Add files using upload-large-folder tool

README.md

@@ -11,6 +11,14 @@ tags:
 task_categories:
 - image-to-image
 configs:
+- config_name: healthy_bone_cell
+  data_files:
+  - split: train
+    path: data/train-*.parquet
+  - split: validation
+    path: data/validation-*.parquet
+  - split: test
+    path: data/test-*.parquet
 - config_name: external_waveplate
   default: true
   data_files:
@@ -39,11 +47,13 @@ Current Hugging Face release status:
 
 - Uploaded: external waveplate test data at 633 nm.
 - Uploaded: external spectral test data at 610, 650, and 690 nm.
-- Not yet uploaded: healthy bone cell `train`, `validation`, and `test` splits.
+- Prepared locally for upload: healthy bone cell `train`, `validation`, and
+  `test` splits.
 
-Because the waveplate tensors are 200 x 200 while the spectral tensors are
-256 x 256, the external data is published as two separate configs:
+Because the waveplate tensors are 200 x 200 while the healthy bone cell and
+spectral tensors are 256 x 256, the data is published as separate configs:
 
+- `healthy_bone_cell`
 - `external_waveplate`
 - `external_spectral`
 
@@ -58,15 +68,14 @@ target modalities. It is not a segmentation or classification dataset.
 
 ## Data Sources
 
-This release contains external test data from `polarization_v3`:
+This release contains healthy bone cell data from `polarization_v2` and external
+test data from `polarization_v3`:
 
+- Healthy bone cell data: source-provided patch splits from 53 sample folders.
 - Waveplate data: `hwp633` and `qwp633`, measured at 633 nm.
 - Multi-wavelength spectral data: selected wavelengths from `mwl_530_690`,
   currently 610, 650, and 690 nm.
 
-The healthy bone cell data from `polarization_v2` is planned for a later upload
-as `train`, `validation`, and `test`.
-
 ## File Structure
 
 ```text
@@ -76,10 +85,15 @@ MMPD-Bench/
 │   ├── external_waveplate-00000-of-00001.parquet
 │   ├── external_spectral_610-00000-of-00001.parquet
 │   ├── external_spectral_650-00000-of-00001.parquet
-│   └── external_spectral_690-00000-of-00001.parquet
+│   ├── external_spectral_690-00000-of-00001.parquet
+│   ├── train-00000-of-00094.parquet
+│   ├── validation-00000-of-00012.parquet
+│   └── test-00000-of-00011.parquet
 └── metadata/
     ├── acquisition_info.json
     ├── channel_order.json
+    ├── healthy_bone_cell_manifest.jsonl
+    ├── healthy_bone_cell_manifest_summary.json
     ├── parameter_ranges.json
     ├── schema.json
     └── split_summary.json
@@ -94,9 +108,9 @@ Common columns:
     "sample_id": str,
     "source_id": str,
     "split": str,
-    "subset": str,          # waveplate or spectral in the current release
-    "specimen_type": str,   # waveplate or spectral in the current release
-    "wavelength_nm": int,
+    "subset": str,          # healthy_bone_cell, waveplate, or spectral
+    "specimen_type": str,   # healthy_bone_cell, waveplate, or spectral
+    "wavelength_nm": int | None,
     "source_path": str,
     "mueller_shape": list[int],
     "target_shape": list[int],
@@ -115,7 +129,7 @@ Waveplate-specific columns:
 }
 ```
 
