Hugging Face's logo

mlx-community
/
tada-tts

Text-to-Speech
Safetensors
MLX
English
mlx-audio
tada
tts
speech-synthesis
apple-silicon
speech
speech generation
voice cloning
Model card Files
xet
 Community
roboalchemist commited on
Commit
75351b5
·
verified ·
1 Parent(s): e1bfe37

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. README.md +54 -0
  2. aligner/weights.safetensors +3 -0
  3. config.json +49 -0
  4. decoder/weights.safetensors +3 -0
  5. encoder/weights.safetensors +3 -0
  6. model.safetensors +3 -0
  7. model.safetensors.index.json +360 -0
  8. model/config.json +49 -0
  9. model/weights.safetensors +3 -0
README.md ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: llama3.2
3
+ library_name: mlx-audio
4
+ language:
5
+ - en
6
+ tags:
7
+ - mlx
8
+ - tts
9
+ - text-to-speech
10
+ - speech-synthesis
11
+ - tada
12
+ - apple-silicon
13
+ - mlx
14
+ - text-to-speech
15
+ - speech
16
+ - speech generation
17
+ - voice cloning
18
+ - tts
19
+ - mlx-audio
20
+ pipeline_tag: text-to-speech
21
+ base_model: meta-llama/Llama-3.2-1B
22
+ arxiv: 2602.23068
23
+ ---
24
+ # mlx-community/tada-tts
25
+
26
+ This model was converted to MLX format from [`HumeAI/mlx-tada-1b`](https://huggingface.co/HumeAI/mlx-tada-1b) using mlx-audio version **0.2.8**.
27
+
28
+ Refer to the [original model card](https://huggingface.co/HumeAI/mlx-tada-1b) for more details on the model.
29
+
30
+ ## Use with mlx-audio
31
+
32
+ ```bash
33
+ pip install -U mlx-audio
34
+ ```
35
+
36
+ ### CLI Example:
37
+ ```bash
38
+ python -m mlx_audio.tts.generate --model mlx-community/tada-tts --text "Hello, this is a test."
39
+ ```
40
+
41
+ ### Python Example:
42
+ ```python
43
+ from mlx_audio.tts.utils import load_model
44
+ from mlx_audio.tts.generate import generate_audio
45
+
46
+ model = load_model("mlx-community/tada-tts")
47
+ generate_audio(
48
+ model=model,
49
+ text="Hello, this is a test.",
50
+ ref_audio="path_to_audio.wav",
51
+ file_prefix="test_audio",
52
+ )
53
+
54
+ ```
aligner/weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af2e603bd1f76bf33dbaf0ebe1d65f7024641d28e2887c703eadb7a3cda1316e
3
+ size 893830649
config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "acoustic_dim": 512,
3
+ "acoustic_from_nth_hidden_state": -1,
4
+ "acoustic_mean": 0.0,
5
+ "acoustic_std": 1.5,
6
+ "add_semantic_to_condition": 0.0,
7
+ "architectures": [
8
+ "TadaForCausalLM"
9
+ ],
10
+ "attention_bias": false,
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 128000,
13
+ "bottleneck_dim": null,
14
+ "context_window": 8,
15
+ "diffusion_head_type": "vibevoice",
16
+ "dist_type": "fixed",
17
+ "dtype": "bfloat16",
18
+ "eos_token_id": 128001,
19
+ "head_dim": 64,
20
+ "head_ffn_ratio": 4.0,
21
+ "head_layers": 6,
22
+ "hidden_act": "silu",
23
+ "hidden_size": 2048,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 8192,
26
+ "latent_dropout": 0.0,
27
+ "max_position_embeddings": 131072,
28
+ "mlp_bias": false,
29
+ "model_type": "tada",
30
+ "num_attention_heads": 32,
31
+ "num_hidden_layers": 16,
32
+ "num_key_value_heads": 8,
33
+ "num_time_classes": 256,
34
+ "pretraining_tp": 1,
35
+ "rms_norm_eps": 1e-05,
36
+ "rope_scaling": {
37
+ "factor": 32.0,
38
+ "high_freq_factor": 4.0,
39
+ "low_freq_factor": 1.0,
40
+ "original_max_position_embeddings": 8192,
41
+ "rope_type": "llama3"
42
+ },
43
+ "rope_theta": 500000.0,
44
+ "shift_acoustic": 5,
45
+ "tie_word_embeddings": true,
46
+ "transformers_version": "4.57.3",
47
+ "use_cache": true,
48
+ "vocab_size": 128256
49
+ }
decoder/weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40310d1e93460f2bea9b77b83dfafe11a5ffbf5dc36224b4ca89db20c1776fcb
3
+ size 237407562
encoder/weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5732c5a73f42475f620a6a8dba36f404e58fc3b4bae1f1766503a1448e062970
3
+ size 186606332
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b52f59e40505f5c8bde0de70bcf0a8164102c6c5288a636bac6af75f90ba8176
3
+ size 3592372707
model.safetensors.index.json ADDED
