Hugging Face's logo

mlx-community
/
sam-audio-small

Audio-to-Audio
Safetensors
MLX
English
mlx-audio
speech
speech generation
voice isolation
Model card Files
xet
Community
prince-canuma commited on
Commit
1be2e40
·
verified ·
1 Parent(s): 1919519

Upload folder using huggingface_hub

Browse files
Files changed (4) hide show
  1. README.md +71 -0
  2. config.json +92 -0
  3. model.safetensors +3 -0
  4. model.safetensors.index.json +563 -0
README.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ base_model:
5
+ - facebook/sam-audio-large
6
+ pipeline_tag: speech-to-speech
7
+ library_name: mlx-audio
8
+ tags:
9
+ - speech-to-speech
10
+ - speech
11
+ - speech generation
12
+ - voice isolation
13
+ - mlx
14
+ ---
15
+ # mlx-community/sam-audio-large
16
+ This model was converted to MLX format from [`facebook/sam-audio-large`](https://huggingface.co/facebook/sam-audio-large) using mlx-audio version **0.2.10**.
17
+ Refer to the [original model card](https://huggingface.co/facebook/sam-audio-large) for more details on the model.
18
+
19
+ ## Use with mlx
20
+ ```bash
21
+ pip install -U mlx-audio
22
+ ```
23
+
24
+ ## Voice Isolation:
25
+ ```python
26
+ from mlx_audio.sts import SAMAudio, SAMAudioProcessor, save_audio
27
+ import mlx.core as mx
28
+
29
+ # Load model and processor
30
+ processor = SAMAudioProcessor.from_pretrained("facebook/sam-audio-large")
31
+ model = SAMAudio.from_pretrained("facebook/sam-audio-large")
32
+
33
+ # Process inputs
34
+ batch = processor(
35
+ descriptions=["speech"],
36
+ audios=["path/to/audio.mp3"],
37
+ # anchors=[[("+", 0.2, 0.5)]], # Optional: temporal
38
+ )
39
+
40
+ # Separate audio
41
+ result = model.separate(
42
+ audios=batch.audios,
43
+ descriptions=batch.descriptions,
44
+ sizes=batch.sizes,
45
+ anchor_ids=batch.anchor_ids,
46
+ anchor_alignment=batch.anchor_alignment,
47
+ ode_decode_chunk_size=50, # Chunked decoding for memory efficiency
48
+ )
49
+
50
+ # For long audio files, use separate_long().
51
+ # Note: This is slower than separate() but it is more memory efficient.
52
+ # result = model.separate_long(
53
+ # audios=batch.audios,
54
+ # descriptions=batch.descriptions,
55
+ # chunk_seconds=10.0,
56
+ # overlap_seconds=3.0,
57
+ # anchor_ids=batch.anchor_ids,
58
+ # anchor_alignment=batch.anchor_alignment,
59
+ # ode_decode_chunk_size=50, # Chunked decoding for memory efficiency
60
+ # )
61
+
62
+ # Save output
63
+ ## Isolated speech
64
+ save_audio(result.target[0], "separated.wav", sample_rate=model.sample_rate)
65
+
66
+ ## Residual audio (background music/noise/other sounds)
67
+ save_audio(result.residual[0], "residual.wav", sample_rate=model.sample_rate)
68
+
69
+ # Check memory usage
70
+ print(f"Peak memory: {result.peak_memory:.2f} GB")
71
+ ```
config.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "in_channels": 768,
3
+ "audio_codec": {
4
+ "encoder_dim": 64,
5
+ "encoder_rates": [
6
+ 2,
7
+ 8,
8
+ 10,
9
+ 12
10
+ ],
11
+ "latent_dim": 1024,
12
+ "decoder_dim": 1536,
13
+ "decoder_rates": [
14
+ 12,
15
+ 10,
16
+ 8,
17
+ 2
18
+ ],
19
+ "n_codebooks": 16,
20
+ "codebook_size": 1024,
21
+ "codebook_dim": 128,
22
+ "quantizer_dropout": false,
23
+ "sample_rate": 48000,
24
+ "mean": 0.0,
25
+ "std": 1.0
26
+ },
27
+ "text_encoder": {
28
+ "dim": 768,
29
+ "name": "t5-base",
30
+ "max_length": 512,
31
+ "pad_mode": "longest"
32
+ },
33
+ "vision_encoder": {
34
+ "dim": 1024,
35
+ "batch_size": 300,
36
+ "name": "PE-Core-L14-336",
37
+ "normalize_feature": true,
38
+ "interpolation_mode": "BICUBIC",
39
+ "image_size": 336
40
+ },
41
+ "transformer": {
42
+ "dim": 1536,
43
+ "n_heads": 12,
44
+ "n_layers": 12,
45
+ "dropout": 0.1,
46
+ "norm_eps": 1e-05,
47
+ "qk_norm": true,
48
+ "fc_bias": false,
49
+ "ffn_exp": 4,
50
+ "ffn_dim_multiplier": 1,
51
+ "multiple_of": 64,
52
+ "non_linearity": "swiglu",
53
+ "use_rope": true,
54
+ "max_positions": 10000,
55
+ "frequency_embedding_dim": 256,
56
+ "timestep_non_linearity": "swiglu",
57
+ "t_block_non_linearity": "silu",
58
+ "t_block_bias": true,
59
+ "context_dim": 1536,
60
+ "context_non_linearity": "swiglu",
61
+ "context_embedder_dropout": 0.0,
62
+ "context_norm": false,
63
+ "out_channels": 256,
64
+ "in_channels": null
65
+ },
66
+ "num_anchors": 3,
67
+ "anchor_embedding_dim": 128,
68
+ "visual_ranker": {
69
+ "checkpoint": null,
70
+ "kind": "imagebind"
71
+ },
72
+ "text_ranker": {
73
+ "rankers": {
74
+ "clap": [
75
+ {
76
+ "checkpoint": null,
77
+ "kind": "clap"
78
+ },
79
+ 5.0
80
+ ],
81
+ "judge": [
82
+ {
83
+ "checkpoint_or_model_id": "facebook/sam-audio-judge",
84
+ "kind": "judge"
85
+ },
86
+ 1.0
87
+ ]
88
+ },
89
+ "kind": "ensemble"
90
+ },
91
+ "span_predictor": "pe-a-frame-large"
92
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae851c17699821e4a7a7c6b0aae44be61c190ceb3c9a603b2a2e5b327d2509f8
3
+ size 2409315890
model.safetensors.index.json ADDED
@@ -0,0 +1,563 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 2409249296,
4
+ "total_parameters": 602199429
5
+ },
6
+ "weight_map": {
7
+ "audio_codec.decoder.blocks.0.block_0.alpha": "model.safetensors",
8
+ "audio_codec.decoder.blocks.0.block_1.bias": "model.safetensors",
9
+ "audio_codec.decoder.blocks.0.block_1.weight_g": "model.safetensors",
10
+ "audio_codec.decoder.blocks.0.block_1.weight_v": "model.safetensors",
11
+ "audio_codec.decoder.blocks.0.block_11.bias": "model.safetensors",
12
+ "audio_codec.decoder.blocks.0.block_11.weight": "model.safetensors",
13
+ "audio_codec.decoder.blocks.0.block_3.bias": "model.safetensors",
14
+ "audio_codec.decoder.blocks.0.block_3.weight": "model.safetensors",
15
+ "audio_codec.decoder.blocks.0.block_4.act1.alpha": "model.safetensors",
16
+ "audio_codec.decoder.blocks.0.block_4.act2.alpha": "model.safetensors",
17
+ "audio_codec.decoder.blocks.0.block_4.conv1.bias": "model.safetensors",
18
