Release depth-v1 controlled realization curriculum

CITATION.cff

@@ -4,7 +4,7 @@ message: >-
   please cite this dataset and the accompanying PixCell project.
 title: "PixCell Dataset: Representation-First Image-to-Code Curriculum"
 type: dataset
-version: core-v1
+version: depth-v1
 date-released: 2026-07-26
 authors:
   - family-names: Agarwal

README.md

@@ -17,7 +17,7 @@ tags:
   - synthetic
   - curriculum-learning
 size_categories:
-  - n<1K
+  - 1K<n<10K
 configs:
   - config_name: all
     default: true
@@ -58,41 +58,53 @@ configs:
     data_files:
       - split: train
         path: data/l4/train-*.parquet
+  - config_name: depth-v1
+    data_files:
+      - split: train
+        path: depth-v1/data/train-*.parquet
+      - split: validation
+        path: depth-v1/data/validation-*.parquet
 ---
 
 # PixCell Dataset
 
-`core-v1` is a 738-row image-to-code training dataset for photonic component
-reconstruction. It pairs exact high-visibility model inputs with physical
-footprint and display-scale context, plus verified primitive-only GDSFactory
-programs. The examples cover 547 representation grammars across five
-curriculum levels, from primitives (L0) to complete components (L4).
+PixCell pairs high-visibility photonic component images and physical footprint
+context with verified primitive-only GDSFactory programs. It is published in
+two complementary configurations:
 
-| Release | Rows | Representations | Split | License |
-|---|---:|---:|---|---|
-| `core-v1` | 738 | 547 | `train` | MIT |
+| Configuration | Rows | Representations | Splits |
+|---|---:|---:|---|
+| `core-v1` | 738 | 547 | `train`: 738 |
+| `depth-v1` | 4,560 | 547 | `train`: 3,468 · `validation`: 1,092 |
 
-[Complete gallery](galleries/all.png) · [L0](galleries/l0.png) ·
+[Core gallery](galleries/all.png) · [L0](galleries/l0.png) ·
 [L1](galleries/l1.png) · [L2](galleries/l2.png) ·
-[L3](galleries/l3.png) · [L4](galleries/l4.png)
+[L3](galleries/l3.png) · [L4](galleries/l4.png) ·
+[Depth overview](depth-v1/galleries/overview.png)
 
 ## Load from Hugging Face
 
 ```bash
-pip install datasets
+pip install "datasets[vision]"
 ```
 
 ```python
 from datasets import load_dataset
 
-dataset = load_dataset(
+core = load_dataset(
     "qpaig-mit/pixcell",
     "core-v1",
     revision="core-v1",
     split="train",
 )
 
-row = dataset[0]
+depth = load_dataset(
+    "qpaig-mit/pixcell",
+    "depth-v1",
+    revision="depth-v1",
+)
+
+row = core[0]
 row["image"]          # exact model input
 row["instruction"]    # footprint and display-scale context
 row["code"]           # supervised program
@@ -120,10 +132,15 @@ components.
 Use both the `core-v1` configuration and `revision="core-v1"` to pin this
 release.
 
-All rows are training data. The paper's held-out evaluation figures are not
-included. If creating a downstream validation split, keep every
-`leakage_group_id` together rather than randomly separating related
-realizations or poses.
+`depth-v1` retains all 738 core rows. It adds seven controlled, model-visible
+geometry realizations to 546 of the 547 approved representation anchors; one
+anchor remains anchor-only because its sole exposed literal changes global
+placement rather than visible shape. Its validation split holds out parameter
+settings, not representation classes: every representation remains present in
+training. Use the `depth-v1`
+configuration and `revision="depth-v1"` together to pin it. The repository
+revision selects the published release, while the configuration selects its
+Parquet files.
 
 ## Row contract
 
@@ -145,6 +162,8 @@ physical aspect.
 
 The full schema and release inventory are in
 [`metadata/catalog.json`](metadata/catalog.json).
+The additional `depth-v1` lineage and controlled-variation fields are
+documented in [`depth-v1/README.md`](depth-v1/README.md).
 
 ## Admission
 
@@ -153,10 +172,12 @@ failed builds, private benchmark figures, and RL rollouts are excluded.
 
 Admission checks compilation and execution, primitive-only source purity,
 declared ports and connectivity, structural witnesses, calibrated geometry,
-footprint agreement, and artifact hashes. The exact policy is in
-[`factory/admission_policy.json`](factory/admission_policy.json).
+footprint agreement, and artifact hashes. The core policy is in
+[`factory/admission_policy.json`](factory/admission_policy.json); the depth
+release binds its accepted rows and evidence in
+[`depth-v1/factory/accepted_manifest.json`](depth-v1/factory/accepted_manifest.json).
 
-The release reports five exact duplicate groups and 41
+The frozen `core-v1` release reports five exact duplicate groups and 41
 rotation/reflection-equivalent silhouette groups. They are retained as
 declared API aliases, pose examples, or calibration cases and are bound by
 `leakage_group_id`.
@@ -164,22 +185,30 @@ declared API aliases, pose examples, or calibration cases and are bound by
 ## Reproduce
 
 The accepted-source capsule under `factory/core-v1/` makes the dataset
-self-contained:
+self-contained. From a clean PixCell checkout:
 