-Spectral-specific columns:
+Patch-based columns for healthy bone cell and spectral rows:
 
 ```python
 {
@@ -126,6 +140,8 @@ Spectral-specific columns:
 
 Current tensor shapes:
 
+- `healthy_bone_cell`: `mueller = [16, 256, 256]`,
+  `target = [6, 256, 256]`.
 - `external_waveplate`: `mueller = [16, 200, 200]`, `target = [6, 200, 200]`.
 - `external_spectral_*`: `mueller = [16, 256, 256]`,
   `target = [6, 256, 256]`.
@@ -170,8 +186,8 @@ theta, psi: [-pi/2, pi/2)
 Important encoding note:
 
 - Waveplate target arrays are stored from the source `.npy` files as float32.
-- Spectral target arrays were converted from grayscale PNG files to float32
-  values normalized to `[0, 1]`; see `target_encoding`.
+- Healthy bone cell and spectral target arrays were converted from grayscale PNG
+  files to float32 values normalized to `[0, 1]`; see `target_encoding`.
 - Mueller matrix tensors are stored as measured/processed values, not forcibly
   clipped to `[-1, 1]`.
 
@@ -241,31 +257,26 @@ annotations or absolute biological ground truth.
 
 | Split | Config | Subset | Samples | Shape | Notes |
 |---|---|---:|---:|---|---|
+| train | healthy_bone_cell | healthy_bone_cell | 6006 | `[16, 256, 256] -> [6, 256, 256]` | 94 shards |
+| validation | healthy_bone_cell | healthy_bone_cell | 713 | `[16, 256, 256] -> [6, 256, 256]` | 12 shards |
+| test | healthy_bone_cell | healthy_bone_cell | 643 | `[16, 256, 256] -> [6, 256, 256]` | 11 shards |
 | external_waveplate | external_waveplate | waveplate | 18 | `[16, 200, 200] -> [6, 200, 200]` | 633 nm HWP/QWP |
 | external_spectral_610 | external_spectral | spectral | 165 | `[16, 256, 256] -> [6, 256, 256]` | 610 nm |
 | external_spectral_650 | external_spectral | spectral | 165 | `[16, 256, 256] -> [6, 256, 256]` | 650 nm |
 | external_spectral_690 | external_spectral | spectral | 165 | `[16, 256, 256] -> [6, 256, 256]` | 690 nm |
-| train | not uploaded yet | healthy_bone_cell | TBD | TBD | planned |
-| validation | not uploaded yet | healthy_bone_cell | TBD | TBD | planned |
-| test | not uploaded yet | healthy_bone_cell | TBD | TBD | planned |
 
 ## Benchmark Protocols
 
-Current release:
+Evaluation configs:
 
-1. External waveplate evaluation: use config `external_waveplate`, split
+1. Healthy bone cell benchmark: use config `healthy_bone_cell`, splits `train`,
+   `validation`, and `test`.
+2. External waveplate evaluation: use config `external_waveplate`, split
    `external_waveplate`.
-2. External spectral evaluation: use config `external_spectral`, then evaluate
+3. External spectral evaluation: use config `external_spectral`, then evaluate
    `external_spectral_610`, `external_spectral_650`, and
    `external_spectral_690`.
 
-Planned full benchmark:
-
-1. Train on healthy bone cell `train`.
-2. Tune on healthy bone cell `validation`.
-3. Report clean performance on healthy bone cell `test`.
-4. Report external generalisation on waveplate and spectral splits.
-
 ## Loading Instructions
 
 Install the Hugging Face datasets package:
@@ -305,6 +316,7 @@ Load via dataset config:
 ```python
 from datasets import load_dataset
 
+healthy = load_dataset("HY2333/MMPD_Bench", "healthy_bone_cell")
 spectral = load_dataset("HY2333/MMPD_Bench", "external_spectral")
 waveplate = load_dataset("HY2333/MMPD_Bench", "external_waveplate")
 ```
@@ -315,10 +327,9 @@ test, use non-streaming loading on a single split as shown above.
 
 ## Ethics and Limitations
 
-The current release contains external physical/spectral evaluation data, not the
-healthy bone cell train/validation/test splits. Biological-data ethics and
-de-identification details should be completed before publishing the healthy bone
-cell splits.
+The current public release focuses on healthy bone cell and external
+physical/spectral evaluation data. Diseased biological samples are not included
+in this release.
 
 The targets are Lu-Chipman reference outputs. Evaluation should be interpreted
 as agreement with a physics-solver reference and related physics consistency,

data/README.md

@@ -11,10 +11,10 @@ external_spectral_650-*.parquet # 165 rows
 external_spectral_690-*.parquet # 165 rows
 ```
 
