Upload model
Browse files- config.json +99 -0
- model.safetensors +3 -0
config.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ESSModel"
|
| 4 |
+
],
|
| 5 |
+
"model": {
|
| 6 |
+
"conformer": {
|
| 7 |
+
"attention_dropout_p": 0.2,
|
| 8 |
+
"conv_dropout_p": 0.2,
|
| 9 |
+
"conv_expansion_factor": 2,
|
| 10 |
+
"conv_kernel_size": 7,
|
| 11 |
+
"decoder_dim": 256,
|
| 12 |
+
"encoder_dim": 256,
|
| 13 |
+
"feed_forward_dropout_p": 0.2,
|
| 14 |
+
"feed_forward_expansion_factor": 4,
|
| 15 |
+
"half_step_residual": true,
|
| 16 |
+
"num_attention_heads": 2,
|
| 17 |
+
"num_decode_layers": 6,
|
| 18 |
+
"num_encode_layers": 4
|
| 19 |
+
},
|
| 20 |
+
"max_seq_len": 1000,
|
| 21 |
+
"mode": "train",
|
| 22 |
+
"num_emotion": 10,
|
| 23 |
+
"reference_encoder": {
|
| 24 |
+
"dropout": 0.2,
|
| 25 |
+
"encoder_dim": 128
|
| 26 |
+
},
|
| 27 |
+
"variance_embedding": {
|
| 28 |
+
"energy_quantization": "linear",
|
| 29 |
+
"n_bins": 256,
|
| 30 |
+
"pitch_quantization": "linear"
|
| 31 |
+
},
|
| 32 |
+
"variance_predictor": {
|
| 33 |
+
"dropout": 0.5,
|
| 34 |
+
"filter_size": 256,
|
| 35 |
+
"kernel_size": 3
|
| 36 |
+
},
|
| 37 |
+
"vocoder": {
|
| 38 |
+
"model": "HiFi-GAN",
|
| 39 |
+
"speaker": "tth"
|
| 40 |
+
}
|
| 41 |
+
},
|
| 42 |
+
"model_type": "emofs2",
|
| 43 |
+
"preprocess": {
|
| 44 |
+
"dataset": "vlsp2023emo",
|
| 45 |
+
"emotion2id": {
|
| 46 |
+
"angry": 3,
|
| 47 |
+
"happy": 1,
|
| 48 |
+
"neutral": 0,
|
| 49 |
+
"sad": 2,
|
| 50 |
+
"surprise": 4
|
| 51 |
+
},
|
| 52 |
+
"id2emotion": {
|
| 53 |
+
"0": "neutral",
|
| 54 |
+
"1": "happy",
|
| 55 |
+
"2": "sad",
|
| 56 |
+
"3": "angry",
|
| 57 |
+
"4": "surprise"
|
| 58 |
+
},
|
| 59 |
+
"path": {
|
| 60 |
+
"corpus_path": "./data/pretrained_tts_dataset/tuyendv.dict",
|
| 61 |
+
"lexicon_path": "../datasets/ess-vlsp2023-lexicon/lexicon.dict",
|
| 62 |
+
"preprocessed_path": "../datasets/ess-vlsp2023-emo-processed-phoneme-level",
|
| 63 |
+
"raw_path": "./data/pretrained_tts_dataset_raw"
|
| 64 |
+
},
|
| 65 |
+
"preprocessing": {
|
| 66 |
+
"audio": {
|
| 67 |
+
"max_wav_value": 32768.0,
|
| 68 |
+
"sampling_rate": 22050
|
| 69 |
+
},
|
| 70 |
+
"energy": {
|
| 71 |
+
"feature": "phoneme_level",
|
| 72 |
+
"normalization": true
|
| 73 |
+
},
|
| 74 |
+
"mel": {
|
| 75 |
+
"mel_fmax": 8000,
|
| 76 |
+
"mel_fmin": 0,
|
| 77 |
+
"n_mel_channels": 80
|
| 78 |
+
},
|
| 79 |
+
"pitch": {
|
| 80 |
+
"feature": "phoneme_level",
|
| 81 |
+
"normalization": true
|
| 82 |
+
},
|
| 83 |
+
"stft": {
|
| 84 |
+
"filter_length": 1024,
|
| 85 |
+
"hop_length": 256,
|
| 86 |
+
"win_length": 1024
|
| 87 |
+
},
|
| 88 |
+
"text": {
|
| 89 |
+
"language": "en",
|
| 90 |
+
"text_cleaners": []
|
| 91 |
+
},
|
| 92 |
+
"val_size": 512
|
| 93 |
+
},
|
| 94 |
+
"smoothing_label": 0.1
|
| 95 |
+
},
|
| 96 |
+
"torch_dtype": "float32",
|
| 97 |
+
"train": null,
|
| 98 |
+
"transformers_version": "4.35.2"
|
| 99 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d47faa7882bb93e619fc5f1c65a13f373e84d921a11e45f3e1b318f0874619be
|
| 3 |
+
size 5402404
|