+ "audio_codec.decoder.blocks.0.block_4.conv1.weight_g": "model.safetensors",
19
+ "audio_codec.decoder.blocks.0.block_4.conv1.weight_v": "model.safetensors",
20
+ "audio_codec.decoder.blocks.0.block_4.conv2.bias": "model.safetensors",
21
+ "audio_codec.decoder.blocks.0.block_4.conv2.weight_g": "model.safetensors",
22
+ "audio_codec.decoder.blocks.0.block_4.conv2.weight_v": "model.safetensors",
23
+ "audio_codec.decoder.blocks.0.block_5.act1.alpha": "model.safetensors",
24
+ "audio_codec.decoder.blocks.0.block_5.act2.alpha": "model.safetensors",
25
+ "audio_codec.decoder.blocks.0.block_5.conv1.bias": "model.safetensors",
26
+ "audio_codec.decoder.blocks.0.block_5.conv1.weight_g": "model.safetensors",
27
+ "audio_codec.decoder.blocks.0.block_5.conv1.weight_v": "model.safetensors",
28
+ "audio_codec.decoder.blocks.0.block_5.conv2.bias": "model.safetensors",
29
+ "audio_codec.decoder.blocks.0.block_5.conv2.weight_g": "model.safetensors",
30
+ "audio_codec.decoder.blocks.0.block_5.conv2.weight_v": "model.safetensors",
31
+ "audio_codec.decoder.blocks.0.block_6.conv1.bias": "model.safetensors",
32
+ "audio_codec.decoder.blocks.0.block_6.conv1.weight": "model.safetensors",
33
+ "audio_codec.decoder.blocks.0.block_6.conv2.bias": "model.safetensors",
34
+ "audio_codec.decoder.blocks.0.block_6.conv2.weight": "model.safetensors",
35
+ "audio_codec.decoder.blocks.0.block_7.conv1.bias": "model.safetensors",
36
+ "audio_codec.decoder.blocks.0.block_7.conv1.weight": "model.safetensors",
37
+ "audio_codec.decoder.blocks.0.block_7.conv2.bias": "model.safetensors",
38
+ "audio_codec.decoder.blocks.0.block_7.conv2.weight": "model.safetensors",
39
+ "audio_codec.decoder.blocks.0.block_8.act1.alpha": "model.safetensors",
40
+ "audio_codec.decoder.blocks.0.block_8.act2.alpha": "model.safetensors",
41
+ "audio_codec.decoder.blocks.0.block_8.conv1.bias": "model.safetensors",
42
+ "audio_codec.decoder.blocks.0.block_8.conv1.weight_g": "model.safetensors",
43
+ "audio_codec.decoder.blocks.0.block_8.conv1.weight_v": "model.safetensors",
44
+ "audio_codec.decoder.blocks.0.block_8.conv2.bias": "model.safetensors",
45
+ "audio_codec.decoder.blocks.0.block_8.conv2.weight_g": "model.safetensors",
46
+ "audio_codec.decoder.blocks.0.block_8.conv2.weight_v": "model.safetensors",
47
+ "audio_codec.decoder.blocks.1.block_0.alpha": "model.safetensors",
48
+ "audio_codec.decoder.blocks.1.block_1.bias": "model.safetensors",
49
+ "audio_codec.decoder.blocks.1.block_1.weight_g": "model.safetensors",
50
+ "audio_codec.decoder.blocks.1.block_1.weight_v": "model.safetensors",
51
+ "audio_codec.decoder.blocks.1.block_11.bias": "model.safetensors",
52
+ "audio_codec.decoder.blocks.1.block_11.weight": "model.safetensors",
53
+ "audio_codec.decoder.blocks.1.block_3.bias": "model.safetensors",
54
+ "audio_codec.decoder.blocks.1.block_3.weight": "model.safetensors",
55
+ "audio_codec.decoder.blocks.1.block_4.act1.alpha": "model.safetensors",
56
+ "audio_codec.decoder.blocks.1.block_4.act2.alpha": "model.safetensors",
57
+ "audio_codec.decoder.blocks.1.block_4.conv1.bias": "model.safetensors",
58
+ "audio_codec.decoder.blocks.1.block_4.conv1.weight_g": "model.safetensors",
59
+ "audio_codec.decoder.blocks.1.block_4.conv1.weight_v": "model.safetensors",
60
+ "audio_codec.decoder.blocks.1.block_4.conv2.bias": "model.safetensors",
61
+ "audio_codec.decoder.blocks.1.block_4.conv2.weight_g": "model.safetensors",
62
+ "audio_codec.decoder.blocks.1.block_4.conv2.weight_v": "model.safetensors",
63
+ "audio_codec.decoder.blocks.1.block_5.act1.alpha": "model.safetensors",
64
+ "audio_codec.decoder.blocks.1.block_5.act2.alpha": "model.safetensors",
65
+ "audio_codec.decoder.blocks.1.block_5.conv1.bias": "model.safetensors",
66
+ "audio_codec.decoder.blocks.1.block_5.conv1.weight_g": "model.safetensors",
67
+ "audio_codec.decoder.blocks.1.block_5.conv1.weight_v": "model.safetensors",
68
+ "audio_codec.decoder.blocks.1.block_5.conv2.bias": "model.safetensors",
69
+ "audio_codec.decoder.blocks.1.block_5.conv2.weight_g": "model.safetensors",
70
+ "audio_codec.decoder.blocks.1.block_5.conv2.weight_v": "model.safetensors",
71
+ "audio_codec.decoder.blocks.1.block_6.conv1.bias": "model.safetensors",
72
+ "audio_codec.decoder.blocks.1.block_6.conv1.weight": "model.safetensors",
73
+ "audio_codec.decoder.blocks.1.block_6.conv2.bias": "model.safetensors",
74
+ "audio_codec.decoder.blocks.1.block_6.conv2.weight": "model.safetensors",
75
+ "audio_codec.decoder.blocks.1.block_7.conv1.bias": "model.safetensors",
76
+ "audio_codec.decoder.blocks.1.block_7.conv1.weight": "model.safetensors",
77
+ "audio_codec.decoder.blocks.1.block_7.conv2.bias": "model.safetensors",
78
+ "audio_codec.decoder.blocks.1.block_7.conv2.weight": "model.safetensors",
79
+ "audio_codec.decoder.blocks.1.block_8.act1.alpha": "model.safetensors",
80
+ "audio_codec.decoder.blocks.1.block_8.act2.alpha": "model.safetensors",
81
+ "audio_codec.decoder.blocks.1.block_8.conv1.bias": "model.safetensors",
82
+ "audio_codec.decoder.blocks.1.block_8.conv1.weight_g": "model.safetensors",
83
+ "audio_codec.decoder.blocks.1.block_8.conv1.weight_v": "model.safetensors",
84
+ "audio_codec.decoder.blocks.1.block_8.conv2.bias": "model.safetensors",
85
+ "audio_codec.decoder.blocks.1.block_8.conv2.weight_g": "model.safetensors",
86
+ "audio_codec.decoder.blocks.1.block_8.conv2.weight_v": "model.safetensors",
87
+ "audio_codec.decoder.blocks.2.block_0.alpha": "model.safetensors",
88
+ "audio_codec.decoder.blocks.2.block_1.bias": "model.safetensors",
89
+ "audio_codec.decoder.blocks.2.block_1.weight_g": "model.safetensors",
90
+ "audio_codec.decoder.blocks.2.block_1.weight_v": "model.safetensors",
91
+ "audio_codec.decoder.blocks.2.block_11.bias": "model.safetensors",
92
+ "audio_codec.decoder.blocks.2.block_11.weight": "model.safetensors",
93
+ "audio_codec.decoder.blocks.2.block_3.bias": "model.safetensors",
94
