    "AlayaLab/MeshAxiom-Base",
+    revision=revision,
+) as model:
+    result = model.generate("image.png", "outputs/object", face_budget=4000)
+```
+
+Install MeshAxiom code version `0.2.0.dev2` for this 100K checkpoint; older
+versions expect different weights. No Hugging Face login is required to download these weights. A local model directory
+can also be used. This repository supplies model data; the source package supplies
+the inference implementation. `model_index.json` is descriptive metadata, not a
+Diffusers pipeline.
+
+Input is a 512×512 RGB image on a black background. The recipe uses MBR8, 20
+DPM++2M steps and topology CFG 3. The requested face budget guides generation;
+actual face counts can differ. OBJ/JSON coordinates are Z-up, and exported GLB
+files include a transform into glTF's Y-up frame.
+
+## Verification and limitations
+
+All 829 vertex EMA tensors exactly match the evaluated 100K checkpoint. The
+unchanged topology and VAE files contain another 1,173 tensors. The vertex model
+loaded strictly on CPU and both geometry stages passed the GPU check below.
+Version `0.2.0.dev2` passes 95 CPU tests. See `WEIGHTS.json`, `VALIDATION.json`
+and `SHA256SUMS` for model identities and the scope of verification.
+
+Sho Kuno contributed the Exact acceleration. His tested platform is H200 with
+NVIDIA PyTorch 25.03. This integrated release also passed a bounded H200 check: one image at 4K and 8K requests, with all four geometry files identical to the reference and resident workers reused between requests. This covers only the recorded inputs, seeds, weights and stack. A cold Docker build remains unverified; run `meshaxiom verify-exact` on your target setup.
+Geometry can have open boundaries or degenerate faces. Earlier paper results and
+speed measurements used the previous 180K vertex checkpoint; they are not
+measurements of this 100K release.
+
+## License
+
+The authors' standalone release license is pending; see `LICENSE.md`. These
+checkpoints contain frozen DINOv3 tensors subject to `external/dinov3/LICENSE.md`.
+The included architecture and processor configurations avoid a separate backbone
+weight download. Third-party terms and notices are retained.
+
+## Paper
+
+The arXiv link and citation will be added when the preprint is available.

SHA256SUMS

@@ -0,0 +1,13 @@
+c493dcf7f25a5d96db1ef997d5a55ed50d93fb0df5e631f6c4c62272faa428ef  .gitattributes
+00ef8af2639ac418a37ff877d07484d74f1cc2fd585d9887c3840768e5b2513c  LICENSE.md
+14bce35fb0ec1d5d29870737122d04ee5e890dc716d79babf8a24f52409be3aa  README.md
+060195caa1df6cac2edfa76a78a6d83b3d1d2461913b80bcf205b088c7bbb7e9  VALIDATION.json
+7a4ba85066fe37827100bcf88c732dad408cbdbbcf384d6e023eea4f81980ee1  WEIGHTS.json
+25d122eb8f5b880fd23c736fb6ea8018ee45c12237e00b8a86d14c653904999e  external/dinov3/LICENSE.md
+135ecd23e34a70b6fbed8b083fdecb319b7e3a54e3d849258bbe4ddcf1783bb5  external/dinov3/config.json
+960c41d1f3a7778b936365769a2d90550b318a6c0a53a0296957adacfe5e0dd7  external/dinov3/preprocessor_config.json
+d4515c61e0b67bc79ac22e28e4378dc50fa1746ab883be09cea69a61d9b930ef  meshaxiom.json
+8506af71108ff3c3c72e4d1419fd5f789b932aed98852c0eb891c99ddecefa48  model_index.json
+7fb0c7cc2035616f811326d3e64c1393de39d8f652ab7aa122ab84a987d2a19f  weights/topology-ema.pt
+7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd  weights/vae-ema.pt
+7743cc03ad05abcadab204b8a00f7d312ad2b26d7de1eb64ba009a75c12054f2  weights/vertex-ema.pt

