Hugging Face's logo

facebook
/
hf-seamless-m4t-large

Text-to-Speech
Transformers
PyTorch
seamless_m4t
feature-extraction
SeamlessM4T
Model card Files Community
19

Upload model

#2
by ylacombe HF staff - opened
base: refs/heads/main
←
from: refs/pr/2
Discussion Files changed
+305
-0
Files changed (3) hide show
  1. config.json +116 -0
  2. generation_config.json +186 -0
  3. pytorch_model.bin +3 -0
config.json ADDED
@@ -0,0 +1,116 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "relu",
4
+ "adaptor_dropout": 0.1,
5
+ "adaptor_kernel_size": 8,
6
+ "adaptor_stride": 8,
7
+ "add_adapter": true,
8
+ "architectures": [
9
+ "SeamlessM4TModel"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 2,
13
+ "control_symbol_vocoder_offset": 4,
14
+ "conv_depthwise_kernel_size": 31,
15
+ "decoder_attention_heads": 16,
16
+ "decoder_ffn_dim": 8192,
17
+ "decoder_layerdrop": 0.05,
18
+ "decoder_layers": 24,
19
+ "decoder_start_token_id": 3,
20
+ "dropout": 0.1,
21
+ "encoder_attention_heads": 16,
22
+ "encoder_ffn_dim": 8192,
23
+ "encoder_layerdrop": 0.05,
24
+ "encoder_layers": 24,
25
+ "eos_token_id": 3,
26
+ "feature_projection_input_dim": 160,
27
+ "hidden_size": 1024,
28
+ "initializer_range": 0.02,
29
+ "is_encoder_decoder": true,
30
+ "lang_embed_dim": 256,
31
+ "layer_norm_eps": 1e-05,
32
+ "leaky_relu_slope": 0.1,
33
+ "max_new_tokens": 256,
34
+ "max_position_embeddings": 1024,
35
+ "max_source_positions": 4096,
36
+ "model_type": "seamless_m4t",
37
+ "num_adapter_layers": 1,
38
+ "num_attention_heads": 16,
39
+ "num_conv_pos_embedding_groups": 16,
40
+ "num_conv_pos_embeddings": 128,
41
+ "num_hidden_layers": 24,
42
+ "pad_token_id": 0,
43
+ "position_embeddings_type": "relative",
44
+ "resblock_dilation_sizes": [
45
+ [
46
+ 1,
47
+ 3,
48
+ 5
49
+ ],
50
+ [
51
+ 1,
52
+ 3,
53
+ 5
54
+ ],
55
+ [
56
+ 1,
57
+ 3,
58
+ 5
59
+ ]
60
+ ],
61
+ "resblock_kernel_sizes": [
62
+ 3,
63
+ 7,
64
+ 11
65
+ ],
66
+ "rotary_embedding_base": 10000,
67
+ "sampling_rate": 16000,
68
+ "scale_embedding": true,
69
+ "speech_encoder_attention_heads": 16,
70
+ "speech_encoder_dropout": 0.0,
71
+ "speech_encoder_hidden_act": "swish",
72
+ "speech_encoder_intermediate_size": 4096,
73
+ "speech_encoder_layerdrop": 0.1,
74
+ "speech_encoder_layers": 24,
75
+ "spkr_embed_dim": 256,
76
+ "t2u_bos_token_id": 0,
77
+ "t2u_decoder_attention_heads": 16,
78
+ "t2u_decoder_ffn_dim": 8192,
79
+ "t2u_decoder_layers": 6,
80
+ "t2u_decoder_start_token_id": 2,
81
+ "t2u_encoder_attention_heads": 16,
82
+ "t2u_encoder_ffn_dim": 8192,
83
+ "t2u_encoder_layers": 6,
84
+ "t2u_eos_token_id": 2,
85
+ "t2u_max_new_tokens": 1024,
86
+ "t2u_max_position_embeddings": 2048,
87
+ "t2u_num_langs": 38,
88
+ "t2u_offset_tgt_lang": 10005,
89
+ "t2u_pad_token_id": 1,
90
+ "torch_dtype": "float32",
91
+ "transformers_version": "4.33.0.dev0",
92
+ "unit_embed_dim": 1280,
93
+ "unit_hifi_gan_vocab_size": 10000,
94
+ "unit_vocab_size": 10082,
95
+ "upsample_initial_channel": 512,
96
+ "upsample_kernel_sizes": [
97
+ 11,
98
+ 8,
99
+ 8,
100
+ 4,
101
+ 4
102
+ ],
103
+ "upsample_rates": [
104
+ 5,
105
+ 4,
106
+ 4,
107
+ 2,
108
+ 2
109
+ ],
110
+ "use_cache": true,
111
+ "var_pred_dropout": 0.5,
112
+ "variance_predictor_kernel_size": 3,
113
+ "vocab_size": 256102,
114
+ "vocoder_num_langs": 36,
115
+ "vocoder_num_spkrs": 200
116
+ }
generation_config.json ADDED
@@ -0,0 +1,186 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "decoder_start_token_id": 3,
4
+ "eos_token_id": 3,
5
+ "max_new_tokens": 256,
6
+ "pad_token_id": 0,
