Datasets:
Explain released splits, Complex source audio preparation, and Selective training
Browse files
README.md
CHANGED
|
@@ -173,6 +173,11 @@ Six tasks: Addition (0), Removal (1), Generation (2), Replacement (3), Volume Mo
|
|
| 173 |
|
| 174 |
Every folder contains `metadata.parquet`, `audio.zip` and `manifest.json` with SHA256 hashes. Audio members preserve their original bytes, member names and directory layout. Audio is 16 kHz mono PCM16, 10 seconds per sample. ZIP paths are stored relative to the archive, such as `audios/val/input/...wav`; changing the public folder name does not change these member paths. Complex val/test and Selective train contain input and reference-target audio. Selective val/test contain input audio only and have empty target paths. Selective val/test are evaluated with task-success rules. No target audio is synthesized or padded into the metadata. Complex training and the full raw pretraining corpus are outside this release.
|
| 175 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 176 |
## Read metadata
|
| 177 |
|
| 178 |
```python
|
|
|
|
| 173 |
|
| 174 |
Every folder contains `metadata.parquet`, `audio.zip` and `manifest.json` with SHA256 hashes. Audio members preserve their original bytes, member names and directory layout. Audio is 16 kHz mono PCM16, 10 seconds per sample. ZIP paths are stored relative to the archive, such as `audios/val/input/...wav`; changing the public folder name does not change these member paths. Complex val/test and Selective train contain input and reference-target audio. Selective val/test contain input audio only and have empty target paths. Selective val/test are evaluated with task-success rules. No target audio is synthesized or padded into the metadata. Complex training and the full raw pretraining corpus are outside this release.
|
| 175 |
|
| 176 |
+
## Continue training
|
| 177 |
+
|
| 178 |
+
- **Complex:** this repository provides `complex_val` and `complex_test` with GT for evaluation. For training, obtain source audio from the [datasets used in our training pools](https://github.com/GeWu-Lab/SteerSound/tree/main/STEER_post_training#complex-training-sources), or use your own recordings. Follow the [audio preparation guide](https://github.com/GeWu-Lab/SteerSound/blob/main/pretrain_and_standard_FM_finetune/dataset/README.md) to create `online_data.json`, then use a [Complex post-training guide](https://github.com/GeWu-Lab/SteerSound/tree/main/STEER_post_training#choose-a-training-workflow). There is no `complex_train` split in this release; editing examples are constructed online from your prepared source audio.
|
| 179 |
+
- **Selective:** download `selective_train`, `selective_val`, and `selective_test` and follow [Emphasis / Selective](https://github.com/GeWu-Lab/SteerSound/tree/main/STEER_post_training/emphasis/selective) or [Recovery / Selective](https://github.com/GeWu-Lab/SteerSound/tree/main/STEER_post_training/recovery/selective). The released training split already contains editing pairs.
|
| 180 |
+
|
| 181 |
## Read metadata
|
| 182 |
|
| 183 |
```python
|