VALIDATION.json

@@ -0,0 +1,246 @@
+{
+  "status": "PASS",
+  "model": "MeshAxiom-Base",
+  "implementation": "exact",
+  "code_version": "0.2.0.dev2",
+  "recipe": {
+    "vertex_step": 100000,
+    "topology_step": 10000,
+    "vae_step": 30000,
+    "weights": "ema",
+    "vertex_conditioning": "image",
+    "vertex_condition_fusion": null
+  },
+  "tensor_counts": {
+    "vertex": 829,
+    "topology": 830,
+    "vae": 343,
+    "total": 2002
+  },
+  "weights": {
+    "status": "PASS",
+    "model": "MeshAxiom-Base",
+    "implementation": "exact",
+    "files": [
+      {
+        "kind": "vertex",
+        "file": "weights/vertex-ema.pt",
+        "step": 100000,
+        "weights": "ema",
+        "tensor_count": 829,
+        "all_tensor_values_and_dtypes_equal": true,
+        "only_metadata_changes": [
+          "inference-only configuration; training paths omitted"
+        ],
+        "source_sha256": "d660061c7ea74a2e3f7fa5f60ef5671fcde13d0f5efd00c7e2fecb68db7b9544",
+        "sha256": "7743cc03ad05abcadab204b8a00f7d312ad2b26d7de1eb64ba009a75c12054f2",
+        "size": 9421814821
+      },
+      {
+        "kind": "topology",
+        "file": "weights/topology-ema.pt",
+        "step": 10000,
+        "weights": "ema",
+        "tensor_count": 830,
+        "all_tensor_values_and_dtypes_equal": true,
+        "only_metadata_changes": [
+          "v113.version"
+        ],
+        "source_sha256": "3818846563217c1d6819c8c9dbf7d8b8cc43b7a6cd2e96d6b37b576fe147ef45",
+        "sha256": "7fb0c7cc2035616f811326d3e64c1393de39d8f652ab7aa122ab84a987d2a19f",
+        "size": 9423507316
+      },
+      {
+        "kind": "vae",
+        "file": "weights/vae-ema.pt",
+        "step": 30000,
+        "weights": "ema",
+        "tensor_count": 343,
+        "all_tensor_values_and_dtypes_equal": true,
+        "only_metadata_changes": [],
+        "source_sha256": "7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd",
+        "sha256": "7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd",
+        "size": 446157647
+      }
+    ],
+    "scope": "All 829 vertex EMA tensors exactly match the evaluated 100K checkpoint; topology and VAE files retain their previous identities."
+  },
+  "cpu_tests": {
+    "status": "PASS",
+    "passed": 95
+  },
+  "cpu_vertex_load": {
+    "status": "PASS",
+    "model": "MeshAxiom-Base",
+    "step": 100000,
+    "strict_state_keys": 829,
+    "all_loaded_tensors_exact": true,
+    "device": "cpu",
+    "forward_calls": 0,
+    "optimizer_included": false,
+    "conditioning": "image",
+    "condition_fusion": null
+  },
+  "environment": {
+    "python": "3.12.3",
+    "torch": "2.7.0a0+7c8ec84dab.nv25.03",
+    "cuda_runtime": "12.8",
+    "packages": {
+      "transformers": "4.57.3",
+      "numpy": "1.26.4",
+      "trimesh": "4.10.1",
+      "huggingface-hub": "0.36.2"
+    }
+  },
+  "gpu_identity": {
+    "status": "PASS",
+    "device": "NVIDIA H200",
+    "input_sha256": "c1289d1201f221148aa7f36f902b5ab66dba4f5ff6ed4efa060f1a363f63d86d",
+    "face_requests": [
+      4000,
+      8000
+    ],
+    "vertex_seed": 2026082301,
+    "topology_seed": 1574365204,
+    "mbr": 8,
+    "steps": 20,
+    "comparisons": {
+      "4000": {
+        "status": "PASS",
+        "files": {
+          "vertices.npz": {
+            "equal": true,
+            "size": 7792,
+            "sha256": "53bc0533866a05d3ee771f519a7de292d4b3da0a1f09111bbb07b9432c8409fe"
+          },