7
+ "t2u_lang_code_to_id": {
8
+ "arb": 0,
9
+ "ben": 1,
10
+ "cat": 2,
11
+ "ces": 3,
12
+ "cmn": 4,
13
+ "cym": 5,
14
+ "dan": 6,
15
+ "deu": 7,
16
+ "eng": 8,
17
+ "est": 9,
18
+ "fin": 10,
19
+ "fra": 11,
20
+ "hin": 12,
21
+ "ind": 13,
22
+ "ita": 14,
23
+ "jpn": 15,
24
+ "kan": 16,
25
+ "kor": 17,
26
+ "mlt": 18,
27
+ "nld": 19,
28
+ "pes": 20,
29
+ "pol": 21,
30
+ "por": 22,
31
+ "ron": 23,
32
+ "rus": 24,
33
+ "slk": 25,
34
+ "spa": 26,
35
+ "swe": 27,
36
+ "swh": 28,
37
+ "tam": 29,
38
+ "tel": 30,
39
+ "tgl": 31,
40
+ "tha": 32,
41
+ "tur": 33,
42
+ "ukr": 34,
43
+ "urd": 35,
44
+ "uzn": 36,
45
+ "vie": 37
46
+ },
47
+ "text_decoder_lang_to_code_id": {
48
+ "afr": 256001,
49
+ "amh": 256002,
50
+ "arb": 256003,
51
+ "ary": 256004,
52
+ "arz": 256005,
53
+ "asm": 256006,
54
+ "azj": 256007,
55
+ "bel": 256008,
56
+ "ben": 256009,
57
+ "bos": 256010,
58
+ "bul": 256011,
59
+ "cat": 256012,
60
+ "ceb": 256013,
61
+ "ces": 256014,
62
+ "ckb": 256015,
63
+ "cmn": 256016,
64
+ "cmn_Hant": 256017,
65
+ "cym": 256018,
66
+ "dan": 256019,
67
+ "deu": 256020,
68
+ "ell": 256021,
69
+ "eng": 256022,
70
+ "est": 256023,
71
+ "eus": 256024,
72
+ "fin": 256025,
73
+ "fra": 256026,
74
+ "fuv": 256027,
75
+ "gaz": 256028,
76
+ "gle": 256029,
77
+ "glg": 256030,
78
+ "guj": 256031,
79
+ "heb": 256032,
80
+ "hin": 256033,
81
+ "hrv": 256034,
82
+ "hun": 256035,
83
+ "hye": 256036,
84
+ "ibo": 256037,
85
+ "ind": 256038,
86
+ "isl": 256039,
87
+ "ita": 256040,
88
+ "jav": 256041,
89
+ "jpn": 256042,
90
+ "kan": 256043,
91
+ "kat": 256044,
92
+ "kaz": 256045,
93
+ "khk": 256046,
94
+ "khm": 256047,
95
+ "kir": 256048,
96
+ "kor": 256049,
97
+ "lao": 256050,
98
+ "lit": 256051,
99
+ "lug": 256052,
100
+ "luo": 256053,
101
+ "lvs": 256054,
102
+ "mai": 256055,
103
+ "mal": 256056,
104
+ "mar": 256057,
105
+ "mkd": 256058,
106
+ "mlt": 256059,
107
+ "mni": 256060,
108
+ "mya": 256061,
109
+ "nld": 256062,
110
+ "nno": 256063,
111
+ "nob": 256064,
112
+ "npi": 256065,
113
+ "nya": 256066,
114
+ "ory": 256067,
115
+ "pan": 256068,
116
+ "pbt": 256069,
117
+ "pes": 256070,
118
+ "pol": 256071,
119
+ "por": 256072,
120
+ "ron": 256073,
121
+ "rus": 256074,
122
+ "sat": 256075,
123
+ "slk": 256076,
124
+ "slv": 256077,
125
+ "sna": 256078,
126
+ "snd": 256079,
127
+ "som": 256080,
128
+ "spa": 256081,
129
+ "srp": 256082,
130
+ "swe": 256083,
131
+ "swh": 256084,
132
+ "tam": 256085,
133
+ "tel": 256086,
134
+ "tgk": 256087,
135
+ "tgl": 256088,
136
+ "tha": 256089,
137
+ "tur": 256090,
138
+ "ukr": 256091,
139
+ "urd": 256092,
140
+ "uzn": 256093,
141
+ "vie": 256094,
142
+ "yor": 256095,
143
+ "yue": 256096,
144
+ "zlm": 256097,
145
+ "zul": 256098
146
+ },
147
+ "transformers_version": "4.33.0.dev0",
148
+ "vocoder_lang_code_to_id": {
149
+ "arb": 0,
150
+ "ben": 1,
151
+ "cat": 2,
152
+ "ces": 3,
153
+ "cmn": 4,
154
+ "cym": 5,
155
+ "dan": 6,
156
+ "deu": 7,
157
+ "eng": 8,
158
+ "est": 9,
159
+ "fin": 10,
160
+ "fra": 11,
161
+ "hin": 12,
162
+ "ind": 13,
163
+ "ita": 14,
164
+ "jpn": 15,
165
+ "kor": 16,
166
+ "mlt": 17,
167
+ "nld": 18,
168
+ "pes": 19,
169
+ "pol": 20,
170
+ "por": 21,
171
+ "ron": 22,
172
+ "rus": 23,
173
+ "slk": 24,
174
+ "spa": 25,
175
+ "swe": 26,
176
+ "swh": 27,
177
+ "tel": 28,
178
+ "tgl": 29,
179
+ "tha": 30,
180
+ "tur": 31,
181
+ "ukr": 32,
182
+ "urd": 33,
183
+ "uzn": 34,
184
+ "vie": 35
185
+ }
186
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dfc9e4eda13d6ae0f444d4dde2678002531ddbc9d40701b6032ce0e3e6f43bad
3
+ size 9440981560