+ "audio_codec.decoder.blocks.2.block_3.weight": "model.safetensors",
95
+ "audio_codec.decoder.blocks.2.block_4.act1.alpha": "model.safetensors",
96
+ "audio_codec.decoder.blocks.2.block_4.act2.alpha": "model.safetensors",
97
+ "audio_codec.decoder.blocks.2.block_4.conv1.bias": "model.safetensors",
98
+ "audio_codec.decoder.blocks.2.block_4.conv1.weight_g": "model.safetensors",
99
+ "audio_codec.decoder.blocks.2.block_4.conv1.weight_v": "model.safetensors",
100
+ "audio_codec.decoder.blocks.2.block_4.conv2.bias": "model.safetensors",
101
+ "audio_codec.decoder.blocks.2.block_4.conv2.weight_g": "model.safetensors",
102
+ "audio_codec.decoder.blocks.2.block_4.conv2.weight_v": "model.safetensors",
103
+ "audio_codec.decoder.blocks.2.block_5.act1.alpha": "model.safetensors",
104
+ "audio_codec.decoder.blocks.2.block_5.act2.alpha": "model.safetensors",
105
+ "audio_codec.decoder.blocks.2.block_5.conv1.bias": "model.safetensors",
106
+ "audio_codec.decoder.blocks.2.block_5.conv1.weight_g": "model.safetensors",
107
+ "audio_codec.decoder.blocks.2.block_5.conv1.weight_v": "model.safetensors",
108
+ "audio_codec.decoder.blocks.2.block_5.conv2.bias": "model.safetensors",
109
+ "audio_codec.decoder.blocks.2.block_5.conv2.weight_g": "model.safetensors",
110
+ "audio_codec.decoder.blocks.2.block_5.conv2.weight_v": "model.safetensors",
111
+ "audio_codec.decoder.blocks.2.block_6.conv1.bias": "model.safetensors",
112
+ "audio_codec.decoder.blocks.2.block_6.conv1.weight": "model.safetensors",
113
+ "audio_codec.decoder.blocks.2.block_6.conv2.bias": "model.safetensors",
114
+ "audio_codec.decoder.blocks.2.block_6.conv2.weight": "model.safetensors",
115
+ "audio_codec.decoder.blocks.2.block_7.conv1.bias": "model.safetensors",
116
+ "audio_codec.decoder.blocks.2.block_7.conv1.weight": "model.safetensors",
117
+ "audio_codec.decoder.blocks.2.block_7.conv2.bias": "model.safetensors",
118
+ "audio_codec.decoder.blocks.2.block_7.conv2.weight": "model.safetensors",
119
+ "audio_codec.decoder.blocks.2.block_8.act1.alpha": "model.safetensors",
120
+ "audio_codec.decoder.blocks.2.block_8.act2.alpha": "model.safetensors",
121
+ "audio_codec.decoder.blocks.2.block_8.conv1.bias": "model.safetensors",
122
+ "audio_codec.decoder.blocks.2.block_8.conv1.weight_g": "model.safetensors",
123
+ "audio_codec.decoder.blocks.2.block_8.conv1.weight_v": "model.safetensors",
124
+ "audio_codec.decoder.blocks.2.block_8.conv2.bias": "model.safetensors",
125
+ "audio_codec.decoder.blocks.2.block_8.conv2.weight_g": "model.safetensors",
126
+ "audio_codec.decoder.blocks.2.block_8.conv2.weight_v": "model.safetensors",
127
+ "audio_codec.decoder.blocks.3.block_0.alpha": "model.safetensors",
128
+ "audio_codec.decoder.blocks.3.block_1.bias": "model.safetensors",
129
+ "audio_codec.decoder.blocks.3.block_1.weight_g": "model.safetensors",
130
+ "audio_codec.decoder.blocks.3.block_1.weight_v": "model.safetensors",
131
+ "audio_codec.decoder.blocks.3.block_11.bias": "model.safetensors",
132
+ "audio_codec.decoder.blocks.3.block_11.weight": "model.safetensors",
133
+ "audio_codec.decoder.blocks.3.block_3.bias": "model.safetensors",
134
+ "audio_codec.decoder.blocks.3.block_3.weight": "model.safetensors",
135
+ "audio_codec.decoder.blocks.3.block_4.act1.alpha": "model.safetensors",
136
+ "audio_codec.decoder.blocks.3.block_4.act2.alpha": "model.safetensors",
137
+ "audio_codec.decoder.blocks.3.block_4.conv1.bias": "model.safetensors",
138
+ "audio_codec.decoder.blocks.3.block_4.conv1.weight_g": "model.safetensors",
139
+ "audio_codec.decoder.blocks.3.block_4.conv1.weight_v": "model.safetensors",
140
+ "audio_codec.decoder.blocks.3.block_4.conv2.bias": "model.safetensors",
141
+ "audio_codec.decoder.blocks.3.block_4.conv2.weight_g": "model.safetensors",
142
+ "audio_codec.decoder.blocks.3.block_4.conv2.weight_v": "model.safetensors",
143
+ "audio_codec.decoder.blocks.3.block_5.act1.alpha": "model.safetensors",
144
+ "audio_codec.decoder.blocks.3.block_5.act2.alpha": "model.safetensors",
145
+ "audio_codec.decoder.blocks.3.block_5.conv1.bias": "model.safetensors",
146
+ "audio_codec.decoder.blocks.3.block_5.conv1.weight_g": "model.safetensors",
147
+ "audio_codec.decoder.blocks.3.block_5.conv1.weight_v": "model.safetensors",
148
+ "audio_codec.decoder.blocks.3.block_5.conv2.bias": "model.safetensors",
149
+ "audio_codec.decoder.blocks.3.block_5.conv2.weight_g": "model.safetensors",
150
+ "audio_codec.decoder.blocks.3.block_5.conv2.weight_v": "model.safetensors",
151
+ "audio_codec.decoder.blocks.3.block_6.conv1.bias": "model.safetensors",
152
+ "audio_codec.decoder.blocks.3.block_6.conv1.weight": "model.safetensors",
153
+ "audio_codec.decoder.blocks.3.block_6.conv2.bias": "model.safetensors",
154
+ "audio_codec.decoder.blocks.3.block_6.conv2.weight": "model.safetensors",
155
+ "audio_codec.decoder.blocks.3.block_7.conv1.bias": "model.safetensors",
156
+ "audio_codec.decoder.blocks.3.block_7.conv1.weight": "model.safetensors",
157
+ "audio_codec.decoder.blocks.3.block_7.conv2.bias": "model.safetensors",
158
+ "audio_codec.decoder.blocks.3.block_7.conv2.weight": "model.safetensors",
159
+ "audio_codec.decoder.blocks.3.block_8.act1.alpha": "model.safetensors",
160
+ "audio_codec.decoder.blocks.3.block_8.act2.alpha": "model.safetensors",
161
+ "audio_codec.decoder.blocks.3.block_8.conv1.bias": "model.safetensors",
162
+ "audio_codec.decoder.blocks.3.block_8.conv1.weight_g": "model.safetensors",
163
+ "audio_codec.decoder.blocks.3.block_8.conv1.weight_v": "model.safetensors",
164
+ "audio_codec.decoder.blocks.3.block_8.conv2.bias": "model.safetensors",
165
+ "audio_codec.decoder.blocks.3.block_8.conv2.weight_g": "model.safetensors",
166
+ "audio_codec.decoder.blocks.3.block_8.conv2.weight_v": "model.safetensors",
167
+ "audio_codec.decoder.conv_in.bias": "model.safetensors",
168
+ "audio_codec.decoder.conv_in.weight_g": "model.safetensors",