+          "topology.npz": {
+            "equal": true,
+            "size": 79907,
+            "sha256": "17270c21e13a1b4e9bb5b712efe4bc2eaafcc3ed579968725e9e80edc99dec97"
+          },
+          "mesh.obj": {
+            "equal": true,
+            "size": 165722,
+            "sha256": "e94cfa2e57360ea3be7e081d5ad2dfc667aa4c5b82e67457f1472b11f49b1cc7"
+          },
+          "mesh.json": {
+            "equal": true,
+            "size": 382849,
+            "sha256": "07037c8a9998fedb30db46e827a040adf11894c9a75d3a4d5f1eddf23814b616"
+          }
+        }
+      },
+      "8000": {
+        "status": "PASS",
+        "files": {
+          "vertices.npz": {
+            "equal": true,
+            "size": 16437,
+            "sha256": "2efd98849db5c0a7cd5b1992921edf522ac54b3d77e66e8aac9cb9257e1acc51"
+          },
+          "topology.npz": {
+            "equal": true,
+            "size": 178198,
+            "sha256": "6e9388958fa22dcaec299a1b2a224d0f34fd7de06ac3e2581fe1537f603600bf"
+          },
+          "mesh.obj": {
+            "equal": true,
+            "size": 364255,
+            "sha256": "956535fde710ae588bbfb744aa622c170665c4a697d5cca0811c31c91c69d5fc"
+          },
+          "mesh.json": {
+            "equal": true,
+            "size": 830713,
+            "sha256": "7182cdb4e044d21546223b01627f26e2f6ccf608318281b4441ddd747ecc236e"
+          }
+        }
+      }
+    },
+    "resident_workers_reused": true,
+    "all_own_workers_exited": true,
+    "scope": "One prepared image, two face requests, fixed seeds and this software stack; no current-checkpoint speed benchmark."
+  },
+  "source_sha256": {
+    "src/meshaxiom/__init__.py": "1e2ea7cc8d2dd06013f1e1c1fa676694c8ec0aafa41e21a84fae7ab7ba0b3e54",
+    "src/meshaxiom/cli.py": "6645a900758c56f4ba4a6275ecc2fea882082f1f92764482b5acaaa10278abbd",
+    "src/meshaxiom/config.py": "775beca8b412e0c4e261c660e567951a4537eae9e9b980cac4882ac43afaf74d",
+    "src/meshaxiom/contracts.py": "4f4a2fa13f806d5f8c8e8cd948c31022485d11a78816715ba55a80f3e6fc3c66",
+    "src/meshaxiom/hub.py": "79204b01b6dd12699d751079ae26493c714b486942ffd3e077d35b7c2f032652",
+    "src/meshaxiom/material_backend/__init__.py": "3e7b48f127d6b6197bb29f92f6021b39b9ffa618e80ddfb19b8f34245471b870",
+    "src/meshaxiom/material_backend/alignment.py": "ddb7a238aa36fa02c008adc212155e04cd222be0ebeefc2a660c415d52678d57",
+    "src/meshaxiom/material_backend/geometry.py": "db1bf37640d55d1a69db7f69c63cf482bdc3e79938b0120fb9048c43dd2ead35",
+    "src/meshaxiom/material_backend/pixal.py": "2fa924f04360f387d8c35c5217ef356955b9773475e4acf75df5f0713e237f43",
+    "src/meshaxiom/material_backend/superresolution.py": "b5dca36ac4a0457eb67677ebebe8b5765a6fe44368b3452c4ae2effd1a848982",
+    "src/meshaxiom/pipeline.py": "e8a1f08d0b0d60bf70457f561bdbe3915af399e63f09ab5eac8e67d2d370b242",
+    "src/meshaxiom/preprocess.py": "d06b4c13680d7ae69d8da11429334cb2f612cc6a947af4bfa9b768f9c53c5475",
+    "src/meshaxiom/server.py": "4436c6c73911c0fe3f490c5d8562fcb452522ac20d4f41627db0bd824d1a47ce",
+    "src/meshaxiom/session.py": "9fa6f1ddf43f7d72088795f6f2a6d987f534c41734890a7342ef00b41c52a4fe",
+    "src/meshaxiom/tools.py": "3d6ca73c56a0fb63cdad3c837baa20445271f357d919abe4e2c1a3eb2f7b2167",
+    "src/meshaxiom/vendor/__init__.py": "3af1ed55a3edc1d6a309aee82426d0c31247a7395d441c35218115075c1101b4",
+    "src/meshaxiom/vendor/kuno/__init__.py": "bd4170391fcc9cece92cfe90c83d26badc816ec4e60266e47e79fff8e82c47b9",
