Hugging Face's logo

smartdigitalnetworks
/
SongGeneration-v2-large-mlx

Text-to-Audio
MLX
Safetensors
apple-silicon
music-generation
song-generation
audiolm
Model card Files
xet
 Community
smartdigitalnetworks ailuntz commited on
Commit
36dba8e
·
0 Parent(s):

Duplicate from mlx-community/SongGeneration-v2-large

Browse files

Co-authored-by: ailuntz <ailuntz@users.noreply.huggingface.co>

This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +35 -0
  2. README.md +141 -0
  3. config.json +198 -0
  4. config.official.yaml +141 -0
  5. mlx_manifest.json +198 -0
  6. model-00001-of-00163.safetensors +3 -0
  7. model-00002-of-00163.safetensors +3 -0
  8. model-00003-of-00163.safetensors +3 -0
  9. model-00004-of-00163.safetensors +3 -0
  10. model-00005-of-00163.safetensors +3 -0
  11. model-00006-of-00163.safetensors +3 -0
  12. model-00007-of-00163.safetensors +3 -0
  13. model-00008-of-00163.safetensors +3 -0
  14. model-00009-of-00163.safetensors +3 -0
  15. model-00010-of-00163.safetensors +3 -0
  16. model-00011-of-00163.safetensors +3 -0
  17. model-00012-of-00163.safetensors +3 -0
  18. model-00013-of-00163.safetensors +3 -0
  19. model-00014-of-00163.safetensors +3 -0
  20. model-00015-of-00163.safetensors +3 -0
  21. model-00016-of-00163.safetensors +3 -0
  22. model-00017-of-00163.safetensors +3 -0
  23. model-00018-of-00163.safetensors +3 -0
  24. model-00019-of-00163.safetensors +3 -0
  25. model-00020-of-00163.safetensors +3 -0
  26. model-00021-of-00163.safetensors +3 -0
  27. model-00022-of-00163.safetensors +3 -0
  28. model-00023-of-00163.safetensors +3 -0
  29. model-00024-of-00163.safetensors +3 -0
  30. model-00025-of-00163.safetensors +3 -0
  31. model-00026-of-00163.safetensors +3 -0
  32. model-00027-of-00163.safetensors +3 -0
  33. model-00028-of-00163.safetensors +3 -0
  34. model-00029-of-00163.safetensors +3 -0
  35. model-00030-of-00163.safetensors +3 -0
  36. model-00031-of-00163.safetensors +3 -0
  37. model-00032-of-00163.safetensors +3 -0
  38. model-00033-of-00163.safetensors +3 -0
  39. model-00034-of-00163.safetensors +3 -0
  40. model-00035-of-00163.safetensors +3 -0
  41. model-00036-of-00163.safetensors +3 -0
  42. model-00037-of-00163.safetensors +3 -0
  43. model-00038-of-00163.safetensors +3 -0
  44. model-00039-of-00163.safetensors +3 -0
  45. model-00040-of-00163.safetensors +3 -0
  46. model-00041-of-00163.safetensors +3 -0
  47. model-00042-of-00163.safetensors +3 -0
  48. model-00043-of-00163.safetensors +3 -0
  49. model-00044-of-00163.safetensors +3 -0
  50. model-00045-of-00163.safetensors +3 -0
.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,141 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ library_name: mlx
4
+ pipeline_tag: text-to-audio
5
+ base_model:
6
+ - tencent/SongGeneration
7
+ tags:
8
+ - mlx
9
+ - apple-silicon
10
+ - music-generation
11
+ - song-generation
12
+ - audiolm
13
+ ---
14
+
15
+ Part of the SongGeneration MLX conversion set. Collection: https://huggingface.co/collections/mlx-community/songgeneration-v2-mlx-6a1bf9342dd0806419737229
16
+
17
+ # SongGeneration-v2-large
18
+
19
+ Apple MLX weights for the autoregressive `audiolm` token generator from Tencent SongGeneration v2-large.
20
+
21
+ This default repository is the bf16 variant and is equivalent to [`mlx-community/SongGeneration-v2-large-bf16`](https://huggingface.co/mlx-community/SongGeneration-v2-large-bf16). Use this repo when you want the recommended default checkpoint.
22
+
23
+ This is not a full-stack pure MLX audio pipeline yet: token generation runs with MLX, while FLAC decoding currently uses the official PyTorch Flow1dVAE / separate-tokenizer bridge in [`ailuntx/SongGeneration-MLX`](https://github.com/ailuntx/SongGeneration-MLX).
24
+
25
+ ## TL;DR
26
+
27
+ | | |
28
+ |---|---|
29
+ | **Variant** | `v2-large` |
30
+ | **Precision** | `bf16` (`bfloat16`) |
31
+ | **Converted component** | SongGeneration `audiolm` token generator |
32
+ | **Runtime** | [`ailuntx/SongGeneration-MLX`](https://github.com/ailuntx/SongGeneration-MLX) |
33
+ | **Official model** | [`tencent/SongGeneration`](https://huggingface.co/tencent/SongGeneration) |
34
+ | **Official code** | [`tencent-ailab/songgeneration`](https://github.com/tencent-ailab/songgeneration) |
35
+
36
+ ## Quick Start
37
+
38
+ ```bash
39
+ git clone https://github.com/ailuntx/SongGeneration-MLX.git
40
+ cd SongGeneration-MLX
41
+ python -m venv .venv
42
+ .venv/bin/pip install -e .