169
+ "audio_codec.decoder.conv_in.weight_v": "model.safetensors",
170
+ "audio_codec.decoder.conv_out.bias": "model.safetensors",
171
+ "audio_codec.decoder.conv_out.weight_g": "model.safetensors",
172
+ "audio_codec.decoder.conv_out.weight_v": "model.safetensors",
173
+ "audio_codec.decoder.snake_out.alpha": "model.safetensors",
174
+ "audio_codec.decoder.wm_model.decoder_block.post_1.bias": "model.safetensors",
175
+ "audio_codec.decoder.wm_model.decoder_block.post_1.weight": "model.safetensors",
176
+ "audio_codec.decoder.wm_model.decoder_block.pre_0.bias": "model.safetensors",
177
+ "audio_codec.decoder.wm_model.decoder_block.pre_0.weight": "model.safetensors",
178
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.0.Wh": "model.safetensors",
179
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.0.Wx": "model.safetensors",
180
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.0.bias": "model.safetensors",
181
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.1.Wh": "model.safetensors",
182
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.1.Wx": "model.safetensors",
183
+ "audio_codec.decoder.wm_model.decoder_block.pre_1.lstm.layers.1.bias": "model.safetensors",
184
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.0.Wh": "model.safetensors",
185
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.0.Wx": "model.safetensors",
186
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.0.bias": "model.safetensors",
187
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.1.Wh": "model.safetensors",
188
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.1.Wx": "model.safetensors",
189
+ "audio_codec.decoder.wm_model.encoder_block.post_0.lstm.layers.1.bias": "model.safetensors",
190
+ "audio_codec.decoder.wm_model.encoder_block.post_2.bias": "model.safetensors",
191
+ "audio_codec.decoder.wm_model.encoder_block.post_2.weight": "model.safetensors",
192
+ "audio_codec.decoder.wm_model.encoder_block.pre_3.bias": "model.safetensors",
193
+ "audio_codec.decoder.wm_model.encoder_block.pre_3.weight": "model.safetensors",
194
+ "audio_codec.decoder.wm_model.msg_processor.msg_processor.weight": "model.safetensors",
195
+ "audio_codec.encoder.blocks.0.conv.bias": "model.safetensors",
196
+ "audio_codec.encoder.blocks.0.conv.weight_g": "model.safetensors",
197
+ "audio_codec.encoder.blocks.0.conv.weight_v": "model.safetensors",
198
+ "audio_codec.encoder.blocks.0.res1.act1.alpha": "model.safetensors",
199
+ "audio_codec.encoder.blocks.0.res1.act2.alpha": "model.safetensors",
200
+ "audio_codec.encoder.blocks.0.res1.conv1.bias": "model.safetensors",
201
+ "audio_codec.encoder.blocks.0.res1.conv1.weight_g": "model.safetensors",
202
+ "audio_codec.encoder.blocks.0.res1.conv1.weight_v": "model.safetensors",
203
+ "audio_codec.encoder.blocks.0.res1.conv2.bias": "model.safetensors",
204
+ "audio_codec.encoder.blocks.0.res1.conv2.weight_g": "model.safetensors",
205
+ "audio_codec.encoder.blocks.0.res1.conv2.weight_v": "model.safetensors",
206
+ "audio_codec.encoder.blocks.0.res2.act1.alpha": "model.safetensors",
207
+ "audio_codec.encoder.blocks.0.res2.act2.alpha": "model.safetensors",
208
+ "audio_codec.encoder.blocks.0.res2.conv1.bias": "model.safetensors",
209
+ "audio_codec.encoder.blocks.0.res2.conv1.weight_g": "model.safetensors",
210
+ "audio_codec.encoder.blocks.0.res2.conv1.weight_v": "model.safetensors",
211
+ "audio_codec.encoder.blocks.0.res2.conv2.bias": "model.safetensors",
212
+ "audio_codec.encoder.blocks.0.res2.conv2.weight_g": "model.safetensors",
213
+ "audio_codec.encoder.blocks.0.res2.conv2.weight_v": "model.safetensors",
214
+ "audio_codec.encoder.blocks.0.res3.act1.alpha": "model.safetensors",
215
+ "audio_codec.encoder.blocks.0.res3.act2.alpha": "model.safetensors",
216
+ "audio_codec.encoder.blocks.0.res3.conv1.bias": "model.safetensors",
217
+ "audio_codec.encoder.blocks.0.res3.conv1.weight_g": "model.safetensors",
218
+ "audio_codec.encoder.blocks.0.res3.conv1.weight_v": "model.safetensors",
219
+ "audio_codec.encoder.blocks.0.res3.conv2.bias": "model.safetensors",
220
+ "audio_codec.encoder.blocks.0.res3.conv2.weight_g": "model.safetensors",
221
+ "audio_codec.encoder.blocks.0.res3.conv2.weight_v": "model.safetensors",
222
+ "audio_codec.encoder.blocks.0.snake.alpha": "model.safetensors",
223
+ "audio_codec.encoder.blocks.1.conv.bias": "model.safetensors",
224
+ "audio_codec.encoder.blocks.1.conv.weight_g": "model.safetensors",
225
+ "audio_codec.encoder.blocks.1.conv.weight_v": "model.safetensors",
226
+ "audio_codec.encoder.blocks.1.res1.act1.alpha": "model.safetensors",
227
+ "audio_codec.encoder.blocks.1.res1.act2.alpha": "model.safetensors",
228
+ "audio_codec.encoder.blocks.1.res1.conv1.bias": "model.safetensors",
229
+ "audio_codec.encoder.blocks.1.res1.conv1.weight_g": "model.safetensors",
230
+ "audio_codec.encoder.blocks.1.res1.conv1.weight_v": "model.safetensors",
231
+ "audio_codec.encoder.blocks.1.res1.conv2.bias": "model.safetensors",
232
+ "audio_codec.encoder.blocks.1.res1.conv2.weight_g": "model.safetensors",
233
+ "audio_codec.encoder.blocks.1.res1.conv2.weight_v": "model.safetensors",
234
+ "audio_codec.encoder.blocks.1.res2.act1.alpha": "model.safetensors",
235
+ "audio_codec.encoder.blocks.1.res2.act2.alpha": "model.safetensors",
236
+ "audio_codec.encoder.blocks.1.res2.conv1.bias": "model.safetensors",
237
+ "audio_codec.encoder.blocks.1.res2.conv1.weight_g": "model.safetensors",
238
+ "audio_codec.encoder.blocks.1.res2.conv1.weight_v": "model.safetensors",
239
+ "audio_codec.encoder.blocks.1.res2.conv2.bias": "model.safetensors",
240
+ "audio_codec.encoder.blocks.1.res2.conv2.weight_g": "model.safetensors",
241
+ "audio_codec.encoder.blocks.1.res2.conv2.weight_v": "model.safetensors",
242
+ "audio_codec.encoder.blocks.1.res3.act1.alpha": "model.safetensors",
243
+ "audio_codec.encoder.blocks.1.res3.act2.alpha": "model.safetensors",