@@ -0,0 +1,360 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 3592330628,
4
+ "total_parameters": 1753575777
5
+ },
6
+ "weight_map": {
7
+ "acoustic_mask_emb.weight": "model.safetensors",
8
+ "acoustic_proj.bias": "model.safetensors",
9
+ "acoustic_proj.weight": "model.safetensors",
10
+ "decoder.decoder_proj.bias": "model.safetensors",
11
+ "decoder.decoder_proj.weight": "model.safetensors",
12
+ "decoder.local_attention_decoder.final_norm.bias": "model.safetensors",
13
+ "decoder.local_attention_decoder.final_norm.weight": "model.safetensors",
14
+ "decoder.local_attention_decoder.layers.0.linear1.bias": "model.safetensors",
15
+ "decoder.local_attention_decoder.layers.0.linear1.weight": "model.safetensors",
16
+ "decoder.local_attention_decoder.layers.0.linear2.bias": "model.safetensors",
17
+ "decoder.local_attention_decoder.layers.0.linear2.weight": "model.safetensors",
18
+ "decoder.local_attention_decoder.layers.0.norm.bias": "model.safetensors",
19
+ "decoder.local_attention_decoder.layers.0.norm.weight": "model.safetensors",
20
+ "decoder.local_attention_decoder.layers.0.self_attn.layer_norm.bias": "model.safetensors",
21
+ "decoder.local_attention_decoder.layers.0.self_attn.layer_norm.weight": "model.safetensors",
22
+ "decoder.local_attention_decoder.layers.0.self_attn.out_proj.bias": "model.safetensors",
23
+ "decoder.local_attention_decoder.layers.0.self_attn.out_proj.weight": "model.safetensors",
24
+ "decoder.local_attention_decoder.layers.0.self_attn.qkv.bias": "model.safetensors",
25
+ "decoder.local_attention_decoder.layers.0.self_attn.qkv.weight": "model.safetensors",
26
+ "decoder.local_attention_decoder.layers.1.linear1.bias": "model.safetensors",
27
+ "decoder.local_attention_decoder.layers.1.linear1.weight": "model.safetensors",
28
+ "decoder.local_attention_decoder.layers.1.linear2.bias": "model.safetensors",
29
+ "decoder.local_attention_decoder.layers.1.linear2.weight": "model.safetensors",
30
+ "decoder.local_attention_decoder.layers.1.norm.bias": "model.safetensors",
31
+ "decoder.local_attention_decoder.layers.1.norm.weight": "model.safetensors",
32
+ "decoder.local_attention_decoder.layers.1.self_attn.layer_norm.bias": "model.safetensors",
33
+ "decoder.local_attention_decoder.layers.1.self_attn.layer_norm.weight": "model.safetensors",
34
+ "decoder.local_attention_decoder.layers.1.self_attn.out_proj.bias": "model.safetensors",
35
+ "decoder.local_attention_decoder.layers.1.self_attn.out_proj.weight": "model.safetensors",
36
+ "decoder.local_attention_decoder.layers.1.self_attn.qkv.bias": "model.safetensors",
37
+ "decoder.local_attention_decoder.layers.1.self_attn.qkv.weight": "model.safetensors",
38
+ "decoder.local_attention_decoder.layers.2.linear1.bias": "model.safetensors",
39
+ "decoder.local_attention_decoder.layers.2.linear1.weight": "model.safetensors",
40
+ "decoder.local_attention_decoder.layers.2.linear2.bias": "model.safetensors",
41
+ "decoder.local_attention_decoder.layers.2.linear2.weight": "model.safetensors",
42
+ "decoder.local_attention_decoder.layers.2.norm.bias": "model.safetensors",
43
+ "decoder.local_attention_decoder.layers.2.norm.weight": "model.safetensors",
44
+ "decoder.local_attention_decoder.layers.2.self_attn.layer_norm.bias": "model.safetensors",
45
+ "decoder.local_attention_decoder.layers.2.self_attn.layer_norm.weight": "model.safetensors",
46
+ "decoder.local_attention_decoder.layers.2.self_attn.out_proj.bias": "model.safetensors",
47
+ "decoder.local_attention_decoder.layers.2.self_attn.out_proj.weight": "model.safetensors",
48
+ "decoder.local_attention_decoder.layers.2.self_attn.qkv.bias": "model.safetensors",
49
+ "decoder.local_attention_decoder.layers.2.self_attn.qkv.weight": "model.safetensors",
50
+ "decoder.local_attention_decoder.layers.3.linear1.bias": "model.safetensors",
51
+ "decoder.local_attention_decoder.layers.3.linear1.weight": "model.safetensors",
52
+ "decoder.local_attention_decoder.layers.3.linear2.bias": "model.safetensors",
53