43
+ .venv/bin/pip install -U huggingface_hub hf_transfer
44
+
45
+ HF_HUB_ENABLE_HF_TRANSFER=1 .venv/bin/hf download mlx-community/SongGeneration-v2-large --local-dir ./models/SongGeneration-v2-large
46
+
47
+ .venv/bin/python -m songgeneration_mlx.cli \
48
+ --model ./models/SongGeneration-v2-large \
49
+ --lyrics '[verse] hello from mlx [chorus] sing it again' \
50
+ --description 'Pop, female vocal, bright production, [Musicality-medium].' \
51
+ --duration 2 \
52
+ --top-k 50 \
53
+ --temperature 0.9 \
54
+ --output tokens_2s.npz
55
+ ```
56
+
57
+ ## Decode Tokens
58
+
59
+ The MLX runtime writes discrete song tokens. To create FLAC audio, use the official decoder bridge in [`ailuntx/SongGeneration-MLX`](https://github.com/ailuntx/SongGeneration-MLX). The bridge needs the official SongGeneration `runtime/` assets, but it does not need the original SongGeneration LM `model.pt` when `--mlx-model` is used.
60
+
61
+ ```bash
62
+ python -m venv .venv-decoder
63
+ .venv-decoder/bin/pip install -U pip
64
+ .venv-decoder/bin/pip install \
65
+ -r third_party/SongGeneration/requirements.txt \
66
+ -r third_party/SongGeneration/requirements_nodeps.txt \
67
+ soundfile
68
+
69
+ HF_HUB_ENABLE_HF_TRANSFER=1 .venv/bin/hf download tencent/SongGeneration \
70
+ --include "runtime/*" \
71
+ --local-dir ./third_party/SongGeneration
72
+
73
+ PYTORCH_ENABLE_MPS_FALLBACK=1 SONGGEN_DEVICE=mps \
74
+ .venv-decoder/bin/python scripts/decode_tokens_official.py \
75
+ --mlx-model ./models/SongGeneration-v2-large \
76
+ --tokens ./tokens_2s.npz \
77
+ --output ./output_2s.flac \
78
+ --device mps
79
+ ```
80
+
81
+ ## Variants
82
+
83
+ | Variant | Disk | Notes |
84
+ |---|---:|---|
85
+ | `SongGeneration-v2-medium-fp32` | 10G | high-precision medium baseline |
86
+ | `SongGeneration-v2-medium-bf16` | 5.2G | recommended medium bf16 quality baseline |
87
+ | `SongGeneration-v2-medium-8bit` | 2.8G | smaller medium checkpoint |
88
+ | `SongGeneration-v2-medium-4bit` | 1.5G | smallest medium checkpoint |
89
+ | `SongGeneration-v2-large-fp32` | 19G | high-precision large baseline |
90
+ | `SongGeneration-v2-large` | 9.5G | default large checkpoint, same weights as bf16 |
91
+ | `SongGeneration-v2-large-bf16` | 9.5G | explicit large bf16 quality baseline |
92
+ | `SongGeneration-v2-large-8bit` | 5.0G | smaller large checkpoint |
93
+ | `SongGeneration-v2-large-4bit` | 2.7G | smallest large checkpoint |
94
+ | `SongGeneration-v2-fast-*` | pending | upstream fast weights were not publicly available when checked on 2026-05-31 |
95
+
96
+ ## Layout
97
+
98
+ ```text
99
+ SongGeneration-v2-large/
100
+ |-- model-00001-of-000xx.safetensors
101
+ |-- model.safetensors.index.json
102
+ |-- config.json
103
+ |-- mlx_manifest.json
104
+ |-- config.official.yaml
105
+ |-- vocab.yaml
106
+ `-- qwen2_tokenizer/
107
+ ```
108
+
109
+ ## Validation
110
+
111
+ Local Apple Silicon validation was run on the medium bf16 path:
112
+
113
+ | Test | Result |
114
+ |---|---|
115
+ | 12s MLX token generation | 550 pattern steps, about 1 minute wall time |
116
+ | 12s official decoder bridge | 73.27s wall time |
117
+ | 12s FLAC | 48kHz stereo, 12.000s, RMS about `0.163` |
118
+
119
+ The official recent-token repetition penalty is implemented in the MLX runtime. Without it, 12s generations collapse into repeated tokens and decode close to silence.
120
+
121
+ ## License
122
+
123
+ License follows the upstream SongGeneration release. Check the official model card and repository for the authoritative model license.
