Hugging Face's logo

nari-labs
/
Dia-1.6B-0626

Text-to-Speech
PyTorch
Safetensors
English
dia
model_hub_mixin
pytorch_model_hub_mixin
Model card Files
xet
Community
5
buttercrab AntonV HF Staff commited on
Commit
2dd3b31
·
verified ·
1 Parent(s): af8370f

hf-version (#1)

Browse files

- upload files (d89859e94175d34e3c274ded92413a925c4c9722)


Co-authored-by: Anton Vlasjuk <[email protected]>

Files changed (9) hide show
  1. audio_tokenizer_config.json +4 -0
  2. config.json +59 -47
  3. generation_config.json +12 -0
  4. model-00001-of-00002.safetensors +3 -0
  5. model-00002-of-00002.safetensors +3 -0
  6. model.safetensors.index.json +342 -0
  7. preprocessor_config.json +7 -4
  8. special_tokens_map.json +16 -0
  9. tokenizer_config.json +36 -0
audio_tokenizer_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "audio_tokenizer_class": "DacModel",
3
+ "audio_tokenizer_name_or_path": "descript/dac_44khz"
4
+ }
config.json CHANGED
@@ -1,50 +1,62 @@
1
  {
2
- "data": {
3
- "audio_bos_value": 1026,
4
- "audio_eos_value": 1024,
5
- "audio_length": 3072,
6
- "audio_pad_value": 1025,
7
- "channels": 9,
8
- "delay_pattern": [
9
- 0,
10
- 8,
11
- 9,
12
- 10,
13
- 11,
14
- 12,
15
- 13,
16
- 14,
17
- 15
18
- ],
19
- "text_length": 1024,
20
- "text_pad_value": 0
 
 
 
 
 
21
  },
22
- "model": {
23
- "decoder": {
24
- "cross_head_dim": 128,
25
- "cross_query_heads": 16,
26
- "gqa_head_dim": 128,
27
- "gqa_query_heads": 16,
28
- "kv_heads": 4,
29
- "n_embd": 2048,
30
- "n_hidden": 8192,
31
- "n_layer": 18
32
- },
33
- "dropout": 0.0,
34
- "encoder": {
35
- "head_dim": 128,
36
- "n_embd": 1024,
37
- "n_head": 16,
38
- "n_hidden": 4096,
39
- "n_layer": 12
40
- },
41
- "normalization_layer_epsilon": 1e-05,
42
- "rope_max_timescale": 10000,
43
- "rope_min_timescale": 1,
44
- "src_vocab_size": 256,
45
- "tgt_vocab_size": 1028,
46
- "weight_dtype": "float32"
 
47
  },
48
- "training": {},
49
- "version": "0.1"
50
- }
 
 
 
 
 
 
 