+ "decoder.local_attention_decoder.layers.3.linear2.weight": "model.safetensors",
54
+ "decoder.local_attention_decoder.layers.3.norm.bias": "model.safetensors",
55
+ "decoder.local_attention_decoder.layers.3.norm.weight": "model.safetensors",
56
+ "decoder.local_attention_decoder.layers.3.self_attn.layer_norm.bias": "model.safetensors",
57
+ "decoder.local_attention_decoder.layers.3.self_attn.layer_norm.weight": "model.safetensors",
58
+ "decoder.local_attention_decoder.layers.3.self_attn.out_proj.bias": "model.safetensors",
59
+ "decoder.local_attention_decoder.layers.3.self_attn.out_proj.weight": "model.safetensors",
60
+ "decoder.local_attention_decoder.layers.3.self_attn.qkv.bias": "model.safetensors",
61
+ "decoder.local_attention_decoder.layers.3.self_attn.qkv.weight": "model.safetensors",
62
+ "decoder.local_attention_decoder.layers.4.linear1.bias": "model.safetensors",
63
+ "decoder.local_attention_decoder.layers.4.linear1.weight": "model.safetensors",
64
+ "decoder.local_attention_decoder.layers.4.linear2.bias": "model.safetensors",
65
+ "decoder.local_attention_decoder.layers.4.linear2.weight": "model.safetensors",
66
+ "decoder.local_attention_decoder.layers.4.norm.bias": "model.safetensors",
67
+ "decoder.local_attention_decoder.layers.4.norm.weight": "model.safetensors",
68
+ "decoder.local_attention_decoder.layers.4.self_attn.layer_norm.bias": "model.safetensors",
69
+ "decoder.local_attention_decoder.layers.4.self_attn.layer_norm.weight": "model.safetensors",
70
+ "decoder.local_attention_decoder.layers.4.self_attn.out_proj.bias": "model.safetensors",
71
+ "decoder.local_attention_decoder.layers.4.self_attn.out_proj.weight": "model.safetensors",
72
+ "decoder.local_attention_decoder.layers.4.self_attn.qkv.bias": "model.safetensors",
73
+ "decoder.local_attention_decoder.layers.4.self_attn.qkv.weight": "model.safetensors",
74
+ "decoder.local_attention_decoder.layers.5.linear1.bias": "model.safetensors",
75
+ "decoder.local_attention_decoder.layers.5.linear1.weight": "model.safetensors",
76
+ "decoder.local_attention_decoder.layers.5.linear2.bias": "model.safetensors",
77
+ "decoder.local_attention_decoder.layers.5.linear2.weight": "model.safetensors",
78
+ "decoder.local_attention_decoder.layers.5.norm.bias": "model.safetensors",
79
+ "decoder.local_attention_decoder.layers.5.norm.weight": "model.safetensors",
80
+ "decoder.local_attention_decoder.layers.5.self_attn.layer_norm.bias": "model.safetensors",
81
+ "decoder.local_attention_decoder.layers.5.self_attn.layer_norm.weight": "model.safetensors",
82
+ "decoder.local_attention_decoder.layers.5.self_attn.out_proj.bias": "model.safetensors",
83
+ "decoder.local_attention_decoder.layers.5.self_attn.out_proj.weight": "model.safetensors",
84
+ "decoder.local_attention_decoder.layers.5.self_attn.qkv.bias": "model.safetensors",
85
+ "decoder.local_attention_decoder.layers.5.self_attn.qkv.weight": "model.safetensors",
86
+ "decoder.wav_decoder.blocks.0.conv_transpose.bias": "model.safetensors",
87
+ "decoder.wav_decoder.blocks.0.conv_transpose.weight": "model.safetensors",
88
+ "decoder.wav_decoder.blocks.0.res1.conv1.bias": "model.safetensors",
89
+ "decoder.wav_decoder.blocks.0.res1.conv1.weight": "model.safetensors",
90
+ "decoder.wav_decoder.blocks.0.res1.conv2.bias": "model.safetensors",
91
+ "decoder.wav_decoder.blocks.0.res1.conv2.weight": "model.safetensors",
92
+ "decoder.wav_decoder.blocks.0.res1.snake1.alpha": "model.safetensors",
93
+ "decoder.wav_decoder.blocks.0.res1.snake2.alpha": "model.safetensors",
94
+ "decoder.wav_decoder.blocks.0.res2.conv1.bias": "model.safetensors",
95
+ "decoder.wav_decoder.blocks.0.res2.conv1.weight": "model.safetensors",
96
+ "decoder.wav_decoder.blocks.0.res2.conv2.bias": "model.safetensors",
97
+ "decoder.wav_decoder.blocks.0.res2.conv2.weight": "model.safetensors",
98
+ "decoder.wav_decoder.blocks.0.res2.snake1.alpha": "model.safetensors",
99
+ "decoder.wav_decoder.blocks.0.res2.snake2.alpha": "model.safetensors",
100
+ "decoder.wav_decoder.blocks.0.res3.conv1.bias": "model.safetensors",