244
+ "audio_codec.encoder.blocks.1.res3.conv1.bias": "model.safetensors",
245
+ "audio_codec.encoder.blocks.1.res3.conv1.weight_g": "model.safetensors",
246
+ "audio_codec.encoder.blocks.1.res3.conv1.weight_v": "model.safetensors",
247
+ "audio_codec.encoder.blocks.1.res3.conv2.bias": "model.safetensors",
248
+ "audio_codec.encoder.blocks.1.res3.conv2.weight_g": "model.safetensors",
249
+ "audio_codec.encoder.blocks.1.res3.conv2.weight_v": "model.safetensors",
250
+ "audio_codec.encoder.blocks.1.snake.alpha": "model.safetensors",
251
+ "audio_codec.encoder.blocks.2.conv.bias": "model.safetensors",
252
+ "audio_codec.encoder.blocks.2.conv.weight_g": "model.safetensors",
253
+ "audio_codec.encoder.blocks.2.conv.weight_v": "model.safetensors",
254
+ "audio_codec.encoder.blocks.2.res1.act1.alpha": "model.safetensors",
255
+ "audio_codec.encoder.blocks.2.res1.act2.alpha": "model.safetensors",
256
+ "audio_codec.encoder.blocks.2.res1.conv1.bias": "model.safetensors",
257
+ "audio_codec.encoder.blocks.2.res1.conv1.weight_g": "model.safetensors",
258
+ "audio_codec.encoder.blocks.2.res1.conv1.weight_v": "model.safetensors",
259
+ "audio_codec.encoder.blocks.2.res1.conv2.bias": "model.safetensors",
260
+ "audio_codec.encoder.blocks.2.res1.conv2.weight_g": "model.safetensors",
261
+ "audio_codec.encoder.blocks.2.res1.conv2.weight_v": "model.safetensors",
262
+ "audio_codec.encoder.blocks.2.res2.act1.alpha": "model.safetensors",
263
+ "audio_codec.encoder.blocks.2.res2.act2.alpha": "model.safetensors",
264
+ "audio_codec.encoder.blocks.2.res2.conv1.bias": "model.safetensors",
265
+ "audio_codec.encoder.blocks.2.res2.conv1.weight_g": "model.safetensors",
266
+ "audio_codec.encoder.blocks.2.res2.conv1.weight_v": "model.safetensors",
267
+ "audio_codec.encoder.blocks.2.res2.conv2.bias": "model.safetensors",
268
+ "audio_codec.encoder.blocks.2.res2.conv2.weight_g": "model.safetensors",
269
+ "audio_codec.encoder.blocks.2.res2.conv2.weight_v": "model.safetensors",
270
+ "audio_codec.encoder.blocks.2.res3.act1.alpha": "model.safetensors",
271
+ "audio_codec.encoder.blocks.2.res3.act2.alpha": "model.safetensors",
272
+ "audio_codec.encoder.blocks.2.res3.conv1.bias": "model.safetensors",
273
+ "audio_codec.encoder.blocks.2.res3.conv1.weight_g": "model.safetensors",
274
+ "audio_codec.encoder.blocks.2.res3.conv1.weight_v": "model.safetensors",
275
+ "audio_codec.encoder.blocks.2.res3.conv2.bias": "model.safetensors",
276
+ "audio_codec.encoder.blocks.2.res3.conv2.weight_g": "model.safetensors",
277
+ "audio_codec.encoder.blocks.2.res3.conv2.weight_v": "model.safetensors",
278
+ "audio_codec.encoder.blocks.2.snake.alpha": "model.safetensors",
279
+ "audio_codec.encoder.blocks.3.conv.bias": "model.safetensors",
280
+ "audio_codec.encoder.blocks.3.conv.weight_g": "model.safetensors",
281
+ "audio_codec.encoder.blocks.3.conv.weight_v": "model.safetensors",
282
+ "audio_codec.encoder.blocks.3.res1.act1.alpha": "model.safetensors",
283
+ "audio_codec.encoder.blocks.3.res1.act2.alpha": "model.safetensors",
284
+ "audio_codec.encoder.blocks.3.res1.conv1.bias": "model.safetensors",
285
+ "audio_codec.encoder.blocks.3.res1.conv1.weight_g": "model.safetensors",
286
+ "audio_codec.encoder.blocks.3.res1.conv1.weight_v": "model.safetensors",
287
+ "audio_codec.encoder.blocks.3.res1.conv2.bias": "model.safetensors",
288
+ "audio_codec.encoder.blocks.3.res1.conv2.weight_g": "model.safetensors",
289
+ "audio_codec.encoder.blocks.3.res1.conv2.weight_v": "model.safetensors",
290
+ "audio_codec.encoder.blocks.3.res2.act1.alpha": "model.safetensors",
291
+ "audio_codec.encoder.blocks.3.res2.act2.alpha": "model.safetensors",
292
+ "audio_codec.encoder.blocks.3.res2.conv1.bias": "model.safetensors",
293
+ "audio_codec.encoder.blocks.3.res2.conv1.weight_g": "model.safetensors",
294
+ "audio_codec.encoder.blocks.3.res2.conv1.weight_v": "model.safetensors",
295
+ "audio_codec.encoder.blocks.3.res2.conv2.bias": "model.safetensors",
296
+ "audio_codec.encoder.blocks.3.res2.conv2.weight_g": "model.safetensors",
297
+ "audio_codec.encoder.blocks.3.res2.conv2.weight_v": "model.safetensors",
298
+ "audio_codec.encoder.blocks.3.res3.act1.alpha": "model.safetensors",
299
+ "audio_codec.encoder.blocks.3.res3.act2.alpha": "model.safetensors",
300
+ "audio_codec.encoder.blocks.3.res3.conv1.bias": "model.safetensors",
301
+ "audio_codec.encoder.blocks.3.res3.conv1.weight_g": "model.safetensors",
302
+ "audio_codec.encoder.blocks.3.res3.conv1.weight_v": "model.safetensors",
303
+ "audio_codec.encoder.blocks.3.res3.conv2.bias": "model.safetensors",
304
+ "audio_codec.encoder.blocks.3.res3.conv2.weight_g": "model.safetensors",
305
+ "audio_codec.encoder.blocks.3.res3.conv2.weight_v": "model.safetensors",
306
+ "audio_codec.encoder.blocks.3.snake.alpha": "model.safetensors",
307
+ "audio_codec.encoder.conv_in.bias": "model.safetensors",
308
+ "audio_codec.encoder.conv_in.weight_g": "model.safetensors",
309
+ "audio_codec.encoder.conv_in.weight_v": "model.safetensors",
310
+ "audio_codec.encoder.conv_out.bias": "model.safetensors",
311
+ "audio_codec.encoder.conv_out.weight_g": "model.safetensors",
312
+ "audio_codec.encoder.conv_out.weight_v": "model.safetensors",
313
+ "audio_codec.encoder.snake_out.alpha": "model.safetensors",
314
+ "audio_codec.quantizer_in_proj.bias": "model.safetensors",
315
+ "audio_codec.quantizer_in_proj.weight_g": "model.safetensors",
316
+ "audio_codec.quantizer_in_proj.weight_v": "model.safetensors",
317
+ "audio_codec.quantizer_out_proj.bias": "model.safetensors",
318
+ "audio_codec.quantizer_out_proj.weight_g": "model.safetensors",
319
+ "audio_codec.quantizer_out_proj.weight_v": "model.safetensors",
320
+ "embed_anchors.embed.weight": "model.safetensors",
321
+ "embed_anchors.gate": "model.safetensors",
322
+ "embed_anchors.proj.weight": "model.safetensors",
323