1
  {
2
+ "architectures": [
3
+ "DiaForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": 1026,
6
+ "decoder_config": {
7
+ "cross_head_dim": 128,
8
+ "cross_hidden_size": 1024,
9
+ "cross_num_attention_heads": 16,
10
+ "cross_num_key_value_heads": 16,
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 3072,
17
+ "model_type": "dia_decoder",
18
+ "norm_eps": 1e-05,
19
+ "num_attention_heads": 16,
20
+ "num_channels": 9,
21
+ "num_hidden_layers": 18,
22
+ "num_key_value_heads": 4,
23
+ "rope_scaling": null,
24
+ "rope_theta": 10000.0,
25
+ "vocab_size": 1028
26
  },
27
+ "delay_pattern": [
28
+ 0,
29
+ 8,
30
+ 9,
31
+ 10,
32
+ 11,
33
+ 12,
34
+ 13,
35
+ 14,
36
+ 15
37
+ ],
38
+ "encoder_config": {
39
+ "head_dim": 128,
40
+ "hidden_act": "silu",
41
+ "hidden_size": 1024,
42
+ "initializer_range": 0.02,
43
+ "intermediate_size": 4096,
44
+ "max_position_embeddings": 1024,
45
+ "model_type": "dia_encoder",
46
+ "norm_eps": 1e-05,
47
+ "num_attention_heads": 16,
48
+ "num_hidden_layers": 12,
49
+ "num_key_value_heads": 16,
50
+ "rope_scaling": null,
51
+ "rope_theta": 10000.0,
52
+ "vocab_size": 256
53
  },
54
+ "eos_token_id": 1024,
55
+ "initializer_range": 0.02,
56
+ "is_encoder_decoder": true,
57
+ "model_type": "dia",
58
+ "norm_eps": 1e-05,
59
+ "pad_token_id": 1025,
60
+ "torch_dtype": "float32",
61
+ "transformers_version": "4.53.0.dev0"
62
+ }
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1026,
3
+ "do_sample": true,
4
+ "eos_token_id": 1024,
5
+ "guidance_scale": 3.0,
6
+ "max_length": 3072,
7
+ "pad_token_id": 1025,
8
+ "temperature": 1.2,
9
+ "top_k": 45,
10
+ "top_p": 0.95,
11
+ "transformers_version": "4.53.0.dev0"
12
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cf8f82d9f87d408f5da9cc0d73a3e9c4d9ca3633a4df3be4d283277b7ff1b1a
3
+ size 4993046400
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54bf3a47ac13e28ba6193659d57bc63a46875685b211b071f782c32efc834343
3
+ size 1451637544
model.safetensors.index.json ADDED
@@ -0,0 +1,342 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 6444642304
4
+ },
5
+ "weight_map": {
6
+ "logits_dense.weight": "model-00002-of-00002.safetensors",
7
+ "model.decoder.embeddings.embed.weight": "model-00001-of-00002.safetensors",
8
+ "model.decoder.layers.0.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
9
+ "model.decoder.layers.0.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
10
+ "model.decoder.layers.0.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
11
+ "model.decoder.layers.0.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
12
+ "model.decoder.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
13
+ "model.decoder.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
14
+ "model.decoder.layers.0.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
15
+ "model.decoder.layers.0.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
16
+ "model.decoder.layers.0.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
17
+ "model.decoder.layers.0.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
18
+ "model.decoder.layers.0.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
19
+ "model.decoder.layers.0.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
20
+ "model.decoder.layers.0.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
21
+ "model.decoder.layers.1.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
22
+ "model.decoder.layers.1.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
23
+ "model.decoder.layers.1.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
24
+ "model.decoder.layers.1.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
25
+ "model.decoder.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
26
+ "model.decoder.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
27
+ "model.decoder.layers.1.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
28
+ "model.decoder.layers.1.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
29
+ "model.decoder.layers.1.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
30
+ "model.decoder.layers.1.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
31
+ "model.decoder.layers.1.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
32
+ "model.decoder.layers.1.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
33
+ "model.decoder.layers.1.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
34
+ "model.decoder.layers.10.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
35
+ "model.decoder.layers.10.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
36
+ "model.decoder.layers.10.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
37
+ "model.decoder.layers.10.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
38
+ "model.decoder.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
39
+ "model.decoder.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
40
+ "model.decoder.layers.10.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
41
+ "model.decoder.layers.10.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
42
+ "model.decoder.layers.10.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
43
+ "model.decoder.layers.10.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
44
+ "model.decoder.layers.10.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
45
+ "model.decoder.layers.10.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
46
+ "model.decoder.layers.10.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
47
+ "model.decoder.layers.11.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
48
+ "model.decoder.layers.11.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
49
+ "model.decoder.layers.11.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
50
+ "model.decoder.layers.11.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
51
+ "model.decoder.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
52
+ "model.decoder.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
53
+ "model.decoder.layers.11.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
54
+ "model.decoder.layers.11.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
55
+ "model.decoder.layers.11.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
56
+ "model.decoder.layers.11.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
57
+ "model.decoder.layers.11.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
58
+ "model.decoder.layers.11.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
