phamhungd commited on
Commit
7b380eb
·
verified ·
1 Parent(s): a200b93

LaziBook model files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +6 -0
  2. LICENSE.md +3 -0
  3. README.md +39 -0
  4. app/clone/codec_encoder.onnx +3 -0
  5. app/clone/constants.json +58 -0
  6. app/clone/denoiser.onnx +3 -0
  7. app/clone/reference_encoder.onnx +3 -0
  8. app/clone/speaker_encoder.onnx +3 -0
  9. app/codec/codec_browser_onnx_meta.json +576 -0
  10. app/codec/moss_audio_tokenizer_decode_step.onnx +3 -0
  11. app/codec/moss_audio_tokenizer_encode.onnx +3 -0
  12. app/g2p/sea_g2p.bin +3 -0
  13. app/matcha/date-zh.fst +0 -0
  14. app/matcha/lexicon.txt +0 -0
  15. app/matcha/model-steps-3.onnx +3 -0
  16. app/matcha/number-zh.fst +0 -0
  17. app/matcha/phone-zh.fst +0 -0
  18. app/matcha/tokens.txt +2190 -0
  19. app/matcha/vocos-16khz-univ.onnx +3 -0
  20. app/nano/.distill4 +1 -0
  21. app/nano/.duration +1 -0
  22. app/nano/.int8 +1 -0
  23. app/nano/.tag +1 -0
  24. app/nano/codec_decoder.onnx +3 -0
  25. app/nano/config.json +172 -0
  26. app/nano/constants.json +15 -0
  27. app/nano/null_spk.f32 +0 -0
  28. app/nano/null_style.f32 +0 -0
  29. app/nano/text_encoder.onnx +3 -0
  30. app/nano/voices.json +0 -0
  31. app/phonon/bundle.json +332 -0
  32. app/phonon/flow-c414-t1.mlmodelc/analytics/coremldata.bin +3 -0
  33. app/phonon/flow-c414-t1.mlmodelc/coremldata.bin +3 -0
  34. app/phonon/flow-c414-t1.mlmodelc/model.mil +0 -0
  35. app/phonon/flow-c414-t1.mlmodelc/weights/weight.bin +3 -0
  36. app/phonon/flow-c414-t48.mlmodelc/analytics/coremldata.bin +3 -0
  37. app/phonon/flow-c414-t48.mlmodelc/coremldata.bin +3 -0
  38. app/phonon/flow-c414-t48.mlmodelc/model.mil +0 -0
  39. app/phonon/flow-c414-t48.mlmodelc/weights/weight.bin +3 -0
  40. app/phonon/host.safetensors +3 -0
  41. app/phonon/mimi-c266.mlmodelc/analytics/coremldata.bin +3 -0
  42. app/phonon/mimi-c266.mlmodelc/coremldata.bin +3 -0
  43. app/phonon/mimi-c266.mlmodelc/model.mil +478 -0
  44. app/phonon/mimi-c266.mlmodelc/weights/weight.bin +3 -0
  45. app/phonon/source.json +5 -0
  46. app/phonon/tokenizer.json +0 -0
  47. app/phonon/voices/Alfie.safetensors +3 -0
  48. app/phonon/voices/Azgorath.safetensors +3 -0
  49. app/phonon/voices/Bryony.safetensors +3 -0
  50. app/phonon/voices/Captain Silas Blackwater.safetensors +3 -0
.gitattributes CHANGED
@@ -33,3 +33,9 @@ saved_model/**/* 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
 
 
 
 
 
 
 
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
36
+ app/turbo/audio_emb.f32 filter=lfs diff=lfs merge=lfs -text
37
+ app/turbo/text_emb.f32 filter=lfs diff=lfs merge=lfs -text
38
+ app/turbo/vieneu_backbone_shared.data filter=lfs diff=lfs merge=lfs -text
39
+ app/turbo/xvec_w.f32 filter=lfs diff=lfs merge=lfs -text
40
+ ming-3.0-tts-0.5B-4bit/tokenizer.json filter=lfs diff=lfs merge=lfs -text
41
+ mt/Hy-MT2-1.8B-1.25Bit.gguf filter=lfs diff=lfs merge=lfs -text
LICENSE.md ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ # Licences
2
+
3
+ This repo mirrors third-party models. Each folder keeps its original licence (LICENSE / README inside the folder). See the table in README.md.
README.md ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: mixed
4
+ license_link: LICENSE.md
5
+ language:
6
+ - vi
7
+ - zh
8
+ - en
9
+ pipeline_tag: text-to-speech
10
+ tags:
11
+ - tts
12
+ - on-device
13
+ - lazibook
14
+ ---
15
+
16
+ # LaziBook — model files
17
+
18
+ Model files used by the LaziBook app (offline e-book reader and audiobook player for iOS, macOS and Android).