+ "memory_proj.bias": "model.safetensors",
324
+ "memory_proj.weight": "model.safetensors",
325
+ "proj.bias": "model.safetensors",
326
+ "proj.weight": "model.safetensors",
327
+ "transformer.final_layer_scale_shift_table": "model.safetensors",
328
+ "transformer.layers.0.attention.k_norm.weight": "model.safetensors",
329
+ "transformer.layers.0.attention.q_norm.weight": "model.safetensors",
330
+ "transformer.layers.0.attention.wk.weight": "model.safetensors",
331
+ "transformer.layers.0.attention.wo.weight": "model.safetensors",
332
+ "transformer.layers.0.attention.wq.weight": "model.safetensors",
333
+ "transformer.layers.0.attention.wv.weight": "model.safetensors",
334
+ "transformer.layers.0.attention_norm.weight": "model.safetensors",
335
+ "transformer.layers.0.cross_attention.k_norm.weight": "model.safetensors",
336
+ "transformer.layers.0.cross_attention.q_norm.weight": "model.safetensors",
337
+ "transformer.layers.0.cross_attention.wk.weight": "model.safetensors",
338
+ "transformer.layers.0.cross_attention.wo.weight": "model.safetensors",
339
+ "transformer.layers.0.cross_attention.wq.weight": "model.safetensors",
340
+ "transformer.layers.0.cross_attention.wv.weight": "model.safetensors",
341
+ "transformer.layers.0.feed_forward.w1.weight": "model.safetensors",
342
+ "transformer.layers.0.feed_forward.w2.weight": "model.safetensors",
343
+ "transformer.layers.0.feed_forward.w3.weight": "model.safetensors",
344
+ "transformer.layers.0.ffn_norm.weight": "model.safetensors",
345
+ "transformer.layers.0.scale_shift_table": "model.safetensors",
346
+ "transformer.layers.1.attention.k_norm.weight": "model.safetensors",
347
+ "transformer.layers.1.attention.q_norm.weight": "model.safetensors",
348
+ "transformer.layers.1.attention.wk.weight": "model.safetensors",
349
+ "transformer.layers.1.attention.wo.weight": "model.safetensors",
350
+ "transformer.layers.1.attention.wq.weight": "model.safetensors",
351
+ "transformer.layers.1.attention.wv.weight": "model.safetensors",
352
+ "transformer.layers.1.attention_norm.weight": "model.safetensors",
353
+ "transformer.layers.1.cross_attention.k_norm.weight": "model.safetensors",
354
+ "transformer.layers.1.cross_attention.q_norm.weight": "model.safetensors",
355
+ "transformer.layers.1.cross_attention.wk.weight": "model.safetensors",
356
+ "transformer.layers.1.cross_attention.wo.weight": "model.safetensors",
357
+ "transformer.layers.1.cross_attention.wq.weight": "model.safetensors",
358
+ "transformer.layers.1.cross_attention.wv.weight": "model.safetensors",
359
+ "transformer.layers.1.feed_forward.w1.weight": "model.safetensors",
360
+ "transformer.layers.1.feed_forward.w2.weight": "model.safetensors",
361
+ "transformer.layers.1.feed_forward.w3.weight": "model.safetensors",
362
+ "transformer.layers.1.ffn_norm.weight": "model.safetensors",
363
+ "transformer.layers.1.scale_shift_table": "model.safetensors",
364
+ "transformer.layers.10.attention.k_norm.weight": "model.safetensors",
365
+ "transformer.layers.10.attention.q_norm.weight": "model.safetensors",
366
+ "transformer.layers.10.attention.wk.weight": "model.safetensors",
367
+ "transformer.layers.10.attention.wo.weight": "model.safetensors",
368
+ "transformer.layers.10.attention.wq.weight": "model.safetensors",
369
+ "transformer.layers.10.attention.wv.weight": "model.safetensors",
370
+ "transformer.layers.10.attention_norm.weight": "model.safetensors",
371
+ "transformer.layers.10.cross_attention.k_norm.weight": "model.safetensors",
372
+ "transformer.layers.10.cross_attention.q_norm.weight": "model.safetensors",
373
+ "transformer.layers.10.cross_attention.wk.weight": "model.safetensors",
374
+ "transformer.layers.10.cross_attention.wo.weight": "model.safetensors",
375
+ "transformer.layers.10.cross_attention.wq.weight": "model.safetensors",
376
+ "transformer.layers.10.cross_attention.wv.weight": "model.safetensors",
377
+ "transformer.layers.10.feed_forward.w1.weight": "model.safetensors",
378
+ "transformer.layers.10.feed_forward.w2.weight": "model.safetensors",
379
+ "transformer.layers.10.feed_forward.w3.weight": "model.safetensors",
380
+ "transformer.layers.10.ffn_norm.weight": "model.safetensors",
381
+ "transformer.layers.10.scale_shift_table": "model.safetensors",
382
+ "transformer.layers.11.attention.k_norm.weight": "model.safetensors",
383
+ "transformer.layers.11.attention.q_norm.weight": "model.safetensors",
384
+ "transformer.layers.11.attention.wk.weight": "model.safetensors",
385
+ "transformer.layers.11.attention.wo.weight": "model.safetensors",
386
+ "transformer.layers.11.attention.wq.weight": "model.safetensors",
387
+ "transformer.layers.11.attention.wv.weight": "model.safetensors",
388
+ "transformer.layers.11.attention_norm.weight": "model.safetensors",
389
+ "transformer.layers.11.cross_attention.k_norm.weight": "model.safetensors",
390
+ "transformer.layers.11.cross_attention.q_norm.weight": "model.safetensors",
391
+ "transformer.layers.11.cross_attention.wk.weight": "model.safetensors",
392
+ "transformer.layers.11.cross_attention.wo.weight": "model.safetensors",
393
+ "transformer.layers.11.cross_attention.wq.weight": "model.safetensors",
394
+ "transformer.layers.11.cross_attention.wv.weight": "model.safetensors",
395
+ "transformer.layers.11.feed_forward.w1.weight": "model.safetensors",
396
+ "transformer.layers.11.feed_forward.w2.weight": "model.safetensors",
397
+ "transformer.layers.11.feed_forward.w3.weight": "model.safetensors",
398
+ "transformer.layers.11.ffn_norm.weight": "model.safetensors",
399
+ "transformer.layers.11.scale_shift_table": "model.safetensors",
400
+ "transformer.layers.2.attention.k_norm.weight": "model.safetensors",
401
+ "transformer.layers.2.attention.q_norm.weight": "model.safetensors",
402
+ "transformer.layers.2.attention.wk.weight": "model.safetensors",
403