59
+ "model.decoder.layers.11.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
60
+ "model.decoder.layers.12.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
61
+ "model.decoder.layers.12.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
62
+ "model.decoder.layers.12.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
63
+ "model.decoder.layers.12.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
64
+ "model.decoder.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
65
+ "model.decoder.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
66
+ "model.decoder.layers.12.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
67
+ "model.decoder.layers.12.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
68
+ "model.decoder.layers.12.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
69
+ "model.decoder.layers.12.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
70
+ "model.decoder.layers.12.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
71
+ "model.decoder.layers.12.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
72
+ "model.decoder.layers.12.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
73
+ "model.decoder.layers.13.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
74
+ "model.decoder.layers.13.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
75
+ "model.decoder.layers.13.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
76
+ "model.decoder.layers.13.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
77
+ "model.decoder.layers.13.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
78
+ "model.decoder.layers.13.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
79
+ "model.decoder.layers.13.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
80
+ "model.decoder.layers.13.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
81
+ "model.decoder.layers.13.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
82
+ "model.decoder.layers.13.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
83
+ "model.decoder.layers.13.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
84
+ "model.decoder.layers.13.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
85
+ "model.decoder.layers.13.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
86
+ "model.decoder.layers.14.cross_attention.k_proj.weight": "model-00002-of-00002.safetensors",
87
+ "model.decoder.layers.14.cross_attention.o_proj.weight": "model-00002-of-00002.safetensors",
88
+ "model.decoder.layers.14.cross_attention.q_proj.weight": "model-00002-of-00002.safetensors",
89
+ "model.decoder.layers.14.cross_attention.v_proj.weight": "model-00002-of-00002.safetensors",
90
+ "model.decoder.layers.14.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
91
+ "model.decoder.layers.14.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
92
+ "model.decoder.layers.14.pre_ca_norm.weight": "model-00002-of-00002.safetensors",
93
+ "model.decoder.layers.14.pre_mlp_norm.weight": "model-00002-of-00002.safetensors",
94
+ "model.decoder.layers.14.pre_sa_norm.weight": "model-00002-of-00002.safetensors",
95
+ "model.decoder.layers.14.self_attention.k_proj.weight": "model-00002-of-00002.safetensors",
96
+ "model.decoder.layers.14.self_attention.o_proj.weight": "model-00002-of-00002.safetensors",
97
+ "model.decoder.layers.14.self_attention.q_proj.weight": "model-00002-of-00002.safetensors",
98
+ "model.decoder.layers.14.self_attention.v_proj.weight": "model-00002-of-00002.safetensors",
99
+ "model.decoder.layers.15.cross_attention.k_proj.weight": "model-00002-of-00002.safetensors",
100
+ "model.decoder.layers.15.cross_attention.o_proj.weight": "model-00002-of-00002.safetensors",
101
+ "model.decoder.layers.15.cross_attention.q_proj.weight": "model-00002-of-00002.safetensors",
102
+ "model.decoder.layers.15.cross_attention.v_proj.weight": "model-00002-of-00002.safetensors",
103
+ "model.decoder.layers.15.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
104
+ "model.decoder.layers.15.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
105
+ "model.decoder.layers.15.pre_ca_norm.weight": "model-00002-of-00002.safetensors",
106
+ "model.decoder.layers.15.pre_mlp_norm.weight": "model-00002-of-00002.safetensors",
107
+ "model.decoder.layers.15.pre_sa_norm.weight": "model-00002-of-00002.safetensors",
108
+ "model.decoder.layers.15.self_attention.k_proj.weight": "model-00002-of-00002.safetensors",
109
+ "model.decoder.layers.15.self_attention.o_proj.weight": "model-00002-of-00002.safetensors",
110
+ "model.decoder.layers.15.self_attention.q_proj.weight": "model-00002-of-00002.safetensors",
111
+ "model.decoder.layers.15.self_attention.v_proj.weight": "model-00002-of-00002.safetensors",
112
+ "model.decoder.layers.16.cross_attention.k_proj.weight": "model-00002-of-00002.safetensors",
113
+ "model.decoder.layers.16.cross_attention.o_proj.weight": "model-00002-of-00002.safetensors",
114
+ "model.decoder.layers.16.cross_attention.q_proj.weight": "model-00002-of-00002.safetensors",
115
+ "model.decoder.layers.16.cross_attention.v_proj.weight": "model-00002-of-00002.safetensors",
116
+ "model.decoder.layers.16.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
117
+ "model.decoder.layers.16.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
118
+ "model.decoder.layers.16.pre_ca_norm.weight": "model-00002-of-00002.safetensors",
119
+ "model.decoder.layers.16.pre_mlp_norm.weight": "model-00002-of-00002.safetensors",
120
+ "model.decoder.layers.16.pre_sa_norm.weight": "model-00002-of-00002.safetensors",
121
+ "model.decoder.layers.16.self_attention.k_proj.weight": "model-00002-of-00002.safetensors",
122
+ "model.decoder.layers.16.self_attention.o_proj.weight": "model-00002-of-00002.safetensors",
123
+ "model.decoder.layers.16.self_attention.q_proj.weight": "model-00002-of-00002.safetensors",
124
+ "model.decoder.layers.16.self_attention.v_proj.weight": "model-00002-of-00002.safetensors",
125
+ "model.decoder.layers.17.cross_attention.k_proj.weight": "model-00002-of-00002.safetensors",
126
+ "model.decoder.layers.17.cross_attention.o_proj.weight": "model-00002-of-00002.safetensors",
127
+ "model.decoder.layers.17.cross_attention.q_proj.weight": "model-00002-of-00002.safetensors",
128
+ "model.decoder.layers.17.cross_attention.v_proj.weight": "model-00002-of-00002.safetensors",
129