101
+ "decoder.wav_decoder.blocks.0.res3.conv1.weight": "model.safetensors",
102
+ "decoder.wav_decoder.blocks.0.res3.conv2.bias": "model.safetensors",
103
+ "decoder.wav_decoder.blocks.0.res3.conv2.weight": "model.safetensors",
104
+ "decoder.wav_decoder.blocks.0.res3.snake1.alpha": "model.safetensors",
105
+ "decoder.wav_decoder.blocks.0.res3.snake2.alpha": "model.safetensors",
106
+ "decoder.wav_decoder.blocks.0.snake.alpha": "model.safetensors",
107
+ "decoder.wav_decoder.blocks.1.conv_transpose.bias": "model.safetensors",
108
+ "decoder.wav_decoder.blocks.1.conv_transpose.weight": "model.safetensors",
109
+ "decoder.wav_decoder.blocks.1.res1.conv1.bias": "model.safetensors",
110
+ "decoder.wav_decoder.blocks.1.res1.conv1.weight": "model.safetensors",
111
+ "decoder.wav_decoder.blocks.1.res1.conv2.bias": "model.safetensors",
112
+ "decoder.wav_decoder.blocks.1.res1.conv2.weight": "model.safetensors",
113
+ "decoder.wav_decoder.blocks.1.res1.snake1.alpha": "model.safetensors",
114
+ "decoder.wav_decoder.blocks.1.res1.snake2.alpha": "model.safetensors",
115
+ "decoder.wav_decoder.blocks.1.res2.conv1.bias": "model.safetensors",
116
+ "decoder.wav_decoder.blocks.1.res2.conv1.weight": "model.safetensors",
117
+ "decoder.wav_decoder.blocks.1.res2.conv2.bias": "model.safetensors",
118
+ "decoder.wav_decoder.blocks.1.res2.conv2.weight": "model.safetensors",
119
+ "decoder.wav_decoder.blocks.1.res2.snake1.alpha": "model.safetensors",
120
+ "decoder.wav_decoder.blocks.1.res2.snake2.alpha": "model.safetensors",
121
+ "decoder.wav_decoder.blocks.1.res3.conv1.bias": "model.safetensors",
122
+ "decoder.wav_decoder.blocks.1.res3.conv1.weight": "model.safetensors",
123
+ "decoder.wav_decoder.blocks.1.res3.conv2.bias": "model.safetensors",
124
+ "decoder.wav_decoder.blocks.1.res3.conv2.weight": "model.safetensors",
125
+ "decoder.wav_decoder.blocks.1.res3.snake1.alpha": "model.safetensors",
126
+ "decoder.wav_decoder.blocks.1.res3.snake2.alpha": "model.safetensors",
127
+ "decoder.wav_decoder.blocks.1.snake.alpha": "model.safetensors",
128
+ "decoder.wav_decoder.blocks.2.conv_transpose.bias": "model.safetensors",
129
+ "decoder.wav_decoder.blocks.2.conv_transpose.weight": "model.safetensors",
130
+ "decoder.wav_decoder.blocks.2.res1.conv1.bias": "model.safetensors",
131
+ "decoder.wav_decoder.blocks.2.res1.conv1.weight": "model.safetensors",
132
+ "decoder.wav_decoder.blocks.2.res1.conv2.bias": "model.safetensors",
133
+ "decoder.wav_decoder.blocks.2.res1.conv2.weight": "model.safetensors",
134
+ "decoder.wav_decoder.blocks.2.res1.snake1.alpha": "model.safetensors",
135
+ "decoder.wav_decoder.blocks.2.res1.snake2.alpha": "model.safetensors",
136
+ "decoder.wav_decoder.blocks.2.res2.conv1.bias": "model.safetensors",
137
+ "decoder.wav_decoder.blocks.2.res2.conv1.weight": "model.safetensors",
138
+ "decoder.wav_decoder.blocks.2.res2.conv2.bias": "model.safetensors",
139
+ "decoder.wav_decoder.blocks.2.res2.conv2.weight": "model.safetensors",
140
+ "decoder.wav_decoder.blocks.2.res2.snake1.alpha": "model.safetensors",
141
+ "decoder.wav_decoder.blocks.2.res2.snake2.alpha": "model.safetensors",
142
+ "decoder.wav_decoder.blocks.2.res3.conv1.bias": "model.safetensors",
143
+ "decoder.wav_decoder.blocks.2.res3.conv1.weight": "model.safetensors",
144
+ "decoder.wav_decoder.blocks.2.res3.conv2.bias": "model.safetensors",
145
+ "decoder.wav_decoder.blocks.2.res3.conv2.weight": "model.safetensors",
146
+ "decoder.wav_decoder.blocks.2.res3.snake1.alpha": "model.safetensors",
147
+ "decoder.wav_decoder.blocks.2.res3.snake2.alpha": "model.safetensors",
148
+ "decoder.wav_decoder.blocks.2.snake.alpha": "model.safetensors",
149
+ "decoder.wav_decoder.blocks.3.conv_transpose.bias": "model.safetensors",
150
+ "decoder.wav_decoder.blocks.3.conv_transpose.weight": "model.safetensors",
151
+ "decoder.wav_decoder.blocks.3.res1.conv1.bias": "model.safetensors",
152
+ "decoder.wav_decoder.blocks.3.res1.conv1.weight": "model.safetensors",