+    "src/meshaxiom/vendor/kuno/p4_vertex_engine.py": "1016132a661f931bf5416a4017f7cc696a7cdb86cbd27324f7b4c8588c913f27",
+    "src/meshaxiom/vendor/kuno/p5_post_fast.py": "92890fabd628ea05e9a4028bcbb5ab53c64b143dc93eacaec3002f789d7f9aca",
+    "src/meshaxiom/vendor/kuno/p7_exact_backend.py": "1c61ddf6f651f2b27feebcbb7a4f27ee899993ace262d0cf6ab88da5e5ab2aa5",
+    "src/meshaxiom/vendor/kuno/topo_fast.py": "b5d610b636ac1e0b96f6d9d3384ffaec97d59c8f675da1ab2760e36f918febec",
+    "src/meshaxiom/vendor/kuno/vertex_engine.py": "a0c2fb8adf320248d9d97ab1e09a06764ee8bfab9fed5f6eabbeb806384467de",
+    "src/meshaxiom/vendor/recipe/__init__.py": "3af1ed55a3edc1d6a309aee82426d0c31247a7395d441c35218115075c1101b4",
+    "src/meshaxiom/vendor/recipe/decode_policy90.py": "3a0da4b826f3a359187d91af6d69e0cf202b5efbeb233ecbc855201d2b3c1937",
+    "src/meshaxiom/vendor/recipe/gen_mbr128.py": "188ad31d9462345e71f3ff39229963fba8eddcc6ed485d7b135dae7a223a92e0",
+    "src/meshaxiom/vendor/recipe/mesh_utils90.py": "e55d1fb6f9e8421d538e578348c6b8a13823b5993956dc73cdc104715d278aee",
+    "src/meshaxiom/vendor/recipe/model118rich.py": "313d35a90f32f669b63f43530df30cf4fdf94a97551b45cb6b4875d0f1e87cc1",
+    "src/meshaxiom/vendor/recipe/orient_faces.py": "90458f3c690429d378a7a35e9e6d87944f1be2011f14ef2c85462bfcdfc325de",
+    "src/meshaxiom/vendor/recipe/runtime.py": "bd9883857a020df31973b2df0d524d425d209527b5a5a2a881a2e34d8854a7ca",
+    "src/meshaxiom/vendor/superresolution/__init__.py": "e2dc2b39ea89853eda532f298472d68e2a549e2d9d84b01995fda4f31d0ea650",
+    "src/meshaxiom/vendor/superresolution/arch_util.py": "7032b86c5ef8f8d5c9b22af89ae8dbe32033500f4d8cc9c2e14e57a46d69fb86",
+    "src/meshaxiom/vendor/superresolution/rrdbnet_arch.py": "928d0a8367794ef1ebe32517b511b1ffdc3be366bfcbdece44136b6a3f9a5193",
+    "src/meshaxiom/vendor/topology/__init__.py": "3af1ed55a3edc1d6a309aee82426d0c31247a7395d441c35218115075c1101b4",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/config.py": "1381bda8e7594e0b54840be343864beea0077f166edaee2e190757a03ccc52e4",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/data/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/data/mesh_ops.py": "c5c66bed4d539b81464ec46f6bc4779a9d4b2ee56665389728c9aeb598b9d545",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/diffusion/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/diffusion/flow.py": "d833a8b34cf721a84fc4e7808d706699175c4a6cb02164a3d9921aa2406a22b6",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/condition.py": "d54f95d2ddc559e24820391514c3496c5dd848d2cd9d2c2f49798acec35b05b3",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/fps.py": "8eefa997216503b9f0c86a596eb5f21d1d0298cbc13c132f46ecc0b13f7ee707",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/layers.py": "be13fba0516a58858b82b233a9f28b0aae4f05b1caceef403d4ad57e20dabb41",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/octree_dit.py": "62ed57f74ccf504a5878deb60d26fd56740970230d817dd0f4bd52d8d2084941",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/topo_dit.py": "de5dba6ac0ec8e799c32053364ff31049d809e0fee844b2db58e647680edf2f3",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/models/topo_vae.py": "3dfc795bc1cf9c1387291fdb4035cf8c319f85ae1b64279bafa91855bfb53633",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/sample.py": "c50833dafbc93e37931427eae376b25da65b1b59093f8a46c38cafdd873909ca",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/topology/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/topology/spacetime.py": "441ec2b60f190057c4c0c1ad517a049eff1a1b91680b04b7f552146b5481df26",