+ "model.decoder.layers.17.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
130
+ "model.decoder.layers.17.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
131
+ "model.decoder.layers.17.pre_ca_norm.weight": "model-00002-of-00002.safetensors",
132
+ "model.decoder.layers.17.pre_mlp_norm.weight": "model-00002-of-00002.safetensors",
133
+ "model.decoder.layers.17.pre_sa_norm.weight": "model-00002-of-00002.safetensors",
134
+ "model.decoder.layers.17.self_attention.k_proj.weight": "model-00002-of-00002.safetensors",
135
+ "model.decoder.layers.17.self_attention.o_proj.weight": "model-00002-of-00002.safetensors",
136
+ "model.decoder.layers.17.self_attention.q_proj.weight": "model-00002-of-00002.safetensors",
137
+ "model.decoder.layers.17.self_attention.v_proj.weight": "model-00002-of-00002.safetensors",
138
+ "model.decoder.layers.2.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
139
+ "model.decoder.layers.2.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
140
+ "model.decoder.layers.2.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
141
+ "model.decoder.layers.2.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
142
+ "model.decoder.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
143
+ "model.decoder.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
144
+ "model.decoder.layers.2.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
145
+ "model.decoder.layers.2.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
146
+ "model.decoder.layers.2.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
147
+ "model.decoder.layers.2.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
148
+ "model.decoder.layers.2.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
149
+ "model.decoder.layers.2.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
150
+ "model.decoder.layers.2.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
151
+ "model.decoder.layers.3.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
152
+ "model.decoder.layers.3.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
153
+ "model.decoder.layers.3.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
154
+ "model.decoder.layers.3.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
155
+ "model.decoder.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
156
+ "model.decoder.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
157
+ "model.decoder.layers.3.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
158
+ "model.decoder.layers.3.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
159
+ "model.decoder.layers.3.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
160
+ "model.decoder.layers.3.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
161
+ "model.decoder.layers.3.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
162
+ "model.decoder.layers.3.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
163
+ "model.decoder.layers.3.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
164
+ "model.decoder.layers.4.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
165
+ "model.decoder.layers.4.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
166
+ "model.decoder.layers.4.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
167
+ "model.decoder.layers.4.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
168
+ "model.decoder.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
169
+ "model.decoder.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
170
+ "model.decoder.layers.4.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
171
+ "model.decoder.layers.4.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
172
+ "model.decoder.layers.4.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
173
+ "model.decoder.layers.4.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
174
+ "model.decoder.layers.4.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
175
+ "model.decoder.layers.4.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
176
+ "model.decoder.layers.4.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
177
+ "model.decoder.layers.5.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
178
+ "model.decoder.layers.5.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
179
+ "model.decoder.layers.5.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
180
+ "model.decoder.layers.5.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
181
+ "model.decoder.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
182
+ "model.decoder.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
183
+ "model.decoder.layers.5.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
184
+ "model.decoder.layers.5.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
185
+ "model.decoder.layers.5.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
186
+ "model.decoder.layers.5.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
187
+ "model.decoder.layers.5.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
188
+ "model.decoder.layers.5.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
189
+ "model.decoder.layers.5.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
190
+ "model.decoder.layers.6.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
191
+ "model.decoder.layers.6.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
192
+ "model.decoder.layers.6.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
193
+ "model.decoder.layers.6.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
194
+ "model.decoder.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
195
+ "model.decoder.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
196
+ "model.decoder.layers.6.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
197
+ "model.decoder.layers.6.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
198
+ "model.decoder.layers.6.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
199
+ "model.decoder.layers.6.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
200
+ "model.decoder.layers.6.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
201
+ "model.decoder.layers.6.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
202
+ "model.decoder.layers.6.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
203
+ "model.decoder.layers.7.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
204
+ "model.decoder.layers.7.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
205
+ "model.decoder.layers.7.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