124
+
125
+ ## Citation
126
+
127
+ ```bibtex
128
+ @misc{songgeneration-mlx,
129
+ title = {SongGeneration-MLX: Apple MLX port of SongGeneration},
130
+ author = {ailuntx},
131
+ year = {2026},
132
+ url = {https://github.com/ailuntx/SongGeneration-MLX},
133
+ }
134
+
135
+ @article{lei2025levo,
136
+ title = {LeVo: High-Quality Song Generation with Multi-Preference Alignment},
137
+ author = {Lei, Shun and Xu, Yaoxun and Lin, Zhiwei and Zhang, Huaicheng and Tan, Wei and Chen, Hangting and Yu, Jianwei and Zhang, Yixuan and Yang, Chenyu and Zhu, Haina and Wang, Shuai and Wu, Zhiyong and Yu, Dong},
138
+ journal = {arXiv preprint arXiv:2506.07520},
139
+ year = {2025},
140
+ }
141
+ ```
config.json ADDED
@@ -0,0 +1,198 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "format": "songgeneration-mlx-audiolm",
3
+ "source": "tencent/SongGeneration",
4
+ "official_code": "https://github.com/tencent-ailab/songgeneration",
5
+ "variant": "v2-large",
6
+ "precision": "bfloat16",
7
+ "runtime": {
8
+ "hidden_size": 2048,
9
+ "intermediate_size": 11008,
10
+ "num_heads": 16,
11
+ "num_layers": 36,
12
+ "num_layers_sub": 12,
13
+ "code_depth": 3,
14
+ "code_size": 16384,
15
+ "prompt_len": 10,
16
+ "frame_rate": 25,
17
+ "max_position_embeddings": 10000,
18
+ "max_position_embeddings_sub": 10000,
19
+ "rope_theta": 500000.0,
20
+ "rope_theta_sub": 500000.0,
21
+ "cfg_coef": 1.5
22
+ },
23
+ "components": {
24
+ "audiolm": "converted to MLX safetensors",
25
+ "qwen_tokenizer": "copied for lyric/style tokenization",
26
+ "audio_decoder": "use official PyTorch Flow1dVAE/separate tokenizer bridge for now"
27
+ },
28
+ "skipped_keys": [
29
+ "audiolm.out_norm.weight",
30
+ "audiolm.out_norm.bias",
31
+ "audiolm.transformer2.lm_head.weight"
32
+ ],
33
+ "weight_files": [
34
+ "model-00001-of-00163.safetensors",
35
+ "model-00002-of-00163.safetensors",
36
+ "model-00003-of-00163.safetensors",
37
+ "model-00004-of-00163.safetensors",
38
+ "model-00005-of-00163.safetensors",
39
+ "model-00006-of-00163.safetensors",
40
+ "model-00007-of-00163.safetensors",
41
+ "model-00008-of-00163.safetensors",
42
+ "model-00009-of-00163.safetensors",
43
+ "model-00010-of-00163.safetensors",
44
+ "model-00011-of-00163.safetensors",
45
+ "model-00012-of-00163.safetensors",
46
+ "model-00013-of-00163.safetensors",
47
+ "model-00014-of-00163.safetensors",
48
+ "model-00015-of-00163.safetensors",
49
+ "model-00016-of-00163.safetensors",
50
+ "model-00017-of-00163.safetensors",
51
+ "model-00018-of-00163.safetensors",
52
+ "model-00019-of-00163.safetensors",
53
+ "model-00020-of-00163.safetensors",
54
+ "model-00021-of-00163.safetensors",
55
+ "model-00022-of-00163.safetensors",
56
+ "model-00023-of-00163.safetensors",
57
+ "model-00024-of-00163.safetensors",
58
+ "model-00025-of-00163.safetensors",
59
+ "model-00026-of-00163.safetensors",
60
+ "model-00027-of-00163.safetensors",
61
+ "model-00028-of-00163.safetensors",
62
+ "model-00029-of-00163.safetensors",
63
+ "model-00030-of-00163.safetensors",
64
+ "model-00031-of-00163.safetensors",
65
+ "model-00032-of-00163.safetensors",
66
+ "model-00033-of-00163.safetensors",
67
+ "model-00034-of-00163.safetensors",
68
+ "model-00035-of-00163.safetensors",
69
+ "model-00036-of-00163.safetensors",
70
+ "model-00037-of-00163.safetensors",
71
+ "model-00038-of-00163.safetensors",
72
+ "model-00039-of-00163.safetensors",
73
+ "model-00040-of-00163.safetensors",
74
+ "model-00041-of-00163.safetensors",
75
+ "model-00042-of-00163.safetensors",
76
+ "model-00043-of-00163.safetensors",
77
+ "model-00044-of-00163.safetensors",
78
+ "model-00045-of-00163.safetensors",
79
+ "model-00046-of-00163.safetensors",
80
+ "model-00047-of-00163.safetensors",
81
+ "model-00048-of-00163.safetensors",
82
+ "model-00049-of-00163.safetensors",
83
+ "model-00050-of-00163.safetensors",
84
+ "model-00051-of-00163.safetensors",
85
+ "model-00052-of-00163.safetensors",
86
+ "model-00053-of-00163.safetensors",
87
+ "model-00054-of-00163.safetensors",
88
+ "model-00055-of-00163.safetensors",
89
+ "model-00056-of-00163.safetensors",
90
+ "model-00057-of-00163.safetensors",
91
+ "model-00058-of-00163.safetensors",
92
+ "model-00059-of-00163.safetensors",
93
+ "model-00060-of-00163.safetensors",
94
+ "model-00061-of-00163.safetensors",
95
+ "model-00062-of-00163.safetensors",
96
+ "model-00063-of-00163.safetensors",
97
+ "model-00064-of-00163.safetensors",
98
+ "model-00065-of-00163.safetensors",
99
+ "model-00066-of-00163.safetensors",