19
+ Each folder keeps the licence of its original model.
20
+
21
+ ## `app/` — bundled in the app (`ios/Models`, Android `voice_*` packs)
22
+
23
+ | Folder | Model | Source | Licence |
24
+ |---|---|---|---|
25
+ | `app/turbo/` | VieNeu-TTS v3 Turbo (backbone MLP layers 1–10 4-bit, the rest 8-bit) | [pnnbao-ump/VieNeu-TTS-v3-Turbo](https://huggingface.co/pnnbao-ump/VieNeu-TTS-v3-Turbo) | Apache-2.0 |
26
+ | `app/nano/` | VieNeu-TTS v3 Nano: text encoder, codec decoder (int8), voices. The estimator and duration predictor are LaziBook's own trained weights and are kept in a private repo | [pnnbao-ump/VieNeu-TTS-v3-Nano](https://huggingface.co/pnnbao-ump/VieNeu-TTS-v3-Nano) | Apache-2.0 |
27
+ | `app/clone/` | Nano encoders for voice cloning (codec, reference, speaker, denoiser) | VieNeu-TTS v3 Nano | Apache-2.0 |
28
+ | `app/codec/` | MOSS Audio Tokenizer Nano (ONNX) | MOSS-Audio-Tokenizer-Nano-ONNX | Apache-2.0 |
29
+ | `app/g2p/` | sea-g2p (Vietnamese G2P) | [pnnbao97/sea-g2p](https://github.com/pnnbao97/sea-g2p) | Apache-2.0 |
30
+ | `app/phonon/` | Phonon 31M English, Core ML 4-bit palettized, precompiled `.mlmodelc` + 38 voices | [gradium/phonon-31m-en](https://huggingface.co/gradium/phonon-31m-en), [gradium/phonon-voices](https://huggingface.co/gradium/phonon-voices) | Apache-2.0 |
31
+ | `app/vitszh/` | VITS Chinese (sherpa-onnx `vits-zh-ll`) | [k2-fsa/sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx) TTS models | see source |
32
+ | `app/matcha/` | Matcha-TTS zh-en + Vocos 16 kHz (sherpa-onnx) | [dengcunqin/matcha_tts_zh_en_20251010](https://modelscope.cn/models/dengcunqin/matcha_tts_zh_en_20251010) | see source |
33
+
34
+ ## Optional downloads
35
+
36
+ | Folder | Used by | Source | Licence |
37
+ |---|---|---|---|
38
+ | `ming-3.0-tts-0.5B-4bit/` | Mac app, Chinese/English voice pack | [jorgepb96/Ming-3.0-tts-0.5B-4bit](https://huggingface.co/jorgepb96/Ming-3.0-tts-0.5B-4bit) @ `d06229c`, from [inclusionAI/Ming-3.0-tts-0.5B](https://huggingface.co/inclusionAI/Ming-3.0-tts-0.5B) | MIT |
39
+ | `mt/` | Offline translation (Hy-MT2 1.8B 1.25-bit GGUF) | [tencent/Hy-MT2-1.8B-1.25Bit-GGUF](https://huggingface.co/tencent/Hy-MT2-1.8B-1.25Bit-GGUF) @ `9df5c82` | Apache-2.0 |
app/clone/codec_encoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6be19085fac217c1f924ff2ffea9edb4f0fcfb5867d6ea43fb4f5d4dde30c6b
3
+ size 28270354
app/clone/constants.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "latent_mean": [
3
+ 0.1389072835445404,
4
+ -0.13309575617313385,
5
+ -0.11205816268920898,
6
+ -0.11297748237848282,
7
+ -0.18527600169181824,
8
+ -0.08897385001182556,
9
+ -0.12180636823177338,
10
+ -0.15749171376228333,
11
+ -0.2606409788131714,
12
+ -0.10214384645223618,
13
+ -0.010901057161390781,
14
+ -0.011725051328539848,
15
+ 0.5120446681976318,
16
+ 0.06387069076299667,
17
+ -0.05569545924663544,
18
+ 0.11140363663434982,
19
+ -0.23515650629997253,
20
+ -0.007840761914849281,
21
+ -0.002412324072793126,
22
+ 0.03336932882666588,
23
+ -0.020932620391249657,
24
+ -0.25093379616737366,
25
+ 0.01933715119957924,
26
+ 0.06063517928123474
27
+ ],
28
+ "latent_std": [
29
+ 0.5559272170066833,
30
+ 0.620286226272583,
31
+ 0.5138673782348633,
32
+ 0.5440409183502197,
33