206
+ "model.decoder.layers.7.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
207
+ "model.decoder.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
208
+ "model.decoder.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
209
+ "model.decoder.layers.7.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
210
+ "model.decoder.layers.7.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
211
+ "model.decoder.layers.7.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
212
+ "model.decoder.layers.7.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
213
+ "model.decoder.layers.7.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
214
+ "model.decoder.layers.7.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
215
+ "model.decoder.layers.7.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
216
+ "model.decoder.layers.8.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
217
+ "model.decoder.layers.8.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
218
+ "model.decoder.layers.8.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
219
+ "model.decoder.layers.8.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
220
+ "model.decoder.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
221
+ "model.decoder.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
222
+ "model.decoder.layers.8.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
223
+ "model.decoder.layers.8.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
224
+ "model.decoder.layers.8.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
225
+ "model.decoder.layers.8.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
226
+ "model.decoder.layers.8.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
227
+ "model.decoder.layers.8.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
228
+ "model.decoder.layers.8.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
229
+ "model.decoder.layers.9.cross_attention.k_proj.weight": "model-00001-of-00002.safetensors",
230
+ "model.decoder.layers.9.cross_attention.o_proj.weight": "model-00001-of-00002.safetensors",
231
+ "model.decoder.layers.9.cross_attention.q_proj.weight": "model-00001-of-00002.safetensors",
232
+ "model.decoder.layers.9.cross_attention.v_proj.weight": "model-00001-of-00002.safetensors",
233
+ "model.decoder.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
234
+ "model.decoder.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
235
+ "model.decoder.layers.9.pre_ca_norm.weight": "model-00001-of-00002.safetensors",
236
+ "model.decoder.layers.9.pre_mlp_norm.weight": "model-00001-of-00002.safetensors",
237
+ "model.decoder.layers.9.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
238
+ "model.decoder.layers.9.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
239
+ "model.decoder.layers.9.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
240
+ "model.decoder.layers.9.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
241
+ "model.decoder.layers.9.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
242
+ "model.decoder.norm.weight": "model-00002-of-00002.safetensors",
243
+ "model.encoder.embedding.weight": "model-00001-of-00002.safetensors",
244
+ "model.encoder.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
245
+ "model.encoder.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
246
+ "model.encoder.layers.0.post_sa_norm.weight": "model-00001-of-00002.safetensors",
247
+ "model.encoder.layers.0.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
248
+ "model.encoder.layers.0.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
249
+ "model.encoder.layers.0.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
250
+ "model.encoder.layers.0.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
251
+ "model.encoder.layers.0.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
252
+ "model.encoder.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
253
+ "model.encoder.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
254
+ "model.encoder.layers.1.post_sa_norm.weight": "model-00001-of-00002.safetensors",
255
+ "model.encoder.layers.1.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
256
+ "model.encoder.layers.1.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
257
+ "model.encoder.layers.1.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
258
+ "model.encoder.layers.1.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
259
+ "model.encoder.layers.1.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
260
+ "model.encoder.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
261
+ "model.encoder.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
262
+ "model.encoder.layers.10.post_sa_norm.weight": "model-00001-of-00002.safetensors",
263
+ "model.encoder.layers.10.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
264
+ "model.encoder.layers.10.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
265
+ "model.encoder.layers.10.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
266
+ "model.encoder.layers.10.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
267
+ "model.encoder.layers.10.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
268
+ "model.encoder.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
269
+ "model.encoder.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
270
+ "model.encoder.layers.11.post_sa_norm.weight": "model-00001-of-00002.safetensors",
271
+ "model.encoder.layers.11.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
272
+ "model.encoder.layers.11.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
273
+ "model.encoder.layers.11.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
274
+ "model.encoder.layers.11.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
275
+ "model.encoder.layers.11.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
276
+ "model.encoder.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
277
+ "model.encoder.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
278
+ "model.encoder.layers.2.post_sa_norm.weight": "model-00001-of-00002.safetensors",
279
+ "model.encoder.layers.2.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
280
+ "model.encoder.layers.2.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
281
+ "model.encoder.layers.2.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
282
