Text-to-Speech
Transformers
Safetensors
parler_tts
text2text-generation
annotation
ylacombe commited on
Commit
e8ae059
1 Parent(s): 2793b94

Upload ParlerTTSForConditionalGeneration

Browse files
Files changed (3) hide show
  1. config.json +277 -0
  2. generation_config.json +10 -0
  3. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,277 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/fsx/yoach/tmp/artefacts/training-multilingual-mini-indic-3/",
3
+ "architectures": [
4
+ "ParlerTTSForConditionalGeneration"
5
+ ],
6
+ "audio_encoder": {
7
+ "_attn_implementation_autoset": false,
8
+ "_name_or_path": "ylacombe/dac_44khz",
9
+ "add_cross_attention": false,
10
+ "architectures": [
11
+ "DacModel"
12
+ ],
13
+ "bad_words_ids": null,
14
+ "begin_suppress_tokens": null,
15
+ "bos_token_id": null,
16
+ "chunk_size_feed_forward": 0,
17
+ "codebook_dim": 8,
18
+ "codebook_loss_weight": 1.0,
19
+ "codebook_size": 1024,
20
+ "commitment_loss_weight": 0.25,
21
+ "cross_attention_hidden_size": null,
22
+ "decoder_hidden_size": 1536,
23
+ "decoder_start_token_id": null,
24
+ "diversity_penalty": 0.0,
25
+ "do_sample": false,
26
+ "downsampling_ratios": [
27
+ 2,
28
+ 4,
29
+ 8,
30
+ 8
31
+ ],
32
+ "early_stopping": false,
33
+ "encoder_hidden_size": 64,
34
+ "encoder_no_repeat_ngram_size": 0,
35
+ "eos_token_id": null,
36
+ "exponential_decay_length_penalty": null,
37
+ "finetuning_task": null,
38
+ "forced_bos_token_id": null,
39
+ "forced_eos_token_id": null,
40
+ "hidden_size": 1024,
41
+ "hop_length": 512,
42
+ "id2label": {
43
+ "0": "LABEL_0",
44
+ "1": "LABEL_1"
45
+ },
46
+ "is_decoder": false,
47
+ "is_encoder_decoder": false,
48
+ "label2id": {
49
+ "LABEL_0": 0,
50
+ "LABEL_1": 1
51
+ },
52
+ "length_penalty": 1.0,
53
+ "max_length": 20,
54
+ "min_length": 0,
55
+ "model_type": "dac",
56
+ "n_codebooks": 9,
57
+ "no_repeat_ngram_size": 0,
58
+ "num_beam_groups": 1,
59
+ "num_beams": 1,
60
+ "num_return_sequences": 1,
61
+ "output_attentions": false,
62
+ "output_hidden_states": false,
63
+ "output_scores": false,
64
+ "pad_token_id": null,
65
+ "prefix": null,
66
+ "problem_type": null,
67
+ "pruned_heads": {},
68
+ "quantizer_dropout": 0.0,
69
+ "remove_invalid_values": false,
70
+ "repetition_penalty": 1.0,
71
+ "return_dict": true,
72
+ "return_dict_in_generate": false,
73
+ "sampling_rate": 44100,
74
+ "sep_token_id": null,
75
+ "suppress_tokens": null,
76
+ "task_specific_params": null,
77
+ "temperature": 1.0,
78
+ "tf_legacy_loss": false,
79
+ "tie_encoder_decoder": false,
80
+ "tie_word_embeddings": true,
81
+ "tokenizer_class": null,
82
+ "top_k": 50,
83
+ "top_p": 1.0,
84
+ "torch_dtype": "float32",
85
+ "torchscript": false,
86
+ "typical_p": 1.0,
87
+ "upsampling_ratios": [
88
+ 8,
89
+ 8,
90
+ 4,
91
+ 2
92
+ ],
93
+ "use_bfloat16": false
94
+ },
95
+ "decoder": {
96
+ "_attn_implementation_autoset": false,