+ 0.5967750549316406,
34
+ 0.5915284752845764,
35
+ 0.5132561922073364,
36
+ 0.6036245226860046,
37
+ 0.734268069267273,
38
+ 0.5521435141563416,
39
+ 0.5918031334877014,
40
+ 0.4939822852611542,
41
+ 0.8453294038772583,
42
+ 0.6165397763252258,
43
+ 0.5982130169868469,
44
+ 0.585174024105072,
45
+ 0.7860732078552246,
46
+ 0.6456387042999268,
47
+ 0.5919169187545776,
48
+ 0.5229639410972595,
49
+ 0.57453852891922,
50
+ 0.5858547687530518,
51
+ 0.5638392567634583,
52
+ 0.5982619524002075
53
+ ],
54
+ "latent_scale": 0.25,
55
+ "group": 6,
56
+ "flow_fps": 15.625,
57
+ "ref_max_frames": 140
58
+ }
app/clone/denoiser.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b7ce31f57559902c9de3685522c8c64a1b3a5a2d79eb703d55bc225a51c999c
3
+ size 21468146
app/clone/reference_encoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b2f386336d209c607c4d8f09efd1f82e3873c7e8649bf533743015c5d1a627a
3
+ size 2939150
app/clone/speaker_encoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c1c15b2999548cc8299bf5f655362eac01be763eb7850b1a61b87c2ab2f010c
3
+ size 14801064
app/codec/codec_browser_onnx_meta.json ADDED
@@ -0,0 +1,576 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "format_version": 2,
3
+ "checkpoint_path": "MOSS-Audio-Tokenizer-Nano",
4
+ "files": {
5
+ "encode": "moss_audio_tokenizer_encode.onnx",
6
+ "decode_full": "moss_audio_tokenizer_decode_full.onnx",
7
+ "decode_step": "moss_audio_tokenizer_decode_step.onnx"
8
+ },
9
+ "external_data_files": {
10
+ "moss_audio_tokenizer_encode.onnx": [
11
+ "moss_audio_tokenizer_encode.data"
12
+ ],
13
+ "moss_audio_tokenizer_decode_full.onnx": [
14
+ "moss_audio_tokenizer_decode_shared.data"
15
+ ],
16
+ "moss_audio_tokenizer_decode_step.onnx": [
17
+ "moss_audio_tokenizer_decode_shared.data"
18
+ ]
19
+ },
20
+ "codec_config": {
21
+ "sample_rate": 48000,
22
+ "channels": 2,
23
+ "downsample_rate": 3840,
24
+ "num_quantizers": 16
25
+ },
26
+ "onnx": {
27
+ "opset": 17,
28
+ "encode_input_names": [
29
+ "waveform",
30
+ "input_lengths"
31
+ ],
32
+ "encode_output_names": [
33
+ "audio_codes",
34
+ "audio_code_lengths"
35
+ ],
36
+ "decode_input_names": [
37
+ "audio_codes",
38
+ "audio_code_lengths"
39
+ ],
40
+ "decode_output_names": [
41
+ "audio",
42
+ "audio_lengths"
43
+ ],
44
+ "decode_step_input_names": [
45
+ "audio_codes",
46
+ "audio_code_lengths",
47
+ "transformer_offset_0",
48
+ "transformer_offset_1",
49
+ "transformer_offset_2",
50
+ "transformer_offset_3",
51
+ "attn_offset_0",
52
+ "attn_cached_keys_0",
53
+ "attn_cached_values_0",
54
+ "attn_cached_positions_0",
55
+ "attn_offset_1",
56
+ "attn_cached_keys_1",
57
+ "attn_cached_values_1",
58
+ "attn_cached_positions_1",
59
+ "attn_offset_2",
60
+ "attn_cached_keys_2",
61
+ "attn_cached_values_2",
62
+ "attn_cached_positions_2",
63
+ "attn_offset_3",
64
+ "attn_cached_keys_3",
65
+ "attn_cached_values_3",
66
+ "attn_cached_positions_3",
67
+ "attn_offset_4",
68
+ "attn_cached_keys_4",
69
+ "attn_cached_values_4",
70
+ "attn_cached_positions_4",
71
+ "attn_offset_5",
72
+ "attn_cached_keys_5",
73
+ "attn_cached_values_5",
74
+ "attn_cached_positions_5",
75
+ "attn_offset_6",
76
+ "attn_cached_keys_6",
77
+ "attn_cached_values_6",
78
+ "attn_cached_positions_6",
79