+    "src/meshaxiom/vendor/topology/meshaxiom_core/topology/streamed_edges.py": "313eed389c488bf4845f06c07ad0f04e21183c47547f0becf57ab8d4f106fe68",
+    "src/meshaxiom/vendor/vertex/__init__.py": "3af1ed55a3edc1d6a309aee82426d0c31247a7395d441c35218115075c1101b4",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/config.py": "1381bda8e7594e0b54840be343864beea0077f166edaee2e190757a03ccc52e4",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/data/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/data/mesh_ops.py": "c5c66bed4d539b81464ec46f6bc4779a9d4b2ee56665389728c9aeb598b9d545",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/diffusion/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/diffusion/flow.py": "d833a8b34cf721a84fc4e7808d706699175c4a6cb02164a3d9921aa2406a22b6",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/__init__.py": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/condition.py": "ab95c04d13a82de6a7334ae8bf5e42bbc58d1afb1c8c94f94e8864b2e40574be",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/fps.py": "70a0005a1b3ca0f8e1f768eeffabcefcab1f8cd05bf888c515709167b3485b21",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/fps_fast153.py": "002154485f0442311db110a2eadf6a2328d6bdf11be7e2e68f78b95084f309b6",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/layers.py": "cedc1ea3e048394c5568ee06d991ea8066f4687790f18aff6da3681771123680",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/octree_dit.py": "b8397cd291a9f323542b5fa52646365a447139491b98391879c8da8811bd27d9",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/models/posenc182.py": "28f7b375bcc2c9f855476f6d44298504d5fc5e8b3b28239dc2a739f9a7be9638",
+    "src/meshaxiom/vendor/vertex/meshaxiom_core/sample.py": "315f6881b51cafb32e0c851fb837627ce7661692169722d3426c2cefc68c4e84",
+    "src/meshaxiom/verification.py": "260f0baaa7eed584fb1e68e7f5857738f01688e2f0927c98158b589391bc6856",
+    "src/meshaxiom/workers/__init__.py": "d2e0cdc3af9ad149d72b6e554af62434d04fc3dbfc6531572e3c3b2b6a6fcf6a",
+    "src/meshaxiom/workers/_geometry_state.py": "a3c700ebc1ffd282d109fa454cee075304649cd23877b1388a5ae7fc778c2978",
+    "src/meshaxiom/workers/material.py": "65e11f200d48fc09d8f34039d784efebb0b94421fb15abf8e6ba55e540a1fa10",
+    "src/meshaxiom/workers/offline_encoder.py": "07f872b5f818955ac413ec7e0789f043f9369fbcd658f7faa69e2c6bde47219a",
+    "src/meshaxiom/workers/reference_topology.py": "0001650277dce2c860d4b5e28dd10a2f777e0efa75ff7ed3b1ba4c69258d5749",
+    "src/meshaxiom/workers/reference_vertex.py": "6db1f34d7838a2be2ceef977b944c72289e51304fd1c4f979c1d532df10b5e9a",
+    "src/meshaxiom/workers/resident.py": "68d2f995bab4cf464e728a1cc2cae9227d5b6a707d2d752f0b719ac611135cf2",
+    "src/meshaxiom/workers/superresolution.py": "503b147de49283df210cadacf4b905226ef9af57e3e5ba85803e14644faf4c34",
+    "src/meshaxiom/workers/topology.py": "607f8e64c9b2b3c6898465dc722b4c60ca51e8795930ea97a53bce2584e5c837",
+    "src/meshaxiom/workers/vertex.py": "3015624fb50b5b14f6cb9263df4ed5fde821356d748dba607a482e0655210028"
+  },
+  "notes": [
+    "Topology and VAE files are unchanged from the prior release.",
+    "Earlier 180K paper and speed measurements do not measure the current 100K vertex weights.",
+    "Cold Docker build and other GPU stacks require target verification."
+  ]
+}