97
+ "_name_or_path": "/fsx/yoach/tmp/artefacts/parler-tts-mini-v2-empty/decoder",
98
+ "activation_dropout": 0.0,
99
+ "activation_function": "gelu",
100
+ "add_cross_attention": true,
101
+ "architectures": [
102
+ "ParlerTTSForCausalLM"
103
+ ],
104
+ "attention_dropout": 0.0,
105
+ "bad_words_ids": null,
106
+ "begin_suppress_tokens": null,
107
+ "bos_token_id": 1025,
108
+ "chunk_size_feed_forward": 0,
109
+ "codebook_weights": null,
110
+ "cross_attention_hidden_size": null,
111
+ "cross_attention_implementation_strategy": null,
112
+ "decoder_start_token_id": null,
113
+ "delay_strategy": "delay",
114
+ "diversity_penalty": 0.0,
115
+ "do_sample": false,
116
+ "dropout": 0.1,
117
+ "early_stopping": false,
118
+ "encoder_no_repeat_ngram_size": 0,
119
+ "eos_token_id": 1024,
120
+ "exponential_decay_length_penalty": null,
121
+ "ffn_dim": 4096,
122
+ "finetuning_task": null,
123
+ "forced_bos_token_id": null,
124
+ "forced_eos_token_id": null,
125
+ "hidden_size": 1024,
126
+ "id2label": {
127
+ "0": "LABEL_0",
128
+ "1": "LABEL_1"
129
+ },
130
+ "initializer_factor": 0.02,
131
+ "is_decoder": true,
132
+ "is_encoder_decoder": false,
133
+ "label2id": {
134
+ "LABEL_0": 0,
135
+ "LABEL_1": 1
136
+ },
137
+ "layerdrop": 0.0,
138
+ "length_penalty": 1.0,
139
+ "max_length": 20,
140
+ "max_position_embeddings": 4096,
141
+ "min_length": 0,
142
+ "model_type": "parler_tts_decoder",
143
+ "no_repeat_ngram_size": 0,
144
+ "num_attention_heads": 16,
145
+ "num_beam_groups": 1,
146
+ "num_beams": 1,
147
+ "num_codebooks": 9,
148
+ "num_cross_attention_key_value_heads": 16,
149
+ "num_hidden_layers": 24,
150
+ "num_key_value_heads": 16,
151
+ "num_return_sequences": 1,
152
+ "output_attentions": false,
153
+ "output_hidden_states": false,
154
+ "output_scores": false,
155
+ "pad_token_id": 1024,
156
+ "prefix": null,
157
+ "problem_type": null,
158
+ "pruned_heads": {},
159
+ "remove_invalid_values": false,
160
+ "repetition_penalty": 1.0,
161
+ "return_dict": true,
162
+ "return_dict_in_generate": false,
163
+ "rope_embeddings": false,
164
+ "rope_theta": 10000.0,
165
+ "scale_embedding": false,
166
+ "sep_token_id": null,
167
+ "suppress_tokens": null,
168
+ "task_specific_params": null,
169
+ "temperature": 1.0,
170
+ "tf_legacy_loss": false,
171
+ "tie_encoder_decoder": false,
172
+ "tie_word_embeddings": false,
173
+ "tokenizer_class": null,
174
+ "top_k": 50,
175
+ "top_p": 1.0,
176
+ "torch_dtype": "float32",
177
+ "torchscript": false,
178
+ "typical_p": 1.0,
179
+ "use_bfloat16": false,
180
+ "use_cache": true,
181
+ "use_fused_lm_heads": true,
182
+ "vocab_size": 1088
183
+ },
184
+ "decoder_start_token_id": 1025,
185
+ "is_encoder_decoder": true,
186
+ "model_type": "parler_tts",
187
+ "pad_token_id": 1024,
188
+ "prompt_cross_attention": false,
189
+ "text_encoder": {
190
+ "_attn_implementation_autoset": false,