153
+ "decoder.wav_decoder.blocks.3.res1.conv2.bias": "model.safetensors",
154
+ "decoder.wav_decoder.blocks.3.res1.conv2.weight": "model.safetensors",
155
+ "decoder.wav_decoder.blocks.3.res1.snake1.alpha": "model.safetensors",
156
+ "decoder.wav_decoder.blocks.3.res1.snake2.alpha": "model.safetensors",
157
+ "decoder.wav_decoder.blocks.3.res2.conv1.bias": "model.safetensors",
158
+ "decoder.wav_decoder.blocks.3.res2.conv1.weight": "model.safetensors",
159
+ "decoder.wav_decoder.blocks.3.res2.conv2.bias": "model.safetensors",
160
+ "decoder.wav_decoder.blocks.3.res2.conv2.weight": "model.safetensors",
161
+ "decoder.wav_decoder.blocks.3.res2.snake1.alpha": "model.safetensors",
162
+ "decoder.wav_decoder.blocks.3.res2.snake2.alpha": "model.safetensors",
163
+ "decoder.wav_decoder.blocks.3.res3.conv1.bias": "model.safetensors",
164
+ "decoder.wav_decoder.blocks.3.res3.conv1.weight": "model.safetensors",
165
+ "decoder.wav_decoder.blocks.3.res3.conv2.bias": "model.safetensors",
166
+ "decoder.wav_decoder.blocks.3.res3.conv2.weight": "model.safetensors",
167
+ "decoder.wav_decoder.blocks.3.res3.snake1.alpha": "model.safetensors",
168
+ "decoder.wav_decoder.blocks.3.res3.snake2.alpha": "model.safetensors",
169
+ "decoder.wav_decoder.blocks.3.snake.alpha": "model.safetensors",
170
+ "decoder.wav_decoder.final_conv.bias": "model.safetensors",
171
+ "decoder.wav_decoder.final_conv.weight": "model.safetensors",
172
+ "decoder.wav_decoder.final_snake.alpha": "model.safetensors",
173
+ "decoder.wav_decoder.initial_conv.bias": "model.safetensors",
174
+ "decoder.wav_decoder.initial_conv.weight": "model.safetensors",
175
+ "model.embed_tokens.weight": "model.safetensors",
176
+ "model.layers.0.input_layernorm.weight": "model.safetensors",
177
+ "model.layers.0.mlp.down_proj.weight": "model.safetensors",
178
+ "model.layers.0.mlp.gate_proj.weight": "model.safetensors",
179
+ "model.layers.0.mlp.up_proj.weight": "model.safetensors",
180
+ "model.layers.0.post_attention_layernorm.weight": "model.safetensors",
181
+ "model.layers.0.self_attn.k_proj.weight": "model.safetensors",
182
+ "model.layers.0.self_attn.o_proj.weight": "model.safetensors",
183
+ "model.layers.0.self_attn.q_proj.weight": "model.safetensors",
184
+ "model.layers.0.self_attn.v_proj.weight": "model.safetensors",
185
+ "model.layers.1.input_layernorm.weight": "model.safetensors",
186
+ "model.layers.1.mlp.down_proj.weight": "model.safetensors",
187
+ "model.layers.1.mlp.gate_proj.weight": "model.safetensors",
188
+ "model.layers.1.mlp.up_proj.weight": "model.safetensors",
189
+ "model.layers.1.post_attention_layernorm.weight": "model.safetensors",
190
+ "model.layers.1.self_attn.k_proj.weight": "model.safetensors",
191
+ "model.layers.1.self_attn.o_proj.weight": "model.safetensors",
192
+ "model.layers.1.self_attn.q_proj.weight": "model.safetensors",
193
+ "model.layers.1.self_attn.v_proj.weight": "model.safetensors",
194
+ "model.layers.10.input_layernorm.weight": "model.safetensors",
195
+ "model.layers.10.mlp.down_proj.weight": "model.safetensors",
196
+ "model.layers.10.mlp.gate_proj.weight": "model.safetensors",
197
+ "model.layers.10.mlp.up_proj.weight": "model.safetensors",
198
+ "model.layers.10.post_attention_layernorm.weight": "model.safetensors",
199
+ "model.layers.10.self_attn.k_proj.weight": "model.safetensors",
200
+ "model.layers.10.self_attn.o_proj.weight": "model.safetensors",
201
+ "model.layers.10.self_attn.q_proj.weight": "model.safetensors",
202
+ "model.layers.10.self_attn.v_proj.weight": "model.safetensors",
203
+ "model.layers.11.input_layernorm.weight": "model.safetensors",
204
+ "model.layers.11.mlp.down_proj.weight": "model.safetensors",
205
+ "model.layers.11.mlp.gate_proj.weight": "model.safetensors",
206
+ "model.layers.11.mlp.up_proj.weight": "model.safetensors",
207
+ "model.layers.11.post_attention_layernorm.weight": "model.safetensors",
208
+ "model.layers.11.self_attn.k_proj.weight": "model.safetensors",
209
+ "model.layers.11.self_attn.o_proj.weight": "model.safetensors",