+ "model.encoder.layers.2.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
283
+ "model.encoder.layers.2.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
284
+ "model.encoder.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
285
+ "model.encoder.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
286
+ "model.encoder.layers.3.post_sa_norm.weight": "model-00001-of-00002.safetensors",
287
+ "model.encoder.layers.3.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
288
+ "model.encoder.layers.3.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
289
+ "model.encoder.layers.3.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
290
+ "model.encoder.layers.3.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
291
+ "model.encoder.layers.3.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
292
+ "model.encoder.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
293
+ "model.encoder.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
294
+ "model.encoder.layers.4.post_sa_norm.weight": "model-00001-of-00002.safetensors",
295
+ "model.encoder.layers.4.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
296
+ "model.encoder.layers.4.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
297
+ "model.encoder.layers.4.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
298
+ "model.encoder.layers.4.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
299
+ "model.encoder.layers.4.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
300
+ "model.encoder.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
301
+ "model.encoder.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
302
+ "model.encoder.layers.5.post_sa_norm.weight": "model-00001-of-00002.safetensors",
303
+ "model.encoder.layers.5.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
304
+ "model.encoder.layers.5.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
305
+ "model.encoder.layers.5.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
306
+ "model.encoder.layers.5.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
307
+ "model.encoder.layers.5.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
308
+ "model.encoder.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
309
+ "model.encoder.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
310
+ "model.encoder.layers.6.post_sa_norm.weight": "model-00001-of-00002.safetensors",
311
+ "model.encoder.layers.6.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
312
+ "model.encoder.layers.6.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
313
+ "model.encoder.layers.6.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
314
+ "model.encoder.layers.6.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
315
+ "model.encoder.layers.6.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
316
+ "model.encoder.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
317
+ "model.encoder.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
318
+ "model.encoder.layers.7.post_sa_norm.weight": "model-00001-of-00002.safetensors",
319
+ "model.encoder.layers.7.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
320
+ "model.encoder.layers.7.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
321
+ "model.encoder.layers.7.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
322
+ "model.encoder.layers.7.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
323
+ "model.encoder.layers.7.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
324
+ "model.encoder.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
325
+ "model.encoder.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
326
+ "model.encoder.layers.8.post_sa_norm.weight": "model-00001-of-00002.safetensors",
327
+ "model.encoder.layers.8.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
328
+ "model.encoder.layers.8.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
329
+ "model.encoder.layers.8.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
330
+ "model.encoder.layers.8.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
331
+ "model.encoder.layers.8.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
332
+ "model.encoder.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
333
+ "model.encoder.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
334
+ "model.encoder.layers.9.post_sa_norm.weight": "model-00001-of-00002.safetensors",
335
+ "model.encoder.layers.9.pre_sa_norm.weight": "model-00001-of-00002.safetensors",
336
+ "model.encoder.layers.9.self_attention.k_proj.weight": "model-00001-of-00002.safetensors",
337
+ "model.encoder.layers.9.self_attention.o_proj.weight": "model-00001-of-00002.safetensors",
338
+ "model.encoder.layers.9.self_attention.q_proj.weight": "model-00001-of-00002.safetensors",
339
+ "model.encoder.layers.9.self_attention.v_proj.weight": "model-00001-of-00002.safetensors",
340
+ "model.encoder.norm.weight": "model-00001-of-00002.safetensors"
341
+ }
342
+ }
preprocessor_config.json CHANGED
@@ -1,7 +1,10 @@
1
  {
2
- "feature_extractor_type": "DacFeatureExtractor",
 
 
 
 
3
  "processor_class": "DiaProcessor",
4
- "audio_eos_value": 1024,
5
- "audio_pad_value": 1025,
6
- "audio_bos_value": 1026
7
  }
 
1
  {
2
+ "feature_extractor_type": "DiaFeatureExtractor",
3
+ "feature_size": 1,
4
+ "hop_length": 512,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
  "processor_class": "DiaProcessor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 44100
 
10
  }
special_tokens_map.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "pad_token": {
3
+ "content": "<pad>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "unk_token": {
10
+ "content": "<pad>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ }
16
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<pad>",
5
+ "lstrip": false,
6
+ "normalized": true,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[S1]",
13
+ "lstrip": false,
14
+ "normalized": true,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "2": {
20
+ "content": "[S2]",
21
+ "lstrip": false,
22
+ "normalized": true,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": false
26
+ }
27
+ },
28
+ "clean_up_tokenization_spaces": false,
29
+ "extra_special_tokens": {},
30
+ "max_length": 1024,
31
+ "model_max_length": 1000000000000000019884624838656,
32
+ "pad_token": "<pad>",
33
+ "processor_class": "DiaProcessor",
34
+ "tokenizer_class": "DiaTokenizer",
35
+ "unk_token": "<pad>"
36
+ }