 ```bash
+cd dataset
 python3.13 -m venv .venv
 source .venv/bin/activate
 python -m pip install -r requirements-lock.txt
 
 python scripts/build_all.py
 python scripts/validate_release.py --root . --execute-programs
+python depth-v1/scripts/build.py --dataset-root . --output depth-v1
+python depth-v1/scripts/validate.py \
+  --release-root depth-v1 \
+  --dataset-root . \
+  --execute-programs
 python -m unittest discover -s tests -v
+python -m pytest depth-v1/tests -q
 ```
 
-The rebuild must reproduce this logical release digest:
+The rebuilds must reproduce these frozen logical release digests:
 
 ```text
-e1c509dca337ce485efcb2b35101a052e06b12331b63a2f21a36c93b3682ea89
+core-v1   e1c509dca337ce485efcb2b35101a052e06b12331b63a2f21a36c93b3682ea89
+depth-v1  676c49134d4d044c7d84426cf8eeecf09302e74ca4aa548956f14b7631a4d80b
 ```
 
 The capsule contains accepted programs, explicit grammar assignments, image
@@ -187,9 +216,10 @@ evidence, and admission sidecars. It excludes historical proposal generators,
 rejected candidates, private benchmarks, and RL runs. See
 [`factory/README.md`](factory/README.md) for the replay contract.
 
-The rows contain executable Python. The validator runs programs in a temporary
-working directory, not a security sandbox; inspect code from any untrusted
-revision before executing it.
+The rows contain executable Python. Core replay uses isolated temporary
+working directories; depth replay executes verified rows in-process for
+speed. Neither is a security sandbox. Inspect code from any untrusted revision
+before executing it.
 
 ## Export
 

THIRD_PARTY_NOTICES.md

@@ -20,6 +20,7 @@ package remains governed by its own license.
 | Hugging Face Datasets | 5.0.0 | Apache-2.0 |
 | Apache Arrow / PyArrow | 24.0.0 | Apache-2.0 |
 | huggingface_hub | 1.22.0 | Apache-2.0 |
+| pytest | 9.1.1 | MIT |
 
 This summary is provided for convenience and is not a replacement for the
 license text shipped by each dependency. A local installation made from

depth-v1/README.md

@@ -0,0 +1,47 @@
+# PixCell depth-v1
+
+`depth-v1` is a separate, core-compatible release of controlled geometry
+realizations around the 547 approved PixCell representations.
+
+The immutable 738-row PixCell core is retained in full. Its 547 semantic rows
+are the `r00` anchors; the other 191 core rows remain as `core-extra` records.
+Exactly 546 anchors have seven additional, model-visible geometry states. One
+L2 anchor remains anchor-only because its sole mutable literal is a global
+translation erased by maximum-visibility normalization; synthesizing its
+descendants would create repeated inputs with different code labels.
+
+| Release split | Contents | Rows |
+|---|---|---:|
+| `train` | 738 core rows plus `r01`–`r05` descendants | 3,468 |
+| `validation` | `r06`–`r07` descendants | 1,092 |
+| **Total** | 546 complete eight-state groups, one anchor-only representation, and core extras | **4,560** |
+
+This release does not alter `core-v1`. Every row binds its parent core example,
+parent code hash, and frozen core digest.
+
+## Row format
+
+Rows preserve all 30 core columns, including the model image, physical-aspect
+target, parametric Python program, footprint, topology, hashes, and curriculum