WEIGHTS.json

@@ -0,0 +1,48 @@
+{
+  "status": "PASS",
+  "model": "MeshAxiom-Base",
+  "implementation": "exact",
+  "files": [
+    {
+      "kind": "vertex",
+      "file": "weights/vertex-ema.pt",
+      "step": 100000,
+      "weights": "ema",
+      "tensor_count": 829,
+      "all_tensor_values_and_dtypes_equal": true,
+      "only_metadata_changes": [
+        "inference-only configuration; training paths omitted"
+      ],
+      "source_sha256": "d660061c7ea74a2e3f7fa5f60ef5671fcde13d0f5efd00c7e2fecb68db7b9544",
+      "sha256": "7743cc03ad05abcadab204b8a00f7d312ad2b26d7de1eb64ba009a75c12054f2",
+      "size": 9421814821
+    },
+    {
+      "kind": "topology",
+      "file": "weights/topology-ema.pt",
+      "step": 10000,
+      "weights": "ema",
+      "tensor_count": 830,
+      "all_tensor_values_and_dtypes_equal": true,
+      "only_metadata_changes": [
+        "v113.version"
+      ],
+      "source_sha256": "3818846563217c1d6819c8c9dbf7d8b8cc43b7a6cd2e96d6b37b576fe147ef45",
+      "sha256": "7fb0c7cc2035616f811326d3e64c1393de39d8f652ab7aa122ab84a987d2a19f",
+      "size": 9423507316
+    },
+    {
+      "kind": "vae",
+      "file": "weights/vae-ema.pt",
+      "step": 30000,
+      "weights": "ema",
+      "tensor_count": 343,
+      "all_tensor_values_and_dtypes_equal": true,
+      "only_metadata_changes": [],
+      "source_sha256": "7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd",
+      "sha256": "7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd",
+      "size": 446157647
+    }
+  ],
+  "scope": "All 829 vertex EMA tensors exactly match the evaluated 100K checkpoint; topology and VAE files retain their previous identities."
+}

