Hugging Face's logo

openai
/
jukebox-5b-lyrics

Feature Extraction
Transformers
PyTorch
jukebox
Inference Endpoints
Model card Files Community
11
ArthurZ HF staff commited on
Commit
272aa88
•
1 Parent(s): 2624161

Upload model

Browse files
Files changed (6) hide show
  1. config.json +227 -0
  2. pytorch_model-00001-of-00004.bin +3 -0
  3. pytorch_model-00002-of-00004.bin +3 -0
  4. pytorch_model-00003-of-00004.bin +3 -0
  5. pytorch_model-00004-of-00004.bin +3 -0
  6. pytorch_model.bin.index.json +0 -0
config.json ADDED
@@ -0,0 +1,227 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "jukebox-5b-lyrics",
3
+ "architectures": [
4
+ "JukeboxModel"
5
+ ],
6
+ "cond_c_res": [
7
+ 0,
8
+ 1,
9
+ 1
10
+ ],
11
+ "cond_depth": [
12
+ 3,
13
+ 16,
14
+ 16
15
+ ],
16
+ "cond_dilation_cycle": [
17
+ null,
18
+ 8,
19
+ 8
20
+ ],
21
+ "cond_dilation_growth_rate": [
22
+ 1,
23
+ 3,
24
+ 3
25
+ ],
26
+ "cond_downs_t": [
27
+ 3,
28
+ 2,
29
+ 2
30
+ ],
31
+ "cond_m_conv": 1,
32
+ "cond_res_scale": [
33
+ null,
34
+ true,
35
+ false
36
+ ],
37
+ "cond_strides_t": [
38
+ 2,
39
+ 2,
40
+ 2
41
+ ],
42
+ "cond_width": [
43
+ 128,
44
+ 1024,
45
+ 1024
46
+ ],
47
+ "cond_zero_out": false,
48
+ "copy_input": false,
49
+ "fp16_params": true,
50
+ "hop_fraction": [
51
+ 0.125,
52
+ 0.5,
53
+ 0.5
54
+ ],
55
+ "init_std": 0.2,
56
+ "lyric_conditioning": [
57
+ true,
58
+ false,
59
+ false
60
+ ],
61
+ "max_duration": 600.0,
62
+ "max_nb_genres": 1,
63
+ "merged_decoder": [
64
+ true,
65
+ false,
66
+ false
67
+ ],
68
+ "metadata_conditioning": true,
69
+ "metadata_dims": [
70
+ [
71
+ 120,
72
+ 4111
73
+ ],
74
+ [
75
+ 120,
76
+ 4111
77
+ ],
78
+ [
79
+ 120,
80
+ 4111
81
+ ]
82
+ ],
83
+ "min_duration": 23.8,
84
+ "model_type": "jukebox",
85
+ "nb_priors": 3,
86
+ "nb_relevant_lyric_tokens": [
87
+ 512,
88
+ 0,
89
+ 0
90
+ ],
91
+ "prime_attn_dropout": 0.0,
92
+ "prime_attn_order": [
93
+ 2,
94
+ 0,
95
+ 0
96
+ ],
97
+ "prime_blocks": 32,
98
+ "prime_depth": [
99
+ 18,
100
+ 3,
101
+ 3
102
+ ],
103
+ "prime_emb_dropout": 0.0,
104
+ "prime_heads": 4,
105
+ "prime_init_scale": [
106
+ 0.1,
107
+ 0.4,
108
+ 0.4
109
+ ],
110
+ "prime_loss_fraction": [
111
+ 0.4,
112
+ 0.0,
113
+ 0.0
114
+ ],
115
+ "prime_m_attn": 0.25,
116
+ "prime_m_mlp": 1.0,
117
+ "prime_n_vocab": 80,
118
+ "prime_pos_init": false,
119
+ "prime_res_scale": false,
120
+ "prime_resid_dropout": 0.0,
121
+ "prime_spread": null,
122
+ "prime_width": [
123
+ 1280,
124
+ 128,
125
+ 128
126
+ ],
127
+ "prime_zero_out": false,
128
+ "prior_alignment_head": [
129