+ "transformer.layers.2.attention.wo.weight": "model.safetensors",
404
+ "transformer.layers.2.attention.wq.weight": "model.safetensors",
405
+ "transformer.layers.2.attention.wv.weight": "model.safetensors",
406
+ "transformer.layers.2.attention_norm.weight": "model.safetensors",
407
+ "transformer.layers.2.cross_attention.k_norm.weight": "model.safetensors",
408
+ "transformer.layers.2.cross_attention.q_norm.weight": "model.safetensors",
409
+ "transformer.layers.2.cross_attention.wk.weight": "model.safetensors",
410
+ "transformer.layers.2.cross_attention.wo.weight": "model.safetensors",
411
+ "transformer.layers.2.cross_attention.wq.weight": "model.safetensors",
412
+ "transformer.layers.2.cross_attention.wv.weight": "model.safetensors",
413
+ "transformer.layers.2.feed_forward.w1.weight": "model.safetensors",
414
+ "transformer.layers.2.feed_forward.w2.weight": "model.safetensors",
415
+ "transformer.layers.2.feed_forward.w3.weight": "model.safetensors",
416
+ "transformer.layers.2.ffn_norm.weight": "model.safetensors",
417
+ "transformer.layers.2.scale_shift_table": "model.safetensors",
418
+ "transformer.layers.3.attention.k_norm.weight": "model.safetensors",
419
+ "transformer.layers.3.attention.q_norm.weight": "model.safetensors",
420
+ "transformer.layers.3.attention.wk.weight": "model.safetensors",
421
+ "transformer.layers.3.attention.wo.weight": "model.safetensors",
422
+ "transformer.layers.3.attention.wq.weight": "model.safetensors",
423
+ "transformer.layers.3.attention.wv.weight": "model.safetensors",
424
+ "transformer.layers.3.attention_norm.weight": "model.safetensors",
425
+ "transformer.layers.3.cross_attention.k_norm.weight": "model.safetensors",
426
+ "transformer.layers.3.cross_attention.q_norm.weight": "model.safetensors",
427
+ "transformer.layers.3.cross_attention.wk.weight": "model.safetensors",
428
+ "transformer.layers.3.cross_attention.wo.weight": "model.safetensors",
429
+ "transformer.layers.3.cross_attention.wq.weight": "model.safetensors",
430
+ "transformer.layers.3.cross_attention.wv.weight": "model.safetensors",
431
+ "transformer.layers.3.feed_forward.w1.weight": "model.safetensors",
432
+ "transformer.layers.3.feed_forward.w2.weight": "model.safetensors",
433
+ "transformer.layers.3.feed_forward.w3.weight": "model.safetensors",
434
+ "transformer.layers.3.ffn_norm.weight": "model.safetensors",
435
+ "transformer.layers.3.scale_shift_table": "model.safetensors",
436
+ "transformer.layers.4.attention.k_norm.weight": "model.safetensors",
437
+ "transformer.layers.4.attention.q_norm.weight": "model.safetensors",
438
+ "transformer.layers.4.attention.wk.weight": "model.safetensors",
439
+ "transformer.layers.4.attention.wo.weight": "model.safetensors",
440
+ "transformer.layers.4.attention.wq.weight": "model.safetensors",
441
+ "transformer.layers.4.attention.wv.weight": "model.safetensors",
442
+ "transformer.layers.4.attention_norm.weight": "model.safetensors",
443
+ "transformer.layers.4.cross_attention.k_norm.weight": "model.safetensors",
444
+ "transformer.layers.4.cross_attention.q_norm.weight": "model.safetensors",
445
+ "transformer.layers.4.cross_attention.wk.weight": "model.safetensors",
446
+ "transformer.layers.4.cross_attention.wo.weight": "model.safetensors",
447
+ "transformer.layers.4.cross_attention.wq.weight": "model.safetensors",
448
+ "transformer.layers.4.cross_attention.wv.weight": "model.safetensors",
449
+ "transformer.layers.4.feed_forward.w1.weight": "model.safetensors",
450
+ "transformer.layers.4.feed_forward.w2.weight": "model.safetensors",
451
+ "transformer.layers.4.feed_forward.w3.weight": "model.safetensors",
452
+ "transformer.layers.4.ffn_norm.weight": "model.safetensors",
453
+ "transformer.layers.4.scale_shift_table": "model.safetensors",
454
+ "transformer.layers.5.attention.k_norm.weight": "model.safetensors",
455
+ "transformer.layers.5.attention.q_norm.weight": "model.safetensors",
456
+ "transformer.layers.5.attention.wk.weight": "model.safetensors",
457
+ "transformer.layers.5.attention.wo.weight": "model.safetensors",
458
+ "transformer.layers.5.attention.wq.weight": "model.safetensors",
459
+ "transformer.layers.5.attention.wv.weight": "model.safetensors",
460
+ "transformer.layers.5.attention_norm.weight": "model.safetensors",
461
+ "transformer.layers.5.cross_attention.k_norm.weight": "model.safetensors",
462
+ "transformer.layers.5.cross_attention.q_norm.weight": "model.safetensors",
463
+ "transformer.layers.5.cross_attention.wk.weight": "model.safetensors",
464
+ "transformer.layers.5.cross_attention.wo.weight": "model.safetensors",
465
+ "transformer.layers.5.cross_attention.wq.weight": "model.safetensors",
466
+ "transformer.layers.5.cross_attention.wv.weight": "model.safetensors",
467
+ "transformer.layers.5.feed_forward.w1.weight": "model.safetensors",
468
+ "transformer.layers.5.feed_forward.w2.weight": "model.safetensors",
469
+ "transformer.layers.5.feed_forward.w3.weight": "model.safetensors",
470
+ "transformer.layers.5.ffn_norm.weight": "model.safetensors",
471
+ "transformer.layers.5.scale_shift_table": "model.safetensors",
472
+ "transformer.layers.6.attention.k_norm.weight": "model.safetensors",
473
+ "transformer.layers.6.attention.q_norm.weight": "model.safetensors",
474
+ "transformer.layers.6.attention.wk.weight": "model.safetensors",
475
+ "transformer.layers.6.attention.wo.weight": "model.safetensors",
476
+ "transformer.layers.6.attention.wq.weight": "model.safetensors",
477
+ "transformer.layers.6.attention.wv.weight": "model.safetensors",
478
+ "transformer.layers.6.attention_norm.weight": "model.safetensors",
479
+ "transformer.layers.6.cross_attention.k_norm.weight": "model.safetensors",
480
+ "transformer.layers.6.cross_attention.q_norm.weight": "model.safetensors",
481
+ "transformer.layers.6.cross_attention.wk.weight": "model.safetensors",
482
+ "transformer.layers.6.cross_attention.wo.weight": "model.safetensors",