+metadata. The appended depth fields record:
+
+- parent and release provenance;
+- geometry and realization identity;
+- controlled transform axes and parameter deltas;
+- dependency, pose, render, and split grouping;
+- verification evidence (hashes and metrics) and generator version; and
+- a deterministic logical-row digest.
+
+Images remain embedded in Parquet, so the release loads without a custom
+dataset script. The Hugging Face feature declaration is provided by
+[`scripts/schema.py`](scripts/schema.py).
+
+## Integrity
+
+`scripts/schema.py` defines the stable column order, Hugging Face features,
+logical row and release digests, deterministic Parquet shard names, the exact
+frozen population contract, and manifest, freeze, and checksum helpers.
+
+The logical digest hashes every non-payload field, including the stored hashes
+of code and images, in ascending row-ID order. Freeze and checksum records bind
+the generated release artifacts without changing any file in the parent core.

depth-v1/data/train-00000-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2d45c383c1ce7323d322f0aa9cbf6611b2e19437bf8e229be163b4918254973
+size 30433999

depth-v1/data/train-00001-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c51c77d5f076bd6960b1452eeccf351fecff58e98189c7ab593ed4d648c617d0
+size 21436090

depth-v1/data/train-00002-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c4ea04eb13ab8ccab55f4301cf8b8292eb2a23d9dbe5bf5a7b1743c7aba1ae
+size 9497154

depth-v1/data/train-00003-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73929f04bdaed1e9ccb72b9c4a53910da71429e32cfb1a6c90c8154c9849ee8b
+size 8309409

depth-v1/data/train-00004-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c59c6af12beebd505c2b4dabdd398498e9a8c218805e89f5af5a9da451327de
+size 9214604

depth-v1/data/train-00005-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0786acb9dd5ea1c05fccab30b7292dd89d59e402bfb52321af8f57f6ba1631f
+size 9470813

depth-v1/data/train-00006-of-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:425467eeeda20953ec62514b447de3034b215c6b7b1d15fa7ff46c425b3b7432
+size 5831429

depth-v1/data/validation-00000-of-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e823f8ce2dbacfd402e1e4db646fef316f165d362b8928ccdb0d1079a792d22
+size 8963776

depth-v1/data/validation-00001-of-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b974b438ecda3d3f53d5dc4b84969d0ef68fa40aecde2195a081726c39a202d
+size 9198283

factory/expansion/preview.png → depth-v1/data/validation-00002-of-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dcb5196ec8e26effb0c2679508fe16c3a550b498c36cd825cbcb6e967edadf4b
-size 137654
+oid sha256:77109cf40718db1e33c25d4998c22aca1d0f13c4ffe01b271e0b36fd22f14a30
+size 994310

depth-v1/factory/core_reference_snapshot.json

@@ -0,0 +1,23 @@
+{
+  "core_digest_sha256": "e1c509dca337ce485efcb2b35101a052e06b12331b63a2f21a36c93b3682ea89",
+  "reference_artifacts": {
+    "data/l0/train-00000-of-00001.parquet": "e0c5bda1b192a901349aeda2cf9287e22d62b5b791c421955e2ffba9ca2785da",
+    "data/l1/train-00000-of-00001.parquet": "fc37aea17af1a127e46f89604c95e490b8c361d5105a4cd77581f1d93dfcfc44",
+    "data/l2/train-00000-of-00001.parquet": "420d2b75f1cd38e2e3629495b9da36a07bc24ec9b81139816528e9332c7d4007",
