| { | |
| "architectures": [ | |
| "BiLSTMAttentionClassifier" | |
| ], | |
| "model_type": "wavlm-bilstm-attention", | |
| "base_encoder": "microsoft/wavlm-base-plus", | |
| "target_sample_rate": 16000, | |
| "input_size": 768, | |
| "hidden_size": 128, | |
| "num_classes": 6, | |
| "dropout": 0.3, | |
| "id2label": { | |
| "0": "Anger", | |
| "1": "Disgust", | |
| "2": "Fear", | |
| "3": "Happy", | |
| "4": "Neutral", | |
| "5": "Sad" | |
| }, | |
| "label2id": { | |
| "Anger": 0, | |
| "Disgust": 1, | |
| "Fear": 2, | |
| "Happy": 3, | |
| "Neutral": 4, | |
| "Sad": 5 | |
| } | |
| } | |