Update README.md
Browse files
README.md
CHANGED
|
@@ -23,15 +23,15 @@ It achieves the following results on the evaluation set:
|
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
| 26 |
-
|
|
|
|
| 27 |
|
| 28 |
-
## Intended uses & limitations
|
| 29 |
-
|
| 30 |
-
More information needed
|
| 31 |
|
| 32 |
## Training and evaluation data
|
| 33 |
|
| 34 |
-
|
|
|
|
|
|
|
| 35 |
|
| 36 |
## Training procedure
|
| 37 |
|
|
|
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
| 26 |
+
Distilhubert is distilled version of the [HuBERT](https://huggingface.co/docs/transformers/model_doc/hubert) and pretrained on data set with 16k frequency. <br/>
|
| 27 |
+
Architecture of this model is CTC or Connectionist Temporal Classification is a technique that is used with encoder-only transformer. <br/>
|
| 28 |
|
|
|
|
|
|
|
|
|
|
| 29 |
|
| 30 |
## Training and evaluation data
|
| 31 |
|
| 32 |
+
Training + Evaluation data set is GTZAN which is a popular dataset of 999 songs for music genre classification. <br/>
|
| 33 |
+
Each song is a 30-second clip from one of 10 genres of music, spanning disco to metal.<br/>
|
| 34 |
+
Train set is 899 songs and Evaluation set is 100 songs remainings.
|
| 35 |
|
| 36 |
## Training procedure
|
| 37 |
|