483
+ "transformer.layers.6.cross_attention.wq.weight": "model.safetensors",
484
+ "transformer.layers.6.cross_attention.wv.weight": "model.safetensors",
485
+ "transformer.layers.6.feed_forward.w1.weight": "model.safetensors",
486
+ "transformer.layers.6.feed_forward.w2.weight": "model.safetensors",
487
+ "transformer.layers.6.feed_forward.w3.weight": "model.safetensors",
488
+ "transformer.layers.6.ffn_norm.weight": "model.safetensors",
489
+ "transformer.layers.6.scale_shift_table": "model.safetensors",
490
+ "transformer.layers.7.attention.k_norm.weight": "model.safetensors",
491
+ "transformer.layers.7.attention.q_norm.weight": "model.safetensors",
492
+ "transformer.layers.7.attention.wk.weight": "model.safetensors",
493
+ "transformer.layers.7.attention.wo.weight": "model.safetensors",
494
+ "transformer.layers.7.attention.wq.weight": "model.safetensors",
495
+ "transformer.layers.7.attention.wv.weight": "model.safetensors",
496
+ "transformer.layers.7.attention_norm.weight": "model.safetensors",
497
+ "transformer.layers.7.cross_attention.k_norm.weight": "model.safetensors",
498
+ "transformer.layers.7.cross_attention.q_norm.weight": "model.safetensors",
499
+ "transformer.layers.7.cross_attention.wk.weight": "model.safetensors",
500
+ "transformer.layers.7.cross_attention.wo.weight": "model.safetensors",
501
+ "transformer.layers.7.cross_attention.wq.weight": "model.safetensors",
502
+ "transformer.layers.7.cross_attention.wv.weight": "model.safetensors",
503
+ "transformer.layers.7.feed_forward.w1.weight": "model.safetensors",
504
+ "transformer.layers.7.feed_forward.w2.weight": "model.safetensors",
505
+ "transformer.layers.7.feed_forward.w3.weight": "model.safetensors",
506
+ "transformer.layers.7.ffn_norm.weight": "model.safetensors",
507
+ "transformer.layers.7.scale_shift_table": "model.safetensors",
508
+ "transformer.layers.8.attention.k_norm.weight": "model.safetensors",
509
+ "transformer.layers.8.attention.q_norm.weight": "model.safetensors",
510
+ "transformer.layers.8.attention.wk.weight": "model.safetensors",
511
+ "transformer.layers.8.attention.wo.weight": "model.safetensors",
512
+ "transformer.layers.8.attention.wq.weight": "model.safetensors",
513
+ "transformer.layers.8.attention.wv.weight": "model.safetensors",
514
+ "transformer.layers.8.attention_norm.weight": "model.safetensors",
515
+ "transformer.layers.8.cross_attention.k_norm.weight": "model.safetensors",
516
+ "transformer.layers.8.cross_attention.q_norm.weight": "model.safetensors",
517
+ "transformer.layers.8.cross_attention.wk.weight": "model.safetensors",
518
+ "transformer.layers.8.cross_attention.wo.weight": "model.safetensors",
519
+ "transformer.layers.8.cross_attention.wq.weight": "model.safetensors",
520
+ "transformer.layers.8.cross_attention.wv.weight": "model.safetensors",
521
+ "transformer.layers.8.feed_forward.w1.weight": "model.safetensors",
522
+ "transformer.layers.8.feed_forward.w2.weight": "model.safetensors",
523
+ "transformer.layers.8.feed_forward.w3.weight": "model.safetensors",
524
+ "transformer.layers.8.ffn_norm.weight": "model.safetensors",
525
+ "transformer.layers.8.scale_shift_table": "model.safetensors",
526
+ "transformer.layers.9.attention.k_norm.weight": "model.safetensors",
527
+ "transformer.layers.9.attention.q_norm.weight": "model.safetensors",
528
+ "transformer.layers.9.attention.wk.weight": "model.safetensors",
529
+ "transformer.layers.9.attention.wo.weight": "model.safetensors",
530
+ "transformer.layers.9.attention.wq.weight": "model.safetensors",
531
+ "transformer.layers.9.attention.wv.weight": "model.safetensors",
532
+ "transformer.layers.9.attention_norm.weight": "model.safetensors",
533
+ "transformer.layers.9.cross_attention.k_norm.weight": "model.safetensors",
534
+ "transformer.layers.9.cross_attention.q_norm.weight": "model.safetensors",
535
+ "transformer.layers.9.cross_attention.wk.weight": "model.safetensors",
536
+ "transformer.layers.9.cross_attention.wo.weight": "model.safetensors",
537
+ "transformer.layers.9.cross_attention.wq.weight": "model.safetensors",
538
+ "transformer.layers.9.cross_attention.wv.weight": "model.safetensors",
539
+ "transformer.layers.9.feed_forward.w1.weight": "model.safetensors",
540
+ "transformer.layers.9.feed_forward.w2.weight": "model.safetensors",
541
+ "transformer.layers.9.feed_forward.w3.weight": "model.safetensors",
542
+ "transformer.layers.9.ffn_norm.weight": "model.safetensors",
543
+ "transformer.layers.9.scale_shift_table": "model.safetensors",
544
+ "transformer.norm.weight": "model.safetensors",
545
+ "transformer.output.weight": "model.safetensors",
546
+ "transformer.t_block.bias": "model.safetensors",
547
+ "transformer.t_block.weight": "model.safetensors",
548
+ "transformer.t_embedder.projection.w1.weight": "model.safetensors",
549
+ "transformer.t_embedder.projection.w2.weight": "model.safetensors",
550
+ "transformer.t_embedder.projection.w3.weight": "model.safetensors",
551
+ "transformer.x_embedder.block.block1.groupnorm.bias": "model.safetensors",
552
+ "transformer.x_embedder.block.block1.groupnorm.weight": "model.safetensors",
553
+ "transformer.x_embedder.block.block1.project.bias": "model.safetensors",
554
+ "transformer.x_embedder.block.block1.project.weight": "model.safetensors",
555
+ "transformer.x_embedder.block.block2.groupnorm.bias": "model.safetensors",
556
+ "transformer.x_embedder.block.block2.groupnorm.weight": "model.safetensors",
557
+ "transformer.x_embedder.block.block2.project.bias": "model.safetensors",
558
+ "transformer.x_embedder.block.block2.project.weight": "model.safetensors",
559
+ "transformer.y_embedder.projection.w1.weight": "model.safetensors",
560
+ "transformer.y_embedder.projection.w2.weight": "model.safetensors",
561
+ "transformer.y_embedder.projection.w3.weight": "model.safetensors"
562
+ }
563
+ }