Card: list the training data (datasets:, base_model:) and add a Training data section
Browse filesFollow-up to braindecode PR #1493: `from_pretrained` reads `datasets:` into `model.card_data`. Ids: MOABB names, NEMAR nm/on ids, OpenNeuro ds ids; evidence per checkpoint.
README.md
CHANGED
|
@@ -1,3 +1,25 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: bsd-3-clause
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: bsd-3-clause
|
| 3 |
+
datasets:
|
| 4 |
+
- TUEG
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
<!-- braindecode:training-data -->
|
| 9 |
+
## Training data
|
| 10 |
+
|
| 11 |
+
The data that shaped these weights. The ids are the `datasets:` field of this card,
|
| 12 |
+
which braindecode reads into `model.card_data.datasets` when you call
|
| 13 |
+
`from_pretrained` (MOABB class names for MOABB datasets, OpenNeuro / NEMAR
|
| 14 |
+
accessions, Hub ids otherwise). `nm`/`on` ids are datasets on [NEMAR](https://nemar.org)
|
| 15 |
+
(`on` = NEMAR's copy of the OpenNeuro dataset with the same number). Rows marked
|
| 16 |
+
`private:` are data the authors did not release; they are not in `datasets:`.
|
| 17 |
+
|
| 18 |
+
| `datasets:` id | Dataset | Stage | Modality |
|
| 19 |
+
|---|---|---|---|
|
| 20 |
+
| `TUEG` | [Temple University Hospital EEG Corpus (TUH EEG)](https://isip.piconepress.com/projects/nedc/html/tuh_eeg/) | pretraining | EEG |
|
| 21 |
+
|
| 22 |
+
- patch-based masked EEG reconstruction; 19 channels, 200 Hz, 30-s samples; 1,109,545 samples (>9,000 h) after cleaning; full TUEG (TUAB/TUEV not removed in the main model)
|
| 23 |
+
- `TUEG` (the TUH EEG Corpus) contains the recordings of its sub-corpora TUAB, TUEV, TUSZ, TUAR, TUSL and TUEP; an evaluation on one of those is not held out from these weights unless the authors removed it.
|
| 24 |
+
- Sources: arXiv:2412.07236 Sec. 3.1 (Pre-training Dataset); arXiv:2412.07236 Appendix E.7; https://github.com/wjq-learning/CBraMod README; tensor hash check: weighting666/CBraMod pretrained_weights.pth vs hosted model.safetensors.
|
| 25 |
+
|