100
+ "model-00067-of-00163.safetensors",
101
+ "model-00068-of-00163.safetensors",
102
+ "model-00069-of-00163.safetensors",
103
+ "model-00070-of-00163.safetensors",
104
+ "model-00071-of-00163.safetensors",
105
+ "model-00072-of-00163.safetensors",
106
+ "model-00073-of-00163.safetensors",
107
+ "model-00074-of-00163.safetensors",
108
+ "model-00075-of-00163.safetensors",
109
+ "model-00076-of-00163.safetensors",
110
+ "model-00077-of-00163.safetensors",
111
+ "model-00078-of-00163.safetensors",
112
+ "model-00079-of-00163.safetensors",
113
+ "model-00080-of-00163.safetensors",
114
+ "model-00081-of-00163.safetensors",
115
+ "model-00082-of-00163.safetensors",
116
+ "model-00083-of-00163.safetensors",
117
+ "model-00084-of-00163.safetensors",
118
+ "model-00085-of-00163.safetensors",
119
+ "model-00086-of-00163.safetensors",
120
+ "model-00087-of-00163.safetensors",
121
+ "model-00088-of-00163.safetensors",
122
+ "model-00089-of-00163.safetensors",
123
+ "model-00090-of-00163.safetensors",
124
+ "model-00091-of-00163.safetensors",
125
+ "model-00092-of-00163.safetensors",
126
+ "model-00093-of-00163.safetensors",
127
+ "model-00094-of-00163.safetensors",
128
+ "model-00095-of-00163.safetensors",
129
+ "model-00096-of-00163.safetensors",
130
+ "model-00097-of-00163.safetensors",
131
+ "model-00098-of-00163.safetensors",
132
+ "model-00099-of-00163.safetensors",
133
+ "model-00100-of-00163.safetensors",
134
+ "model-00101-of-00163.safetensors",
135
+ "model-00102-of-00163.safetensors",
136
+ "model-00103-of-00163.safetensors",
137
+ "model-00104-of-00163.safetensors",
138
+ "model-00105-of-00163.safetensors",
139
+ "model-00106-of-00163.safetensors",
140
+ "model-00107-of-00163.safetensors",
141
+ "model-00108-of-00163.safetensors",
142
+ "model-00109-of-00163.safetensors",
143
+ "model-00110-of-00163.safetensors",
144
+ "model-00111-of-00163.safetensors",
145
+ "model-00112-of-00163.safetensors",
146
+ "model-00113-of-00163.safetensors",
147
+ "model-00114-of-00163.safetensors",
148
+ "model-00115-of-00163.safetensors",
149
+ "model-00116-of-00163.safetensors",
150
+ "model-00117-of-00163.safetensors",
151
+ "model-00118-of-00163.safetensors",
152
+ "model-00119-of-00163.safetensors",
153
+ "model-00120-of-00163.safetensors",
154
+ "model-00121-of-00163.safetensors",
155
+ "model-00122-of-00163.safetensors",
156
+ "model-00123-of-00163.safetensors",
157
+ "model-00124-of-00163.safetensors",
158
+ "model-00125-of-00163.safetensors",
159
+ "model-00126-of-00163.safetensors",
160
+ "model-00127-of-00163.safetensors",
161
+ "model-00128-of-00163.safetensors",
162
+ "model-00129-of-00163.safetensors",
163
+ "model-00130-of-00163.safetensors",
164
+ "model-00131-of-00163.safetensors",
165
+ "model-00132-of-00163.safetensors",
166
+ "model-00133-of-00163.safetensors",
167
+ "model-00134-of-00163.safetensors",
168
+ "model-00135-of-00163.safetensors",
169
+ "model-00136-of-00163.safetensors",
170
+ "model-00137-of-00163.safetensors",
171
+ "model-00138-of-00163.safetensors",
172
+ "model-00139-of-00163.safetensors",
173
+ "model-00140-of-00163.safetensors",
174
+ "model-00141-of-00163.safetensors",
175
+ "model-00142-of-00163.safetensors",
176
+ "model-00143-of-00163.safetensors",
177
+ "model-00144-of-00163.safetensors",
178
+ "model-00145-of-00163.safetensors",
179
+ "model-00146-of-00163.safetensors",
180
+ "model-00147-of-00163.safetensors",
181
+ "model-00148-of-00163.safetensors",
182
+ "model-00149-of-00163.safetensors",
183
+ "model-00150-of-00163.safetensors",
184
+ "model-00151-of-00163.safetensors",
185
+ "model-00152-of-00163.safetensors",
186
+ "model-00153-of-00163.safetensors",
187
+ "model-00154-of-00163.safetensors",
188
+ "model-00155-of-00163.safetensors",
189
+ "model-00156-of-00163.safetensors",
190
+ "model-00157-of-00163.safetensors",
191
+ "model-00158-of-00163.safetensors",
192
+ "model-00159-of-00163.safetensors",
193
+ "model-00160-of-00163.safetensors",
194
+ "model-00161-of-00163.safetensors",
195
+ "model-00162-of-00163.safetensors",
196
+ "model-00163-of-00163.safetensors"
197
+ ]
198
+ }
config.official.yaml ADDED
@@ -0,0 +1,141 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # ================ Train Config ================ #
2
+ lyric_processor:
3
+ max_dur: 270
4
+ min_dur: 30
5
+ prompt_len: 10
6
+ pad_to_max: true
7
+
8
+
9
+ # ================ Audio tokenzier ================ #
10
+ audio_tokenizer_checkpoint: Flow1dVAE1rvq_./ckpt/model_1rvq/model_2_fixed.safetensors
11