+    "data/l3/train-00000-of-00001.parquet": "0dea653b7f8a7d5c5e773606ead614f460fae3f40e8588e67d48d314b053c4b2",
+    "data/l4/train-00000-of-00001.parquet": "05c8162b50feb869224e998a7d43fe81eee40d12371019fb27ae6aef3557a16d",
+    "galleries/all.png": "5fe87e14eb72e0a76cb39bcac1d7cc739861948dc299c400f1a746763c2793d9",
+    "galleries/l0.png": "838dc208f2bddb3c9cfa140ef09cb8d350f866c9b380d285b3c4d6afc3bd9c0a",
+    "galleries/l1.png": "e6cb27d4d4230ceb2c0b8eb5cde759af53759b2c3c05abbd9fd72b05302400d8",
+    "galleries/l2.png": "e730702cd6c321f50d48305ff7952434990ed48dc87c1a41699bf893e364489b",
+    "galleries/l3.png": "5fb0387dfc4be86ac455528e75f9dc33f8979ea5a860812159445f851f2eefd9",
+    "galleries/l4.png": "eca24f18ece945ee76f316e9c97fd48d5c9a52a52b8fdc04eba0857b49376463",
+    "metadata/audit_report.json": "2cb6122042f1b231adc4c6de642f64663f42759a2bf3dcee305c9d255064b0ae",
+    "metadata/catalog.json": "c22125e4201e3ccac233b0d6898a86cda232a5371ad8a68e19b44112aa680511",
+    "metadata/source_reports/l0.json": "654d69c5a9954f4a251450f85ce6143635bbfb26e6f1f01f8e22777223c98d23",
+    "metadata/source_reports/l1.json": "6e95ccff6ea3dfeec8df3c31cbbafbb487986d94736d8ec74d17189edc3da290",
+    "metadata/source_reports/l2.json": "95d17fe874a904df8ef2a2aca5a563d3cc5e72237e611278083181bf1ec9465a",
+    "metadata/source_reports/l3.json": "b5dbd40ef517ec23353d98a6326f8ec649b0ec88ec2f0d2b2fdc622b4d8a6f1a",
+    "metadata/source_reports/l4.json": "a7cfc8be204209ef7997447f20db339a52e327ad6ecdabbcb921c54111de9eeb"
+  }
+}

depth-v1/factory/freeze.json

@@ -0,0 +1,88 @@
+{
+  "anchor_only_parent_ids": [
+    "l2_02_rotated_edge_aligned_cluster__s03__proposal_b"
+  ],
+  "anchor_only_representation_count": 1,
+  "approved_representation_count": 547,
+  "complete_depth_representation_count": 546,
+  "depth_schema_version": "pixcell-depth-row-v1",
+  "geometry_state_count": 8,
+  "logical_release_sha256": "676c49134d4d044c7d84426cf8eeecf09302e74ca4aa548956f14b7631a4d80b",
+  "manifest": {
+    "path": "metadata/manifest.json",
+    "sha256": "e40017affd7bc3fc1acf39ff05167681c8436fc38a1cbe8ba953149079373133"
+  },
+  "parent_core_digest_sha256": "e1c509dca337ce485efcb2b35101a052e06b12331b63a2f21a36c93b3682ea89",
+  "reference_artifacts": {
+    "data/train-00000-of-00007.parquet": "c2d45c383c1ce7323d322f0aa9cbf6611b2e19437bf8e229be163b4918254973",
+    "data/train-00001-of-00007.parquet": "c51c77d5f076bd6960b1452eeccf351fecff58e98189c7ab593ed4d648c617d0",
+    "data/train-00002-of-00007.parquet": "24c4ea04eb13ab8ccab55f4301cf8b8292eb2a23d9dbe5bf5a7b1743c7aba1ae",
+    "data/train-00003-of-00007.parquet": "73929f04bdaed1e9ccb72b9c4a53910da71429e32cfb1a6c90c8154c9849ee8b",
+    "data/train-00004-of-00007.parquet": "7c59c6af12beebd505c2b4dabdd398498e9a8c218805e89f5af5a9da451327de",
+    "data/train-00005-of-00007.parquet": "e0786acb9dd5ea1c05fccab30b7292dd89d59e402bfb52321af8f57f6ba1631f",
+    "data/train-00006-of-00007.parquet": "425467eeeda20953ec62514b447de3034b215c6b7b1d15fa7ff46c425b3b7432",
+    "data/validation-00000-of-00003.parquet": "8e823f8ce2dbacfd402e1e4db646fef316f165d362b8928ccdb0d1079a792d22",