210
+ "model.layers.11.self_attn.q_proj.weight": "model.safetensors",
211
+ "model.layers.11.self_attn.v_proj.weight": "model.safetensors",
212
+ "model.layers.12.input_layernorm.weight": "model.safetensors",
213
+ "model.layers.12.mlp.down_proj.weight": "model.safetensors",
214
+ "model.layers.12.mlp.gate_proj.weight": "model.safetensors",
215
+ "model.layers.12.mlp.up_proj.weight": "model.safetensors",
216
+ "model.layers.12.post_attention_layernorm.weight": "model.safetensors",
217
+ "model.layers.12.self_attn.k_proj.weight": "model.safetensors",
218
+ "model.layers.12.self_attn.o_proj.weight": "model.safetensors",
219
+ "model.layers.12.self_attn.q_proj.weight": "model.safetensors",
220
+ "model.layers.12.self_attn.v_proj.weight": "model.safetensors",
221
+ "model.layers.13.input_layernorm.weight": "model.safetensors",
222
+ "model.layers.13.mlp.down_proj.weight": "model.safetensors",
223
+ "model.layers.13.mlp.gate_proj.weight": "model.safetensors",
224
+ "model.layers.13.mlp.up_proj.weight": "model.safetensors",
225
+ "model.layers.13.post_attention_layernorm.weight": "model.safetensors",
226
+ "model.layers.13.self_attn.k_proj.weight": "model.safetensors",
227
+ "model.layers.13.self_attn.o_proj.weight": "model.safetensors",
228
+ "model.layers.13.self_attn.q_proj.weight": "model.safetensors",
229
+ "model.layers.13.self_attn.v_proj.weight": "model.safetensors",
230
+ "model.layers.14.input_layernorm.weight": "model.safetensors",
231
+ "model.layers.14.mlp.down_proj.weight": "model.safetensors",
232
+ "model.layers.14.mlp.gate_proj.weight": "model.safetensors",
233
+ "model.layers.14.mlp.up_proj.weight": "model.safetensors",
234
+ "model.layers.14.post_attention_layernorm.weight": "model.safetensors",
235
+ "model.layers.14.self_attn.k_proj.weight": "model.safetensors",
236
+ "model.layers.14.self_attn.o_proj.weight": "model.safetensors",
237
+ "model.layers.14.self_attn.q_proj.weight": "model.safetensors",
238
+ "model.layers.14.self_attn.v_proj.weight": "model.safetensors",
239
+ "model.layers.15.input_layernorm.weight": "model.safetensors",
240
+ "model.layers.15.mlp.down_proj.weight": "model.safetensors",
241
+ "model.layers.15.mlp.gate_proj.weight": "model.safetensors",
242
+ "model.layers.15.mlp.up_proj.weight": "model.safetensors",
243
+ "model.layers.15.post_attention_layernorm.weight": "model.safetensors",
244
+ "model.layers.15.self_attn.k_proj.weight": "model.safetensors",
245
+ "model.layers.15.self_attn.o_proj.weight": "model.safetensors",
246
+ "model.layers.15.self_attn.q_proj.weight": "model.safetensors",
247
+ "model.layers.15.self_attn.v_proj.weight": "model.safetensors",
248
+ "model.layers.2.input_layernorm.weight": "model.safetensors",
249
+ "model.layers.2.mlp.down_proj.weight": "model.safetensors",
250
+ "model.layers.2.mlp.gate_proj.weight": "model.safetensors",
251
+ "model.layers.2.mlp.up_proj.weight": "model.safetensors",
252
+ "model.layers.2.post_attention_layernorm.weight": "model.safetensors",
253
+ "model.layers.2.self_attn.k_proj.weight": "model.safetensors",
254
+ "model.layers.2.self_attn.o_proj.weight": "model.safetensors",
255
+ "model.layers.2.self_attn.q_proj.weight": "model.safetensors",
256
+ "model.layers.2.self_attn.v_proj.weight": "model.safetensors",
257
+ "model.layers.3.input_layernorm.weight": "model.safetensors",
258
+ "model.layers.3.mlp.down_proj.weight": "model.safetensors",
259
+ "model.layers.3.mlp.gate_proj.weight": "model.safetensors",
260
+ "model.layers.3.mlp.up_proj.weight": "model.safetensors",
261
+ "model.layers.3.post_attention_layernorm.weight": "model.safetensors",
262
+ "model.layers.3.self_attn.k_proj.weight": "model.safetensors",
263
+ "model.layers.3.self_attn.o_proj.weight": "model.safetensors",
264
+ "model.layers.3.self_attn.q_proj.weight": "model.safetensors",
265
+ "model.layers.3.self_attn.v_proj.weight": "model.safetensors",
266
+ "model.layers.4.input_layernorm.weight": "model.safetensors",
267
+ "model.layers.4.mlp.down_proj.weight": "model.safetensors",