+ audio_tokenizer_frame_rate: 25
12
+ audio_tokenizer_code_depth: 1
13
+ sample_rate: 48000
14
+
15
+ audio_tokenizer_checkpoint_sep: Flow1dVAESeparate_./ckpt/model_septoken/model_2.safetensors
16
+ audio_tokenizer_frame_rate_sep: 25
17
+ audio_tokenizer_code_depth_sep: 2
18
+ sample_rate_sep: 48000
19
+
20
+ # ================ VAE ================ #
21
+ vae_config: ./ckpt/vae/stable_audio_1920_vae.json
22
+ vae_model: ./ckpt/vae/autoencoder_music_1320k.ckpt
23
+
24
+ # ================== LM =========================== #
25
+ lm:
26
+ lm_type: Llama # [Llama]
27
+ dim: 2048
28
+ intermediate_size: 11008
29
+ num_heads: 16
30
+ num_layers: 36
31
+ num_layers_sub: 12
32
+ code_depth: 3
33
+ code_size: 16384
34
+ max_position_embeddings: 10000
35
+ max_position_embeddings_sub: 10000
36
+ rope_theta: 500000.0
37
+ rope_theta_sub: 500000.0
38
+ dropout: 0.0
39
+ use_flash_attn_2: true
40
+ activation: gelu
41
+ norm_first: true
42
+ bias_ff: false
43
+ bias_attn: false
44
+ causal: true
45
+ custom: false
46
+ memory_efficient: true
47
+ attention_as_float32: false
48
+ layer_scale: null
49
+ positional_embedding: sin
50
+ xpos: false
51
+ checkpointing: torch
52
+ weight_init: gaussian
53
+ depthwise_init: current
54
+ zero_bias_init: true
55
+ norm: layer_norm
56
+ cross_attention: false
57
+ qk_layer_norm: false
58
+ qk_layer_norm_cross: false
59
+ attention_dropout: null
60
+ kv_repeat: 1
61
+
62
+ codebooks_pattern:
63
+ modeling: delay
64
+ delay:
65
+ delays: [ 0, 250, 250 ]
66
+ flatten_first: 0
67
+ empty_initial: 0
68
+
69
+ # ================ Conditioners ===================== #
70
+ classifier_free_guidance:
71
+ # drop all conditions simultaneously
72
+ training_dropout: 0.15
73
+ inference_coef: 1.5
74
+
75
+ attribute_dropout:
76
+ # drop each condition separately
77
+ args:
78
+ active_on_eval: false
79
+ text:
80
+ description: 0.0
81
+ type_info: 0.2
82
+ audio:
83
+ prompt_audio: 0.5
84
+
85
+
86
+ use_text_training: True
87
+ fuser:
88
+ sum: []
89
+ prepend: [ description, prompt_audio, type_info ] # this order is the SAME with the input concatenation order
90
+
91
+ conditioners:
92
+ prompt_audio:
93
+ model: qt_embedding
94
+ qt_embedding:
95
+ code_size: 16384
96
+ code_depth: 3
97
+ max_len: ${eval:${prompt_len}*${audio_tokenizer_frame_rate}+2} # 25*10+2+1
98
+ description:
99
+ model: QwTokenizer
100
+ QwTokenizer:
101
+ token_path: third_party/Qwen2-7B
102
+ max_len: 600
103
+ add_token_list: ${load_yaml:conf/vocab.yaml}
104
+ type_info:
105
+ model: QwTextTokenizer
106
+ QwTextTokenizer:
107
+ token_path: third_party/Qwen2-7B
108
+ max_len: 100
109
+
110
+ offload:
111
+ audiolm:
112
+ offload_module: self
113
+ cpu_mem_gb: 0
114
+ pre_copy_step: 1
115
+ clean_cache_after_forward: false
116
+ dtype: torch.float16
117
+ offload_layer_dict:
118
+ transformer: 4
119
+ transformer2: 4
120
+ ignore_layer_list: []
121
+ clean_cache_wrapper:
122
+ module: self
123
+ method_name: _sample_next_token
124
+ diff_mem_gb_thre: 5
125
+ debug: false
126
+
127
+ wav_tokenizer_diffusion:
128
+ offload_module: self.model.model
129
+ pre_copy_step: 1
130
+ clean_cache_after_forward: false
131
+ cpu_mem_gb: -1
132
+ dtype: null
133
+ offload_layer_dict:
134
+ cfm_wrapper: 5
135
+ hubert: 4
136
+ ignore_layer_list: []
137
+ clean_cache_wrapper:
138
+ module: self.model.model.cfm_wrapper.estimator
139
+ method_name: forward
140
+ diff_mem_gb_thre: 1
141
+ debug: false
mlx_manifest.json ADDED
@@ -0,0 +1,198 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "format": "songgeneration-mlx-audiolm",
3
+ "source": "tencent/SongGeneration",
4
+ "official_code": "https://github.com/tencent-ailab/songgeneration",
5
+ "variant": "v2-large",
6
+ "precision": "bfloat16",
7
+ "runtime": {
8
+ "hidden_size": 2048,
9
+ "intermediate_size": 11008,
10
+ "num_heads": 16,
11
+ "num_layers": 36,
12
+ "num_layers_sub": 12,
13
+ "code_depth": 3,
14
+ "code_size": 16384,
15
+ "prompt_len": 10,
16
+ "frame_rate": 25,
17
+ "max_position_embeddings": 10000,
18
+ "max_position_embeddings_sub": 10000,
19
+ "rope_theta": 500000.0,
20
+ "rope_theta_sub": 500000.0,
21
+ "cfg_coef": 1.5
22
+ },
23
+ "components": {
24
+ "audiolm": "converted to MLX safetensors",
25
+ "qwen_tokenizer": "copied for lyric/style tokenization",
26
+ "audio_decoder": "use official PyTorch Flow1dVAE/separate tokenizer bridge for now"
27
+ },
28
+ "skipped_keys": [
29
+ "audiolm.out_norm.weight",
30
+ "audiolm.out_norm.bias",
31
+ "audiolm.transformer2.lm_head.weight"