external/dinov3/LICENSE.md

@@ -0,0 +1,66 @@
+# DINOv3 License
+
+*Last Updated: August 19, 2025*
+
+**“Agreement”** means the terms and conditions for use, reproduction, distribution and modification of the DINO Materials set forth herein.
+
+**“DINO Materials”** means, collectively, Documentation and the models, software and algorithms, including machine-learning model code, trained model weights, inference-enabling code, training-enabling code, fine-tuning enabling code, and other elements of the foregoing distributed by Meta and made available under this Agreement.
+
+**“Documentation”** means the specifications, manuals and documentation accompanying
+DINO Materials distributed by Meta.
+
+**“Licensee”** or **“you”** means you, or your employer or any other person or entity (if you are entering into this Agreement on such person or entity’s behalf), of the age required under applicable laws, rules or regulations to provide legal consent and that has legal authority to bind your employer or such other person or entity if you are entering in this Agreement on their behalf.
+
+**“Meta”** or **“we”** means Meta Platforms Ireland Limited (if you are located in or, if you are an entity, your principal place of business is in the EEA or Switzerland) or Meta Platforms, Inc. (if you are located outside of the EEA or Switzerland).
+
+**“Sanctions”** means any economic or trade sanctions or restrictions administered or enforced by the United States (including the Office of Foreign Assets Control of the U.S. Department of the Treasury (“OFAC”), the U.S. Department of State and the U.S. Department of Commerce), the United Nations, the European Union, or the United Kingdom.
+
+**“Trade Controls”** means any of the following: Sanctions and applicable export and import controls.
+
+By clicking “I Accept” below or by using or distributing any portion or element of the DINO Materials, you agree to be bound by this Agreement.
+
+## 1. License Rights and Redistribution.
+
+a. <ins>Grant of Rights</ins>. You are granted a non-exclusive, worldwide, non-transferable and royalty-free limited license under Meta’s intellectual property or other rights owned by Meta embodied in the DINO Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the DINO Materials.
+
+b. <ins>Redistribution and Use</ins>.
+
+i. Distribution of DINO Materials, and any derivative works thereof, are subject to the terms of this Agreement. If you distribute or make the DINO Materials, or any derivative works thereof, available to a third party, you may only do so under the terms of this Agreement and you shall provide a copy of this Agreement with any such DINO Materials.
+
+ii.  If you submit for publication the results of research you perform on, using, or otherwise in connection with DINO Materials, you must acknowledge the use of DINO Materials in your publication.
+
+iii. Your use of the DINO Materials must comply with applicable laws and regulations, including Trade Control Laws and applicable privacy and data protection laws.
+
+iv. Your use of the DINO Materials will not involve or encourage others to reverse engineer, decompile or discover the underlying components of the DINO Materials.
+
+v. You are not the target of Trade Controls and your use of DINO Materials must comply with Trade Controls. You agree not to use, or permit others to use, DINO Materials for any activities subject to the International Traffic in Arms Regulations (ITAR) or end uses prohibited by Trade Controls, including those related to military or warfare purposes, nuclear industries or applications, espionage, or the development or use of guns or illegal weapons.
+
+## 2. User Support.
+
+Your use of the DINO Materials is done at your own discretion; Meta does not process any information nor provide any service in relation to such use.  Meta is under no obligation to provide any support services for the DINO Materials. Any support provided is “as is”, “with all faults”, and without warranty of any kind.
+
+## 3. Disclaimer of Warranty.
+
+UNLESS REQUIRED BY APPLICABLE LAW, THE DINO MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, AND META DISCLAIMS ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE DINO MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE DINO MATERIALS AND ANY OUTPUT AND RESULTS.
+
+## 4. Limitation of Liability.
+
+IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY DIRECT OR INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.
+
+## 5. Intellectual Property.
+
+a. Subject to Meta’s ownership of DINO Materials and derivatives made by or for Meta, with respect to any derivative works and modifications of the DINO Materials that are made by you, as between you and Meta, you are and will be the owner of such derivative works and modifications.
+
+b. If you institute litigation or other proceedings against Meta or any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the DINO Materials, outputs or results, or any portion of any of the foregoing, constitutes infringement of intellectual property or other rights owned or licensable by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold harmless Meta from and against any claim by any third party arising out of or related to your use or distribution of the DINO Materials.
+
+## 6. Term and Termination.
+
+The term of this Agreement will commence upon your acceptance of this Agreement or access to the DINO Materials and will continue in full force and effect until terminated in accordance with the terms and conditions herein. Meta may terminate this Agreement if you are in breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete and cease use of the DINO Materials. Sections 3, 4 and 7 shall survive the termination of this Agreement.
+
+## 7. Governing Law and Jurisdiction.
+
+This Agreement will be governed and construed under the laws of the State of California without regard to choice of law principles, and the UN Convention on Contracts for the International Sale of Goods does not apply to this Agreement. The courts of California shall have exclusive jurisdiction of any dispute arising out of this Agreement.
+
+## 8. Modifications and Amendments.
+
+Meta may modify this Agreement from time to time; provided that they are similar in spirit to the current version of the Agreement, but may differ in detail to address new problems or concerns. All such changes will be effective immediately. Your continued use of the DINO Materials after any modification to this Agreement constitutes your agreement to such modification. Except as provided in this Agreement, no modification or addition to any provision of this Agreement will be binding unless it is in writing and signed by an authorized representative of both you and Meta.