-Planned but not yet uploaded:
+Generated locally for the healthy_bone_cell upload:
 
 ```text
-train-*.parquet                 # healthy_bone_cell
-validation-*.parquet            # healthy_bone_cell
-test-*.parquet                  # healthy_bone_cell
+train-*.parquet                 # healthy_bone_cell, 6006 rows, 94 shards
+validation-*.parquet            # healthy_bone_cell, 713 rows, 12 shards
+test-*.parquet                  # healthy_bone_cell, 643 rows, 11 shards
 ```

data/test-00007-of-00011.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1ec85ba8f481770f182334437ef90de84ff98bf95d83a875d6c47b8fca0fe83
+size 248098890

data/test-00008-of-00011.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21e6f5849b0f7966668ad5e8fc865d92027e65f4582f7bc652bada24a0ba8e18
+size 248570491

data/test-00009-of-00011.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4a87a29d3fe43572353786e28f43a6645e52bfc2d1f39ac240dafd472fdbb53
+size 247866094

data/train-00000-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3b256fa129e66840fb8d7253bb9ac3f862ff3e1393a69a8afc6c6c25fbd6780
+size 246872076

data/train-00001-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e531f2100e49c5a3bdb7414c12b7659eb5e6cb5e22ee4f927ae311a8a0d2606
+size 247525181

data/train-00002-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2136257ae9f3a5e58380354652da7d0ff16662be02cb78435cf092b001a1cc9c
+size 247410356

data/train-00003-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7974e515728126eff7ecf282cd34a01ff28a7265459609dc8eddd19523dc4fbf
+size 246393485

data/train-00004-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f74ac824ce0dcbd117aaaa6a1dab23463f28aa8a8271edf69608541d09bf2e80
+size 246514272

data/train-00005-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40c0a03440b5084fa7dd7730cd39d42eb44ac8f510e29a3df089fc3bd269ff20
+size 246377523

data/train-00006-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:053f84330ed633bdffd38cbebd3e181b3859d5964674c71f71d89c425ae54c31
+size 246393733

data/train-00007-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b018199f29b9525eaa68cf51e329c6d6de4d811f28958670cd562abb2c3e3b7b
+size 249206392

data/train-00008-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d3b10ac87f3b165ee25b85dcd564aeccb3cbb447670dc463d903437233594b9
+size 248055074

data/train-00009-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afa086a68f2ad07a55b548fd2cb66c3d44c4151ad8af58cd88343ef173b93858
+size 248079707

data/train-00010-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86b8f4e731e9b52637a8a41ca3e81f83232643e5e14b7bc9153a2cc383be108c
+size 247075332

data/train-00011-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5de1574be76ac06ca7f05261d218ca33f39b9bcb800de0174cc1298b6eb6cf39
+size 248151802

data/train-00012-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7536ef79048c815c0d21860183e43cd3e67c0e8e4a8605cbd2223680f4f09c65
+size 247665264

data/train-00013-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65c9eda2bacc80ddab3c86ddab0a6d75aff51a984f5be5ec3e6624bfcb005147
+size 248192924

data/train-00014-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbef1dd5111fb5fb4d90427dfa4e3de047f29052422c6b095dd82832c942c527
+size 248368589

data/train-00015-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ccd688edd14ae4ab361fec431cac3f7da79eb57ed5f4eec03246d2d6d0413f3
+size 248247825

data/train-00016-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e14d7878091d79c7916ca34ce54a249e449bc19c721e31584ccc78d6041b55fc
+size 247967640

data/train-00017-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b833c6d6167e9dda9448428df994e4050365ca097fb6ab83c1e0c2fd29ccd3d
+size 247482880

data/train-00018-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45de916940293d317fa70a3b11ccb08f06f47f111dc4335a7ab5514cf13cf146
+size 248841155

data/train-00019-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fa28c96f8c5b3b3309ee0f46b2a93759f7543732dfa244226dbebeb691bf289
+size 250708411