+    "data/validation-00001-of-00003.parquet": "8b974b438ecda3d3f53d5dc4b84969d0ef68fa40aecde2195a081726c39a202d",
+    "data/validation-00002-of-00003.parquet": "77109cf40718db1e33c25d4998c22aca1d0f13c4ffe01b271e0b36fd22f14a30",
+    "factory/accepted_manifest.json": "c89f6131d8ee41b26e4cbacef6fe935fe7f98c01726f846689ff67ed131b0a72",
+    "factory/anchor_manifest.json": "89f50e714b0f07bf8c053575f964f3294076298c312e4f31233540dbd6fb5887",
+    "factory/core_reference_snapshot.json": "363818781a7e4708de2b8fc99e0f4119dd5c91fc6b57cb2752542d503b5df088",
+    "factory/generation_plan.json": "f81cab2e287e2d7cb7c3ff809c59a4f57719fa24474aefde9202ee7b8c11eec5",
+    "factory/variation_contracts.json": "02b5ae3838fcd4f66d0e7bec00ba7733145868287256d179a7655f5bb9a14c8d",
+    "galleries/overview.png": "4d98499c23cab6dd507052cf3ff915516a1b850e0309c632f4db2b9b0b8115b7",
+    "galleries/progressions/l0-01.png": "84f4f96980efcf42ddd397934d0550d223e4bf43883368e575c3a48c7cf5bd0b",
+    "galleries/progressions/l0-02.png": "842b666c1673caa193e7620c48becdbd6a762754d8e8998745b39d4ba7e790c2",
+    "galleries/progressions/l0-03.png": "28ab0c3c1d78e0d1e9970d664a0f226eca331776f314290b2e6041eeccb49b7c",
+    "galleries/progressions/l0-04.png": "8d088f596dbbbba6fb211ec26f4e0103964244a70f720b4b5f6cb7dbb656bc91",
+    "galleries/progressions/l0-05.png": "9420c982329ddda49bf882b18390fe5a3447bae2dae73e31c48d00264541bca7",
+    "galleries/progressions/l0-06.png": "228dea635951049f7450b9a403f68ea8881b588c0d76ad2f4ed1b3a3f19008be",
+    "galleries/progressions/l0-07.png": "20fae52cceeec61826d6426f18b48f7e7532b54fddce69d886119026fb2c90c0",
+    "galleries/progressions/l0-08.png": "64eeeea2f3e30fc07725860dbfde1745408f35a660bd4f2e24d5b6be14d0710e",
+    "galleries/progressions/l1-01.png": "d9e8fe52985a0698c920a47888bebd19e5732a268509cf6031abc83f03b62f54",
+    "galleries/progressions/l1-02.png": "ad7d2f868761cd3e025fab75c4df6e7b7777b8ec49c4f66cd0f766dc821c25bd",
+    "galleries/progressions/l1-03.png": "70f0e93478bf585a9adb99359ed46e069a810f26ac09f547ba07793ab879c161",
+    "galleries/progressions/l1-04.png": "dc0b3e473e3eb66baca629c1979c0172b0c77719d644eab149404f5b12c2bf49",
+    "galleries/progressions/l1-05.png": "1a6d4e8994ed91773fc60f97c583e1e081c504b9888729ea2d49a88d0f8fdd2d",
+    "galleries/progressions/l1-06.png": "546403b734044b7511a39c77fcd623956b6cbe86a789133b8613049dcb91f672",
+    "galleries/progressions/l1-07.png": "fdec99ac62658676ab8b79d94b75a934e6db2878c460b50c8ae1afa6e243d80b",
+    "galleries/progressions/l1-08.png": "9dc603aa2c849b21948712ed3f8026d7103a17441f8371216bd87aead9fc8e76",
+    "galleries/progressions/l1-09.png": "38a23268d00e2bbd910e488c24ae8dc8e01cbb36e70922c8756a55626a2e443d",
+    "galleries/progressions/l1-10.png": "a75d1f4d434428ed649db069f31a4f71ad8b7e4e0e181cb51dd82f94c644f613",
+    "galleries/progressions/l1-11.png": "aab5c0228808575112115a9746a88fadf7cdb24d164bad065e00dab883d188c7",
+    "galleries/progressions/l2-01.png": "003e36d604140e050296361166b920a12665ebfaa9b8b8212223296e43cf7eab",
+    "galleries/progressions/l2-02.png": "c6d0d714b7c69ad7858622920863586f09e6e7d2db10d8de1793c5d125ce45f2",
+    "galleries/progressions/l2-03.png": "8b995f8533766e34aa96ef261b34fac48f852a6511b15bf447d3958fc1b82400",