external/dinov3/config.json

@@ -0,0 +1,32 @@
+{
+  "architectures": [
+    "DINOv3ViTModel"
+  ],
+  "attention_dropout": 0.0,
+  "drop_path_rate": 0.0,
+  "hidden_act": "gelu",
+  "hidden_size": 1024,
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "key_bias": false,
+  "layer_norm_eps": 1e-05,
+  "layerscale_value": 1.0,
+  "mlp_bias": true,
+  "model_type": "dinov3_vit",
+  "num_attention_heads": 16,
+  "num_channels": 3,
+  "num_hidden_layers": 24,
+  "num_register_tokens": 4,
+  "patch_size": 16,
+  "pos_embed_jitter": null,
+  "pos_embed_rescale": 2.0,
+  "pos_embed_shift": null,
+  "proj_bias": true,
+  "query_bias": true,
+  "rope_theta": 100.0,
+  "torch_dtype": "float32",
+  "transformers_version": "4.56.0.dev0",
+  "use_gated_mlp": false,
+  "value_bias": true
+}

external/dinov3/preprocessor_config.json

@@ -0,0 +1,31 @@
+{
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "disable_grouping": null,
+  "do_center_crop": null,
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "DINOv3ViTImageProcessorFast",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "input_data_format": null,
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "return_tensors": null,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}

meshaxiom.json

@@ -0,0 +1,35 @@
+{
+  "schema_version": 1,
+  "geometry": {
+    "vertex_checkpoint": {
+      "path": "weights/vertex-ema.pt",
+      "sha256": "7743cc03ad05abcadab204b8a00f7d312ad2b26d7de1eb64ba009a75c12054f2",
+      "step": 100000,
+      "weights": "ema"
+    },
+    "topology_checkpoint": {
+      "path": "weights/topology-ema.pt",
+      "sha256": "7fb0c7cc2035616f811326d3e64c1393de39d8f652ab7aa122ab84a987d2a19f",
+      "step": 10000,
+      "weights": "ema"
+    },
+    "vae_checkpoint": {
+      "path": "weights/vae-ema.pt",
+      "sha256": "7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd",
+      "step": 30000,
+      "weights": "ema"
+    },
+    "image_model": "external/dinov3",
+    "topology_seed": 1574365204,
+    "implementation": "exact"
+  },
+  "runtime": {
+    "vertex_python": "python",
+    "topology_python": "python",
+    "device": "cuda:0",
+    "gpu_lock_pattern": "cuda-{index}.lock",
+    "minimum_free_disk_gib": 0,
+    "minimum_free_workspace_gib": 0,
+    "timeout_seconds": 3600
+  }
+}

model_index.json

@@ -0,0 +1,13 @@
+{
+  "_class_name": "MeshAxiomPipeline",
+  "_library_name": "meshaxiom",
+  "schema_version": 1,
+  "model": "MeshAxiom-Base",
+  "implementation": "exact",
+  "configuration": "meshaxiom.json",
+  "vertex_step": 100000,
+  "topology_step": 10000,
+  "vae_step": 30000,
+  "ema_only": true,
+  "optimizer_included": false
+}

weights/topology-ema.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fb0c7cc2035616f811326d3e64c1393de39d8f652ab7aa122ab84a987d2a19f
+size 9423507316

weights/vae-ema.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d6833425af71e923756d2ebdb2863e704857bffb8b365ab0c2f5ec3bfb6f4bd
+size 446157647

weights/vertex-ema.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7743cc03ad05abcadab204b8a00f7d312ad2b26d7de1eb64ba009a75c12054f2
+size 9421814821