data/train-00020-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24a789de234cf7fdab62279d9af1836b47378219bacc96a26ca88145fe86f130
+size 248119103

data/train-00021-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e39550ce99e7c64c78c329ff2a287298151940075c4ba7f4b7e2c937cd51f00
+size 248195557

data/train-00022-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59256e9a9ca136b6b86007928f368d007226aa85aee7184b0a2e96cdcb0c6421
+size 246808836

data/train-00023-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e161763895c1cce385b0a1ed18f4f6cce98930f6ed2596660a909aa6b4a2e900
+size 247529473

data/train-00024-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79f23e564c5edfae2620f9c9fe4ab0ce3d44c0ec220171cbec13699e6add7f86
+size 247941542

data/train-00025-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a7c33e4916062ba66379515854f4074a1f621f17c7173d929d966aaea9eae6e
+size 250761736

data/train-00026-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8441d1852ea500633d39dc18370da03f3c4f9b896a3ef83fea157ce9684ee7c4
+size 248875062

data/train-00027-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46a2ff79e3ad9e911dbfb4575c8c13bf64bc819edfddfdc700beb919a06d6f06
+size 247179918

data/train-00028-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d8b06faece314ac81b96619a2e0ad1f8776b9175f03d5eda4ddb63cdcff2cc4
+size 247032890

data/train-00029-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f08fc511ad78cd865b6d1e6d01adac386278dd2569463beff3b874cb5636137
+size 245995183

data/train-00030-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63087c5bcc4546fa9bc83fd4e15fa3a04a81e5c5c1dacbc31635ae9c2903f771
+size 246006482

data/train-00031-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a33549a5dd4151cac9825486d9a0de346b67a3588afd8ad420854353e0e74b45
+size 246377008

data/train-00032-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bc178aada068c9832cae91e4c062f0e8a2ed60d5905ce63f86cb1e7a4ba1836
+size 246503223

data/train-00033-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85537d0de271975c5992ae989941cb098988cd5e1ab199512a3e69aad8510816
+size 246808384

data/train-00034-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeee11d51e06d1eae7955df14b12eb57d4ecf3984de91111551f1dca5eb4acf0
+size 247145744

data/train-00035-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af4201ddaed459bff66f4e54c60c011a2d32f4ee02386d69b28ce89629aaa267
+size 247220086

data/train-00036-of-00094.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7196443ceed275676fd6db2a1e5f3e7a032b656a99ac8c18346b4d3cc676578b
+size 245281347