+ 2,
130
+ null,
131
+ null
132
+ ],
133
+ "prior_alignment_layer": [
134
+ 68,
135
+ null,
136
+ null
137
+ ],
138
+ "prior_attn_dropout": 0,
139
+ "prior_attn_order": [
140
+ 10,
141
+ 2,
142
+ 2
143
+ ],
144
+ "prior_blocks": 128,
145
+ "prior_depth": [
146
+ 79,
147
+ 72,
148
+ 72
149
+ ],
150
+ "prior_emb_dropout": 0,
151
+ "prior_init_scale": [
152
+ 0.2,
153
+ 1,
154
+ 1
155
+ ],
156
+ "prior_latent_dim": 2048,
157
+ "prior_m_attn": 0.25,
158
+ "prior_n_ctx": [
159
+ 8192,
160
+ 8192,
161
+ 8192
162
+ ],
163
+ "prior_n_heads": [
164
+ 8,
165
+ 1,
166
+ 1
167
+ ],
168
+ "prior_pos_init": false,
169
+ "prior_res_scale": false,
170
+ "prior_resid_dropout": 0,
171
+ "prior_spread": null,
172
+ "prior_width": [
173
+ 4800,
174
+ 1920,
175
+ 1920
176
+ ],
177
+ "prior_zero_out": false,
178
+ "sample_length": 1058304,
179
+ "sampling_rate": 44100,
180
+ "single_enc_dec": [
181
+ false,
182
+ false,
183
+ false
184
+ ],
185
+ "timing_dims": 128,
186
+ "torch_dtype": "float32",
187
+ "transformers_version": "4.22.0.dev0",
188
+ "vqvae_codebook_dimension": 2048,
189
+ "vqvae_commit": 0.02,
190
+ "vqvae_conv_block_depth": 4,
191
+ "vqvae_conv_block_width": 32,
192
+ "vqvae_depth": 4,
193
+ "vqvae_dilation_cycle": null,
194
+ "vqvae_dilation_growth_rate": 3,
195
+ "vqvae_downs_t": [
196
+ 3,
197
+ 2,
198
+ 2
199
+ ],
200
+ "vqvae_emmbedding_width": 64,
201
+ "vqvae_levels": 3,
202
+ "vqvae_lmu": 0.99,
203
+ "vqvae_m_conv": 1,
204
+ "vqvae_multipliers": [
205
+ 2,
206
+ 1,
207
+ 1
208
+ ],
209
+ "vqvae_music_tokens_shapes": [
210
+ [
211
+ 8268
212
+ ],
213
+ [
214
+ 33072
215
+ ],
216
+ [
217
+ 132288
218
+ ]
219
+ ],
220
+ "vqvae_reverse_decoder_dilation": 1,
221
+ "vqvae_strides_t": [
222
+ 2,
223
+ 2,
224
+ 2
225
+ ],
226
+ "vqvae_width": 64
227
+ }
pytorch_model-00001-of-00004.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e39936eb64985f559f496871e544cd36b0d0387bc399d767cc6c5b6d083bd84
3
+ size 9911113867
pytorch_model-00002-of-00004.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7018f8e2e8817f8ccd09731101a00b3558eed729ad57d04b7eb70adf754f2a7a
3
+ size 9958785645
pytorch_model-00003-of-00004.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:163ba8a3a74431a5c48a61716931260c1a9cecdb377ef7623ba73fe101511722
3
+ size 9958785685
pytorch_model-00004-of-00004.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e9a5cdaa4bdbb9aabe609486df9c86e727e74413f90e1b87dd843490fa26865
3
+ size 592587412
pytorch_model.bin.index.json ADDED
The diff for this file is too large to render. See raw diff