268
+ "model.layers.4.mlp.gate_proj.weight": "model.safetensors",
269
+ "model.layers.4.mlp.up_proj.weight": "model.safetensors",
270
+ "model.layers.4.post_attention_layernorm.weight": "model.safetensors",
271
+ "model.layers.4.self_attn.k_proj.weight": "model.safetensors",
272
+ "model.layers.4.self_attn.o_proj.weight": "model.safetensors",
273
+ "model.layers.4.self_attn.q_proj.weight": "model.safetensors",
274
+ "model.layers.4.self_attn.v_proj.weight": "model.safetensors",
275
+ "model.layers.5.input_layernorm.weight": "model.safetensors",
276
+ "model.layers.5.mlp.down_proj.weight": "model.safetensors",
277
+ "model.layers.5.mlp.gate_proj.weight": "model.safetensors",
278
+ "model.layers.5.mlp.up_proj.weight": "model.safetensors",
279
+ "model.layers.5.post_attention_layernorm.weight": "model.safetensors",
280
+ "model.layers.5.self_attn.k_proj.weight": "model.safetensors",
281
+ "model.layers.5.self_attn.o_proj.weight": "model.safetensors",
282
+ "model.layers.5.self_attn.q_proj.weight": "model.safetensors",
283
+ "model.layers.5.self_attn.v_proj.weight": "model.safetensors",
284
+ "model.layers.6.input_layernorm.weight": "model.safetensors",
285
+ "model.layers.6.mlp.down_proj.weight": "model.safetensors",
286
+ "model.layers.6.mlp.gate_proj.weight": "model.safetensors",
287
+ "model.layers.6.mlp.up_proj.weight": "model.safetensors",
288
+ "model.layers.6.post_attention_layernorm.weight": "model.safetensors",
289
+ "model.layers.6.self_attn.k_proj.weight": "model.safetensors",
290
+ "model.layers.6.self_attn.o_proj.weight": "model.safetensors",
291
+ "model.layers.6.self_attn.q_proj.weight": "model.safetensors",
292
+ "model.layers.6.self_attn.v_proj.weight": "model.safetensors",
293
+ "model.layers.7.input_layernorm.weight": "model.safetensors",
294
+ "model.layers.7.mlp.down_proj.weight": "model.safetensors",
295
+ "model.layers.7.mlp.gate_proj.weight": "model.safetensors",
296
+ "model.layers.7.mlp.up_proj.weight": "model.safetensors",
297
+ "model.layers.7.post_attention_layernorm.weight": "model.safetensors",
298
+ "model.layers.7.self_attn.k_proj.weight": "model.safetensors",
299
+ "model.layers.7.self_attn.o_proj.weight": "model.safetensors",
300
+ "model.layers.7.self_attn.q_proj.weight": "model.safetensors",
301
+ "model.layers.7.self_attn.v_proj.weight": "model.safetensors",
302
+ "model.layers.8.input_layernorm.weight": "model.safetensors",
303
+ "model.layers.8.mlp.down_proj.weight": "model.safetensors",
304
+ "model.layers.8.mlp.gate_proj.weight": "model.safetensors",
305
+ "model.layers.8.mlp.up_proj.weight": "model.safetensors",
306
+ "model.layers.8.post_attention_layernorm.weight": "model.safetensors",
307
+ "model.layers.8.self_attn.k_proj.weight": "model.safetensors",
308
+ "model.layers.8.self_attn.o_proj.weight": "model.safetensors",
309
+ "model.layers.8.self_attn.q_proj.weight": "model.safetensors",
310
+ "model.layers.8.self_attn.v_proj.weight": "model.safetensors",
311
+ "model.layers.9.input_layernorm.weight": "model.safetensors",
312
+ "model.layers.9.mlp.down_proj.weight": "model.safetensors",
313
+ "model.layers.9.mlp.gate_proj.weight": "model.safetensors",
314
+ "model.layers.9.mlp.up_proj.weight": "model.safetensors",
315
+ "model.layers.9.post_attention_layernorm.weight": "model.safetensors",
316
+ "model.layers.9.self_attn.k_proj.weight": "model.safetensors",
317
+ "model.layers.9.self_attn.o_proj.weight": "model.safetensors",
318
+ "model.layers.9.self_attn.q_proj.weight": "model.safetensors",
319
+ "model.layers.9.self_attn.v_proj.weight": "model.safetensors",
320
+ "model.norm.weight": "model.safetensors",
321
+ "prediction_head.cond_proj.weight": "model.safetensors",
322
+ "prediction_head.final_layer.adaLN_modulation.layers.1.weight": "model.safetensors",
323
+ "prediction_head.final_layer.linear.weight": "model.safetensors",
324
+ "prediction_head.layers.0.adaLN_modulation.layers.1.weight": "model.safetensors",
325
+ "prediction_head.layers.0.ffn.down_proj.weight": "model.safetensors",