+    "galleries/progressions/l2-04.png": "862c081ed46f03b2fbce974ef3f51dd481bf1e8f6f2c6b826ab4e11072644170",
+    "galleries/progressions/l2-05.png": "f2c2998c6ead103bbd51d301d7c80007cd82c1ef348447c5de74e161cae64d2c",
+    "galleries/progressions/l2-06.png": "a611f563eb5b6704682d81fed3dac4cc2272eb13fe9cb275328a3ad458e7adcc",
+    "galleries/progressions/l2-07.png": "69f3d5f524a3e83c2db4fb90cb8b359067799cac02ccfc65fdb496cebdf38b56",
+    "galleries/progressions/l2-08.png": "7011e780a0c8ff1e0e9876c6f2386374498144f95f8045a76ecf458fd9d2cd12",
+    "galleries/progressions/l2-09.png": "aac57155eb4ff276f636fa57eddeb499eedf1730f3297aa919a8b8dc76c05992",
+    "galleries/progressions/l3-01.png": "034a81a1ec99f1a368d905a457cb992699ac1c61bd01be49ff786376e89d0ebd",
+    "galleries/progressions/l3-02.png": "271e1cb5177395c848da2dbdf9ad370a737381fc59116c769a158e33de769987",
+    "galleries/progressions/l3-03.png": "ef09ba9b692c535892fa86b33a0d1685c6b59b3daed6829bab1cf1f60235c9e7",
+    "galleries/progressions/l3-04.png": "6bd68e776c54a457c57ec8afc8355540295503c2c7073170b3e980702e56dee1",
+    "galleries/progressions/l3-05.png": "f5201d064da975252f8e654ab8fbcf027a654f936397dd8bace2c6eff87ebaa9",
+    "galleries/progressions/l3-06.png": "b48175848b919308e2008d8b23d68f19b6c15df4d841949cc69533c32a3bd7e1",
+    "galleries/progressions/l3-07.png": "ff8832022c99d94ba5f635969bd18a4405c5c8bd08ea06f11574522ae0ff9868",
+    "galleries/progressions/l3-08.png": "bc3d8b67c1c25f1a03f49d1292727ce35d4b7f45c097b62cc5d0518da8c066ef",
+    "galleries/progressions/l3-09.png": "77415db961e4bad9d12ecc8a27e26ac9691c6eb38026fb5f862ef783a80225e4",
+    "galleries/progressions/l3-10.png": "f7f8858e43efe1a80e763d47af8a429bc24091191b92580720339eb2d4d9f2b9",
+    "galleries/progressions/l4-01.png": "2b7f0aac0abe59c8bc91f84ab951f514a4eb2a4d53b54285510b738063342fcd",
+    "galleries/progressions/l4-02.png": "bce171b535d4401a79c5ad90681e4a0cc5ecb48d853ba01aaa4789da6d3ebf47",
+    "galleries/progressions/l4-03.png": "50dfcd1503b2870c243e78f86f2f6efd29b0b0e078d3a1bc6be4a7736c882ec5",
+    "galleries/progressions/l4-04.png": "3e0967ec887f865dda0a06ba0219803e05880719080694ec37ba540348a83b7c",
+    "galleries/progressions/l4-05.png": "86ad628c4ab7b972edba974f78464dfe3352f812d290c9d04c9d64cf5afc4e8b",
+    "galleries/progressions/l4-06.png": "af6489a9d6b8563ff1e6b1a6e37c6632311d8816ddd79a7f48d99365e43aa10c",
+    "galleries/progressions/l4-07.png": "aa3fca29b8a9f5389d03937eea41d80a8dd4fedba64598bde970d0b833beebd2",
+    "galleries/progressions/l4-08.png": "cae839ffdc53c2764e1dedb82be1c3c6479c4e74e23e827497457d782bb3abab",
+    "galleries/progressions/l4-09.png": "9b8ff9d52155f66cf7f0f0a748bbda92b69d31db0a06c0e746bd2a5741329eb4",
+    "metadata/audit_report.json": "b8b26d46f48d3b5e882c3f7b0e906f31eaa5449d9bf7db4f6b168fd7d484fa11",
+    "metadata/catalog.json": "595b700aaea0e8f9883d0f5aad588d5c1dc98f1aa12db8d19ac37e3095b90032",
+    "metadata/gallery_manifest.json": "5f62dae4cd32f4cfa268c785368d6833e56b6f213e118301afad9407bb5fa640",
+    "metadata/manifest.json": "e40017affd7bc3fc1acf39ff05167681c8436fc38a1cbe8ba953149079373133"
+  },
+  "release_name": "depth-v1",
+  "row_count": 4560,
+  "schema_version": "pixcell-depth-freeze-v1"
+}