metadata/healthy_bone_cell_manifest_summary.json

@@ -0,0 +1,277 @@
+{
+  "counts": {
+    "test": 643,
+    "train": 6006,
+    "validation": 713
+  },
+  "errors": [],
+  "per_source_counts": {
+    "x_0_y_0": {
+      "test": 10,
+      "train": 114,
+      "validation": 13
+    },
+    "x_0_y_2": {
+      "test": 18,
+      "train": 168,
+      "validation": 20
+    },
+    "x_0_y_3": {
+      "test": 14,
+      "train": 130,
+      "validation": 15
+    },
+    "x_10_y_0": {
+      "test": 4,
+      "train": 36,
+      "validation": 4
+    },
+    "x_10_y_1": {
+      "test": 5,
+      "train": 45,
+      "validation": 5
+    },
+    "x_10_y_2": {
+      "test": 10,
+      "train": 96,
+      "validation": 11
+    },
+    "x_10_y_3": {
+      "test": 12,
+      "train": 108,
+      "validation": 13
+    },
+    "x_11_y_0": {
+      "test": 3,
+      "train": 27,
+      "validation": 3
+    },
+    "x_11_y_1": {
+      "test": 4,
+      "train": 38,
+      "validation": 4
+    },
+    "x_11_y_2": {
+      "test": 8,
+      "train": 68,
+      "validation": 8
+    },
+    "x_11_y_3": {
+      "test": 9,
+      "train": 82,
+      "validation": 10
+    },
+    "x_12_y_0": {
+      "test": 7,
+      "train": 67,
+      "validation": 8
+    },
+    "x_12_y_1": {
+      "test": 9,
+      "train": 88,
+      "validation": 10
+    },
+    "x_12_y_2": {
+      "test": 9,
+      "train": 85,
+      "validation": 10
+    },
+    "x_12_y_3": {
+      "test": 14,
+      "train": 130,
+      "validation": 16
+    },
+    "x_13_y_0": {
+      "test": 9,
+      "train": 81,
+      "validation": 9
+    },
+    "x_13_y_1": {
+      "test": 9,
+      "train": 89,
+      "validation": 10
+    },
+    "x_13_y_2": {
+      "test": 5,
+      "train": 81,
+      "validation": 9
+    },
+    "x_13_y_3": {
+      "test": 13,
+      "train": 164,
+      "validation": 19
+    },
+    "x_14_y_0": {
+      "test": 14,
+      "train": 115,
+      "validation": 14
+    },
+    "x_14_y_1": {
+      "test": 20,
+      "train": 161,
+      "validation": 20
+    },
+    "x_14_y_2": {
+      "test": 20,
+      "train": 187,
+      "validation": 23
+    },
+    "x_14_y_3": {
+      "test": 17,
+      "train": 158,
+      "validation": 19
+    },
+    "x_1_y_0": {
+      "test": 11,
+      "train": 89,
+      "validation": 11
+    },
+    "x_1_y_1": {
+      "test": 18,
+      "train": 178,
+      "validation": 21
+    },
+    "x_1_y_2": {
+      "test": 21,
+      "train": 204,
+      "validation": 25
+    },
+    "x_1_y_3": {
+      "test": 19,
+      "train": 161,
+      "validation": 19
+    },
+    "x_2_y_0": {
+      "test": 6,
+      "train": 50,
+      "validation": 6
+    },
+    "x_2_y_1": {
+      "test": 17,
+      "train": 136,
+      "validation": 17
+    },
+    "x_2_y_2": {
+      "test": 20,
+      "train": 192,
+      "validation": 23
+    },
+    "x_2_y_3": {
+      "test": 19,
+      "train": 172,
+      "validation": 21
+    },
+    "x_3_y_0": {
+      "test": 3,
+      "train": 35,
+      "validation": 4
+    },
+    "x_3_y_1": {
+      "test": 11,
+      "train": 101,
+      "validation": 12
+    },
+    "x_3_y_2": {
+      "test": 16,
+      "train": 133,
+      "validation": 16
+    },
+    "x_3_y_3": {
+      "test": 21,
+      "train": 171,
+      "validation": 21
+    },
+    "x_5_y_1": {
+      "test": 5,
+      "train": 75,
+      "validation": 8
+    },
+    "x_5_y_2": {
+      "test": 15,
+      "train": 131,
+      "validation": 16
+    },
+    "x_5_y_3": {
+      "test": 16,
+      "train": 144,
+      "validation": 17
+    },
+    "x_6_y_1": {
+      "test": 11,
+      "train": 94,
+      "validation": 11
+    },
+    "x_6_y_2": {
+      "test": 17,
+      "train": 161,
+      "validation": 19
+    },
+    "x_6_y_3": {
+      "test": 15,
+      "train": 127,
+      "validation": 15
+    },
+    "x_7_y_0": {
+      "test": 4,
+      "train": 44,
+      "validation": 5
+    },
+    "x_7_y_1": {
+      "test": 12,
+      "train": 132,
+      "validation": 16
+    },
+    "x_7_y_2": {
+      "test": 17,
+      "train": 181,
+      "validation": 21
+    },
+    "x_7_y_3": {
+      "test": 10,
+      "train": 100,
+      "validation": 12
+    },
+    "x_8_y_0": {
+      "test": 6,
+      "train": 72,
+      "validation": 8
+    },
+    "x_8_y_1": {
+      "test": 14,
+      "train": 124,
+      "validation": 15
+    },
+    "x_8_y_2": {
+      "test": 19,
+      "train": 165,
+      "validation": 20
+    },
+    "x_8_y_3": {
+      "test": 12,
+      "train": 97,
+      "validation": 12
+    },
+    "x_9_y_0": {
+      "test": 7,
+      "train": 58,
+      "validation": 7
+    },
+    "x_9_y_1": {
+      "test": 10,
+      "train": 115,
+      "validation": 13
+    },
+    "x_9_y_2": {
+      "test": 17,
+      "train": 145,
+      "validation": 17
+    },
+    "x_9_y_3": {
+      "test": 11,
+      "train": 101,
+      "validation": 12
+    }
+  },
+  "source_dirs": 53,
+  "total": 7362
+}