326
+ "prediction_head.layers.0.ffn.gate_proj.weight": "model.safetensors",
327
+ "prediction_head.layers.0.ffn.up_proj.weight": "model.safetensors",
328
+ "prediction_head.layers.0.norm.weight": "model.safetensors",
329
+ "prediction_head.layers.1.adaLN_modulation.layers.1.weight": "model.safetensors",
330
+ "prediction_head.layers.1.ffn.down_proj.weight": "model.safetensors",
331
+ "prediction_head.layers.1.ffn.gate_proj.weight": "model.safetensors",
332
+ "prediction_head.layers.1.ffn.up_proj.weight": "model.safetensors",
333
+ "prediction_head.layers.1.norm.weight": "model.safetensors",
334
+ "prediction_head.layers.2.adaLN_modulation.layers.1.weight": "model.safetensors",
335
+ "prediction_head.layers.2.ffn.down_proj.weight": "model.safetensors",
336
+ "prediction_head.layers.2.ffn.gate_proj.weight": "model.safetensors",
337
+ "prediction_head.layers.2.ffn.up_proj.weight": "model.safetensors",
338
+ "prediction_head.layers.2.norm.weight": "model.safetensors",
339
+ "prediction_head.layers.3.adaLN_modulation.layers.1.weight": "model.safetensors",
340
+ "prediction_head.layers.3.ffn.down_proj.weight": "model.safetensors",
341
+ "prediction_head.layers.3.ffn.gate_proj.weight": "model.safetensors",
342
+ "prediction_head.layers.3.ffn.up_proj.weight": "model.safetensors",
343
+ "prediction_head.layers.3.norm.weight": "model.safetensors",
344
+ "prediction_head.layers.4.adaLN_modulation.layers.1.weight": "model.safetensors",
345
+ "prediction_head.layers.4.ffn.down_proj.weight": "model.safetensors",
346
+ "prediction_head.layers.4.ffn.gate_proj.weight": "model.safetensors",
347
+ "prediction_head.layers.4.ffn.up_proj.weight": "model.safetensors",
348
+ "prediction_head.layers.4.norm.weight": "model.safetensors",
349
+ "prediction_head.layers.5.adaLN_modulation.layers.1.weight": "model.safetensors",
350
+ "prediction_head.layers.5.ffn.down_proj.weight": "model.safetensors",
351
+ "prediction_head.layers.5.ffn.gate_proj.weight": "model.safetensors",
352
+ "prediction_head.layers.5.ffn.up_proj.weight": "model.safetensors",
353
+ "prediction_head.layers.5.norm.weight": "model.safetensors",
354
+ "prediction_head.noisy_images_proj.weight": "model.safetensors",
355
+ "prediction_head.t_embedder.mlp.layers.0.weight": "model.safetensors",
356
+ "prediction_head.t_embedder.mlp.layers.2.weight": "model.safetensors",
357
+ "time_end_embed.weight": "model.safetensors",
358
+ "time_start_embed.weight": "model.safetensors"
359
+ }
360
+ }
model/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "acoustic_dim": 512,
3
+ "acoustic_from_nth_hidden_state": -1,
4
+ "acoustic_mean": 0.0,
5
+ "acoustic_std": 1.5,
6
+ "add_semantic_to_condition": 0.0,
7
+ "architectures": [
8
+ "TadaForCausalLM"
9
+ ],
10
+ "attention_bias": false,
11
+ "attention_dropout": 0.0,
12
+ "bos_token_id": 128000,
13
+ "bottleneck_dim": null,
14
+ "context_window": 8,
15
+ "diffusion_head_type": "vibevoice",
16
+ "dist_type": "fixed",
17
+ "dtype": "bfloat16",
18
+ "eos_token_id": 128001,
19
+ "head_dim": 64,
20
+ "head_ffn_ratio": 4.0,
21
+ "head_layers": 6,
22
+ "hidden_act": "silu",
23
+ "hidden_size": 2048,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 8192,
26
+ "latent_dropout": 0.0,
27
+ "max_position_embeddings": 131072,
28
+ "mlp_bias": false,
29
+ "model_type": "llama",
30
+ "num_attention_heads": 32,
31
+ "num_hidden_layers": 16,
32
+ "num_key_value_heads": 8,
33
+ "num_time_classes": 256,
34
+ "pretraining_tp": 1,
35
+ "rms_norm_eps": 1e-05,
36
+ "rope_scaling": {
37
+ "factor": 32.0,
38
+ "high_freq_factor": 4.0,
39
+ "low_freq_factor": 1.0,
40
+ "original_max_position_embeddings": 8192,
41
+ "rope_type": "llama3"
42
+ },
43
+ "rope_theta": 500000.0,
44
+ "shift_acoustic": 5,
45
+ "tie_word_embeddings": true,
46
+ "transformers_version": "4.57.3",
47
+ "use_cache": true,
48
+ "vocab_size": 128256
49
+ }
model/weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45b45dcbc3faa9efa11c6aa6a6a84290f26a1c6944b860521526be4dc30d4e4b
3
+ size 3269784687