191
+ "_name_or_path": "google/flan-t5-large",
192
+ "add_cross_attention": false,
193
+ "architectures": [
194
+ "T5ForConditionalGeneration"
195
+ ],
196
+ "bad_words_ids": null,
197
+ "begin_suppress_tokens": null,
198
+ "bos_token_id": null,
199
+ "chunk_size_feed_forward": 0,
200
+ "classifier_dropout": 0.0,
201
+ "cross_attention_hidden_size": null,
202
+ "d_ff": 2816,
203
+ "d_kv": 64,
204
+ "d_model": 1024,
205
+ "decoder_start_token_id": 0,
206
+ "dense_act_fn": "gelu_new",
207
+ "diversity_penalty": 0.0,
208
+ "do_sample": false,
209
+ "dropout_rate": 0.1,
210
+ "early_stopping": false,
211
+ "encoder_no_repeat_ngram_size": 0,
212
+ "eos_token_id": 1,
213
+ "exponential_decay_length_penalty": null,
214
+ "feed_forward_proj": "gated-gelu",
215
+ "finetuning_task": null,
216
+ "forced_bos_token_id": null,
217
+ "forced_eos_token_id": null,
218
+ "id2label": {
219
+ "0": "LABEL_0",
220
+ "1": "LABEL_1"
221
+ },
222
+ "initializer_factor": 1.0,
223
+ "is_decoder": false,
224
+ "is_encoder_decoder": true,
225
+ "is_gated_act": true,
226
+ "label2id": {
227
+ "LABEL_0": 0,
228
+ "LABEL_1": 1
229
+ },
230
+ "layer_norm_epsilon": 1e-06,
231
+ "length_penalty": 1.0,
232
+ "max_length": 20,
233
+ "min_length": 0,
234
+ "model_type": "t5",
235
+ "n_positions": 512,
236
+ "no_repeat_ngram_size": 0,
237
+ "num_beam_groups": 1,
238
+ "num_beams": 1,
239
+ "num_decoder_layers": 24,
240
+ "num_heads": 16,
241
+ "num_layers": 24,
242
+ "num_return_sequences": 1,
243
+ "output_attentions": false,
244
+ "output_hidden_states": false,
245
+ "output_past": true,
246
+ "output_scores": false,
247
+ "pad_token_id": 0,
248
+ "prefix": null,
249
+ "problem_type": null,
250
+ "pruned_heads": {},
251
+ "relative_attention_max_distance": 128,
252
+ "relative_attention_num_buckets": 32,
253
+ "remove_invalid_values": false,
254
+ "repetition_penalty": 1.0,
255
+ "return_dict": true,
256
+ "return_dict_in_generate": false,
257
+ "sep_token_id": null,
258
+ "suppress_tokens": null,
259
+ "task_specific_params": null,
260
+ "temperature": 1.0,
261
+ "tf_legacy_loss": false,
262
+ "tie_encoder_decoder": false,
263
+ "tie_word_embeddings": false,
264
+ "tokenizer_class": null,
265
+ "top_k": 50,
266
+ "top_p": 1.0,
267
+ "torch_dtype": null,
268
+ "torchscript": false,
269
+ "typical_p": 1.0,
270
+ "use_bfloat16": false,
271
+ "use_cache": true,
272
+ "vocab_size": 32128
273
+ },
274
+ "torch_dtype": "float32",
275
+ "transformers_version": "4.46.1",
276
+ "vocab_size": 90714
277
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1025,
4
+ "decoder_start_token_id": 1025,
5
+ "do_sample": true,
6
+ "eos_token_id": 1024,
7
+ "max_length": 2610,
8
+ "pad_token_id": 1024,
9
+ "transformers_version": "4.46.1"
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19869c69b20b7fd90b35da4ec77fc6b1c54e4a56fc4ebf872d38b0621eae7a49
3
+ size 3751321772