32
+ ],
33
+ "weight_files": [
34
+ "model-00001-of-00163.safetensors",
35
+ "model-00002-of-00163.safetensors",
36
+ "model-00003-of-00163.safetensors",
37
+ "model-00004-of-00163.safetensors",
38
+ "model-00005-of-00163.safetensors",
39
+ "model-00006-of-00163.safetensors",
40
+ "model-00007-of-00163.safetensors",
41
+ "model-00008-of-00163.safetensors",
42
+ "model-00009-of-00163.safetensors",
43
+ "model-00010-of-00163.safetensors",
44
+ "model-00011-of-00163.safetensors",
45
+ "model-00012-of-00163.safetensors",
46
+ "model-00013-of-00163.safetensors",
47
+ "model-00014-of-00163.safetensors",
48
+ "model-00015-of-00163.safetensors",
49
+ "model-00016-of-00163.safetensors",
50
+ "model-00017-of-00163.safetensors",
51
+ "model-00018-of-00163.safetensors",
52
+ "model-00019-of-00163.safetensors",
53
+ "model-00020-of-00163.safetensors",
54
+ "model-00021-of-00163.safetensors",
55
+ "model-00022-of-00163.safetensors",
56
+ "model-00023-of-00163.safetensors",
57
+ "model-00024-of-00163.safetensors",
58
+ "model-00025-of-00163.safetensors",
59
+ "model-00026-of-00163.safetensors",
60
+ "model-00027-of-00163.safetensors",
61
+ "model-00028-of-00163.safetensors",
62
+ "model-00029-of-00163.safetensors",
63
+ "model-00030-of-00163.safetensors",
64
+ "model-00031-of-00163.safetensors",
65
+ "model-00032-of-00163.safetensors",
66
+ "model-00033-of-00163.safetensors",
67
+ "model-00034-of-00163.safetensors",
68
+ "model-00035-of-00163.safetensors",
69
+ "model-00036-of-00163.safetensors",
70
+ "model-00037-of-00163.safetensors",
71
+ "model-00038-of-00163.safetensors",
72
+ "model-00039-of-00163.safetensors",
73
+ "model-00040-of-00163.safetensors",
74
+ "model-00041-of-00163.safetensors",
75
+ "model-00042-of-00163.safetensors",
76
+ "model-00043-of-00163.safetensors",
77
+ "model-00044-of-00163.safetensors",
78
+ "model-00045-of-00163.safetensors",
79
+ "model-00046-of-00163.safetensors",
80
+ "model-00047-of-00163.safetensors",
81
+ "model-00048-of-00163.safetensors",
82
+ "model-00049-of-00163.safetensors",
83
+ "model-00050-of-00163.safetensors",
84
+ "model-00051-of-00163.safetensors",
85
+ "model-00052-of-00163.safetensors",
86
+ "model-00053-of-00163.safetensors",
87
+ "model-00054-of-00163.safetensors",
88
+ "model-00055-of-00163.safetensors",
89
+ "model-00056-of-00163.safetensors",
90
+ "model-00057-of-00163.safetensors",
91
+ "model-00058-of-00163.safetensors",
92
+ "model-00059-of-00163.safetensors",
93
+ "model-00060-of-00163.safetensors",
94
+ "model-00061-of-00163.safetensors",
95
+ "model-00062-of-00163.safetensors",
96
+ "model-00063-of-00163.safetensors",
97
+ "model-00064-of-00163.safetensors",
98
+ "model-00065-of-00163.safetensors",
99
+ "model-00066-of-00163.safetensors",
100
+ "model-00067-of-00163.safetensors",
101
+ "model-00068-of-00163.safetensors",
102
+ "model-00069-of-00163.safetensors",
103
+ "model-00070-of-00163.safetensors",
104
+ "model-00071-of-00163.safetensors",
105
+ "model-00072-of-00163.safetensors",
106
+ "model-00073-of-00163.safetensors",
107
+ "model-00074-of-00163.safetensors",
108
+ "model-00075-of-00163.safetensors",
109
+ "model-00076-of-00163.safetensors",
110
+ "model-00077-of-00163.safetensors",
111
+ "model-00078-of-00163.safetensors",
112
+ "model-00079-of-00163.safetensors",
113
+ "model-00080-of-00163.safetensors",
114
+ "model-00081-of-00163.safetensors",
115
+ "model-00082-of-00163.safetensors",
116
+ "model-00083-of-00163.safetensors",
117
+ "model-00084-of-00163.safetensors",
118
+ "model-00085-of-00163.safetensors",
119
+ "model-00086-of-00163.safetensors",
120
+ "model-00087-of-00163.safetensors",
121
+ "model-00088-of-00163.safetensors",
122
+ "model-00089-of-00163.safetensors",
123
+ "model-00090-of-00163.safetensors",
124
+ "model-00091-of-00163.safetensors",
125
+ "model-00092-of-00163.safetensors",
126
+ "model-00093-of-00163.safetensors",
127
+ "model-00094-of-00163.safetensors",
128
+ "model-00095-of-00163.safetensors",
129
+ "model-00096-of-00163.safetensors",
130
+ "model-00097-of-00163.safetensors",
131
+ "model-00098-of-00163.safetensors",
132
+ "model-00099-of-00163.safetensors",
133
+ "model-00100-of-00163.safetensors",
134
+ "model-00101-of-00163.safetensors",
135
+ "model-00102-of-00163.safetensors",
136
+ "model-00103-of-00163.safetensors",
137
+ "model-00104-of-00163.safetensors",
138
+ "model-00105-of-00163.safetensors",
139
+ "model-00106-of-00163.safetensors",
140
+ "model-00107-of-00163.safetensors",
141
+ "model-00108-of-00163.safetensors",