metadata/schema.json

@@ -27,18 +27,18 @@
     },
     "wavelength_nm": {
       "type": "int32",
-      "required": true,
-      "description": "Measurement wavelength in nanometers."
+      "required": false,
+      "description": "Measurement wavelength in nanometers. External rows provide this value; healthy_bone_cell rows use null when no single wavelength label is available."
     },
     "patch_id": {
       "type": "string",
       "required": false,
-      "description": "Patch identifier for external_spectral_* rows, e.g. patch_0000."
+      "description": "Patch identifier for patch-based rows, e.g. patch_0000."
     },
     "target_encoding": {
       "type": "string",
       "required": false,
-      "description": "Encoding of target values. Spectral pilot rows use png_uint8_normalized_to_float32_0_1."
+      "description": "Encoding of target values. Spectral and healthy_bone_cell PNG-derived rows use png_uint8_normalized_to_float32_0_1."
     },
     "plate_type": {
       "type": "string",
@@ -63,12 +63,12 @@
     "mueller_shape": {
       "type": "int32[3]",
       "required": true,
-      "description": "Stored tensor shape. External waveplate rows are [16, 200, 200]; external_spectral_* rows are [16, 256, 256]."
+      "description": "Stored tensor shape. External waveplate rows are [16, 200, 200]; external_spectral_* and healthy_bone_cell rows are [16, 256, 256]."
     },
     "target_shape": {
       "type": "int32[3]",
       "required": true,
-      "description": "Stored tensor shape. External waveplate rows are [6, 200, 200]; external_spectral_* rows are [6, 256, 256]."
+      "description": "Stored tensor shape. External waveplate rows are [6, 200, 200]; external_spectral_* and healthy_bone_cell rows are [6, 256, 256]."
     },
     "mueller": {
       "type": "float32[channel,height,width]",
@@ -82,7 +82,6 @@
     }
   },
   "todo": [
-    "Confirm final column set before Parquet conversion.",
     "Decide whether to include preview_png or summary statistics."
   ]
 }

metadata/split_summary.json

@@ -1,23 +1,29 @@
 {
-  "split_policy": "TODO: specimen_level_or_patch_level",
-  "leakage_note": "TODO",
+  "split_policy": "Source-provided patch split from polarization_v2 healthy_bone_cell folders. Each source folder contains train, val, and test subfolders.",
+  "leakage_note": "healthy_bone_cell splits follow the existing source-folder patch split. Multiple patches from the same source_id can appear across train, validation, and test, so users who need specimen-level isolation should rebuild splits from metadata/healthy_bone_cell_manifest.jsonl.",
   "splits": {
     "train": {
       "subset": "healthy_bone_cell",
       "specimen_type": "healthy_bone_cell",
-      "samples": null,
+      "samples": 6006,
+      "source_dirs": 53,
+      "shards": 94,
       "path_pattern": "data/train-*.parquet"
     },
     "validation": {
       "subset": "healthy_bone_cell",
       "specimen_type": "healthy_bone_cell",
-      "samples": null,
+      "samples": 713,
+      "source_dirs": 53,
+      "shards": 12,
       "path_pattern": "data/validation-*.parquet"
     },
     "test": {
       "subset": "healthy_bone_cell",
       "specimen_type": "healthy_bone_cell",
-      "samples": null,
+      "samples": 643,
+      "source_dirs": 53,
+      "shards": 11,
       "path_pattern": "data/test-*.parquet"
     },
     "external_waveplate": {