142
+ "model-00109-of-00163.safetensors",
143
+ "model-00110-of-00163.safetensors",
144
+ "model-00111-of-00163.safetensors",
145
+ "model-00112-of-00163.safetensors",
146
+ "model-00113-of-00163.safetensors",
147
+ "model-00114-of-00163.safetensors",
148
+ "model-00115-of-00163.safetensors",
149
+ "model-00116-of-00163.safetensors",
150
+ "model-00117-of-00163.safetensors",
151
+ "model-00118-of-00163.safetensors",
152
+ "model-00119-of-00163.safetensors",
153
+ "model-00120-of-00163.safetensors",
154
+ "model-00121-of-00163.safetensors",
155
+ "model-00122-of-00163.safetensors",
156
+ "model-00123-of-00163.safetensors",
157
+ "model-00124-of-00163.safetensors",
158
+ "model-00125-of-00163.safetensors",
159
+ "model-00126-of-00163.safetensors",
160
+ "model-00127-of-00163.safetensors",
161
+ "model-00128-of-00163.safetensors",
162
+ "model-00129-of-00163.safetensors",
163
+ "model-00130-of-00163.safetensors",
164
+ "model-00131-of-00163.safetensors",
165
+ "model-00132-of-00163.safetensors",
166
+ "model-00133-of-00163.safetensors",
167
+ "model-00134-of-00163.safetensors",
168
+ "model-00135-of-00163.safetensors",
169
+ "model-00136-of-00163.safetensors",
170
+ "model-00137-of-00163.safetensors",
171
+ "model-00138-of-00163.safetensors",
172
+ "model-00139-of-00163.safetensors",
173
+ "model-00140-of-00163.safetensors",
174
+ "model-00141-of-00163.safetensors",
175
+ "model-00142-of-00163.safetensors",
176
+ "model-00143-of-00163.safetensors",
177
+ "model-00144-of-00163.safetensors",
178
+ "model-00145-of-00163.safetensors",
179
+ "model-00146-of-00163.safetensors",
180
+ "model-00147-of-00163.safetensors",
181
+ "model-00148-of-00163.safetensors",
182
+ "model-00149-of-00163.safetensors",
183
+ "model-00150-of-00163.safetensors",
184
+ "model-00151-of-00163.safetensors",
185
+ "model-00152-of-00163.safetensors",
186
+ "model-00153-of-00163.safetensors",
187
+ "model-00154-of-00163.safetensors",
188
+ "model-00155-of-00163.safetensors",
189
+ "model-00156-of-00163.safetensors",
190
+ "model-00157-of-00163.safetensors",
191
+ "model-00158-of-00163.safetensors",
192
+ "model-00159-of-00163.safetensors",
193
+ "model-00160-of-00163.safetensors",
194
+ "model-00161-of-00163.safetensors",
195
+ "model-00162-of-00163.safetensors",
196
+ "model-00163-of-00163.safetensors"
197
+ ]
198
+ }
model-00001-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5eae90eb8547cb736e88fc92ff9125f01ada51f373d69c101146981f5778b8de
3
+ size 621195376
model-00002-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04940f8f7a3fe47e75baf103cb785b855f5340ba9bfbf2ee9088c3bf413e59c2
3
+ size 819312
model-00003-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:138f009b09dbe33b2c67924d5050bce291309dd5b4289e673654f8684277e834
3
+ size 67117152
model-00004-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ed9b6c77565a7f6d5af8084ef5f3bfb8ddd4db3da0df22e208c724ae01a699f
3
+ size 67117152
model-00005-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2cd0f258c71ff0dae249bda13c4fca74457612c0d1f898ed31957102170521f
3
+ size 67117152
model-00006-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d182fe7115527917e62c587d272f4f06b6bd3a7a405da09eb85bbb0f40306aa
3
+ size 67117152
model-00007-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53b7071817cb38da1ff36287d305c17232e575f2199550271545fdfd9f8b918a
3
+ size 67113056
model-00008-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afb70510a7142486bda4f99697887ed5bb212be07eadb157da96d23bee5f8d5f
3
+ size 67113056
model-00009-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95f6a135803a7453d7643c06041fc2574ff3c2acb0a3443c264d6861ed5bab20
3
+ size 25174368
model-00010-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2dd68b61fbd2696aae7e40cde5fdfde5ce3d4de3e37b53b6e5c397aed0d119b
3
+ size 67117160
model-00011-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7207818089fb0cbdd8a3cd0838dfaecacdab9a6686d2e72bf626380471c80fb1
3
+ size 67117160
model-00012-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2380b8c402fdca9e86bb6b8062ec81d1b6cae78a36ddf5e7c88ccffd410c57f
3
+ size 67117160
model-00013-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b50c5e739527af6f78e6ddfc7034bab8b333a4562fd8aa3e1860c915453040e
3
+ size 8368
model-00014-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75b483adfd06c3f3d927061538347a05a9bea1b0048d5d9dfa6b4d19911c9100
3
+ size 67113072
model-00015-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5585143e34bc2747b75aea3ed7450dbaf9b004433ef6ffb19beacd439e784c7c
3
+ size 45093088
model-00016-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84cacf51ba955d0e6225ccb2cc37843c57e8ab4695fc71cfa90c51437cce34d6
3
+ size 45088888
model-00017-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79a718c39a1fae56e800b92ac7da7e11be966d4264ec357f1a2443039fb00b76
3
+ size 61870560
model-00018-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20cf3e70b44362932f462aee077731147e83c4d3693488d3cb25d17ca4cb0bf5
3
+ size 61870552
model-00019-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0759f77b8cf7f6cd158eae98bcd6e1e6f782f3598b4449dbea2c1a24d8063ab
3
+ size 45088888
model-00020-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7133fae6585086b62d9d78db835dcacd88d2a3efe67f31b11bc9393bf0f77f6f
3
+ size 61870560
model-00021-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0ae0c6a2ffe99f5ca2f946ca9f9e2f4b5f3d7ac50a1f87f8424eba5dd7f6e34
3
+ size 61870552
model-00022-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19fe484a701d92f99a3db6b294b84229beb0d4c812ba0d0fbbf249b589ff379e
3
+ size 45088888
model-00023-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d70a5b1271f75a1980f733108c3599265708d84687e8f8945ef02106376d95c
3
+ size 61870568
model-00024-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87da3095c85858917e3c838c6529ddd5be72d1ac9ed1fdc305eb64153ae3c0dc
3
+ size 61870552
model-00025-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:321f4ed360a03edc1e388cfb4b8ea772798c86778f016253befe125c1de25323
3
+ size 45088888
model-00026-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b150719006463c1eba8fbddd4aeaaf14ecad224c47b5a2d1cfdce2b8bdb73ac9
3
+ size 61870568
model-00027-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db0e66ba28cc53b3d80a5d743b822fa9d94f720d0024f7015d833aef9260cbb9
3
+ size 61870552
model-00028-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acd50678e6ed9a2ec7aa43e0ba734ace92178b849a4c58873dec8d06c97d9175
3
+ size 45088888
model-00029-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e6cb36d8006b38630e6aae77d2a58cfbed99e98eef1bf5d123dc60318d34246
3
+ size 61870568
model-00030-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05f27b6df9f66d3b05aa0b43f0d8c6ba8faae1c75166452579244c6403cbcf18
3
+ size 61870552
model-00031-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7d83af8bccd88e15880d79a561f84422a818a632acc41f42dcd3c75d7b2de88
3
+ size 45088888
model-00032-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0489caf1dd919f3edf1bc5bcca5b2c51316f0bccf3db9573a02ca5806aadfb83
3
+ size 61870568
model-00033-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57a2a0f707df3c4d46f721ad45e1104c9a05db9e46ed333a31bb50c8da2d4898
3
+ size 61870552
model-00034-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18a0d0aeb089e75a7a4802f0996c631eac2ba816005e915ed501d0e0bdfe1a63
3
+ size 45088888
model-00035-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:694d52ad18ea25044ce7ad9ac0b13d5c462353cc27ab96100f3ef8f435ada828
3
+ size 61870568
model-00036-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:816c749499b51ffc05387770ec69d60750c5324ce2b4bfa08c113b934895504c
3
+ size 61870552
model-00037-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d538a63d10206e508df1749e5f59adad164bb0cbeb668be4bca210532f516e0f
3
+ size 45088888
model-00038-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3d4197dbbe50c747d412e6637223b6f4d1ee077482654128e45ed25d1e30104
3
+ size 61870568
model-00039-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0135b0e9214058f67733cd7e108497d1934f9e0a0c8daef47869ced2ac5018dc
3
+ size 61870552
model-00040-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b7955c84fd00c1f584b2fb228a8d078aba4221f38ad902dd9a50e13402d0b55a
3
+ size 45088888
model-00041-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e49b403e7a033291bfcd7841f7b0f940ca272cea24c908ae58222dd2d6ae739
3
+ size 61870568
model-00042-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd0185b53fbbac8838d3fc7c96fa21ca28c0612a7baf2d2c75f9495a0e577634
3
+ size 61870552
model-00043-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95fe89d74feb855a3b7914ff6901b265be77d23b6b75901c74d5a12d46c5f8a9
3
+ size 45088888
model-00044-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:841ce10e125f40340f28017bfa9f8c573639b9707a140cb96e48e88dd858d76a
3
+ size 61870568
model-00045-of-00163.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0677932cdfed754d64d636dea5c5deccd2b420c19c52f96631e7b2e9e1dba02
3
+ size 61870552