+ "attn_offset_7",
80
+ "attn_cached_keys_7",
81
+ "attn_cached_values_7",
82
+ "attn_cached_positions_7",
83
+ "attn_offset_8",
84
+ "attn_cached_keys_8",
85
+ "attn_cached_values_8",
86
+ "attn_cached_positions_8",
87
+ "attn_offset_9",
88
+ "attn_cached_keys_9",
89
+ "attn_cached_values_9",
90
+ "attn_cached_positions_9",
91
+ "attn_offset_10",
92
+ "attn_cached_keys_10",
93
+ "attn_cached_values_10",
94
+ "attn_cached_positions_10",
95
+ "attn_offset_11",
96
+ "attn_cached_keys_11",
97
+ "attn_cached_values_11",
98
+ "attn_cached_positions_11"
99
+ ],
100
+ "decode_step_output_names": [
101
+ "audio",
102
+ "audio_lengths",
103
+ "transformer_offset_out_0",
104
+ "transformer_offset_out_1",
105
+ "transformer_offset_out_2",
106
+ "transformer_offset_out_3",
107
+ "attn_offset_out_0",
108
+ "attn_cached_keys_out_0",
109
+ "attn_cached_values_out_0",
110
+ "attn_cached_positions_out_0",
111
+ "attn_offset_out_1",
112
+ "attn_cached_keys_out_1",
113
+ "attn_cached_values_out_1",
114
+ "attn_cached_positions_out_1",
115
+ "attn_offset_out_2",
116
+ "attn_cached_keys_out_2",
117
+ "attn_cached_values_out_2",
118
+ "attn_cached_positions_out_2",
119
+ "attn_offset_out_3",
120
+ "attn_cached_keys_out_3",
121
+ "attn_cached_values_out_3",
122
+ "attn_cached_positions_out_3",
123
+ "attn_offset_out_4",
124
+ "attn_cached_keys_out_4",
125
+ "attn_cached_values_out_4",
126
+ "attn_cached_positions_out_4",
127
+ "attn_offset_out_5",
128
+ "attn_cached_keys_out_5",
129
+ "attn_cached_values_out_5",
130
+ "attn_cached_positions_out_5",
131
+ "attn_offset_out_6",
132
+ "attn_cached_keys_out_6",
133
+ "attn_cached_values_out_6",
134
+ "attn_cached_positions_out_6",
135
+ "attn_offset_out_7",
136
+ "attn_cached_keys_out_7",
137
+ "attn_cached_values_out_7",
138
+ "attn_cached_positions_out_7",
139
+ "attn_offset_out_8",
140
+ "attn_cached_keys_out_8",
141
+ "attn_cached_values_out_8",
142
+ "attn_cached_positions_out_8",
143
+ "attn_offset_out_9",
144
+ "attn_cached_keys_out_9",
145
+ "attn_cached_values_out_9",
146
+ "attn_cached_positions_out_9",
147
+ "attn_offset_out_10",
148
+ "attn_cached_keys_out_10",
149
+ "attn_cached_values_out_10",
150
+ "attn_cached_positions_out_10",
151
+ "attn_offset_out_11",
152
+ "attn_cached_keys_out_11",
153
+ "attn_cached_values_out_11",
154
+ "attn_cached_positions_out_11"
155
+ ]
156
+ },
157
+ "streaming_decode": {
158
+ "batch_size": 1,
159
+ "transformer_offsets": [
160
+ {
161
+ "index": 0,
162
+ "decoder_index": 1,
163
+ "input_name": "transformer_offset_0",
164
+ "output_name": "transformer_offset_out_0",
165
+ "shape": [
166
+ 1
167
+ ],
168
+ "dtype": "int32"
169
+ },
170
+ {
171
+ "index": 1,
172
+ "decoder_index": 3,
173
+ "input_name": "transformer_offset_1",
174
+ "output_name": "transformer_offset_out_1",
175
+ "shape": [
176
+ 1
177
+ ],
178
+ "dtype": "int32"
179
+ },
180
+ {
181
+ "index": 2,
182
+ "decoder_index": 5,
183
+ "input_name": "transformer_offset_2",
184
+ "output_name": "transformer_offset_out_2",
185
+ "shape": [
186
+ 1
187
+ ],
188
+ "dtype": "int32"
189
+ },
190
+ {
191
+ "index": 3,
192
+ "decoder_index": 7,
193
+ "input_name": "transformer_offset_3",
194
+ "output_name": "transformer_offset_out_3",
195
+ "shape": [
196
+ 1
197
+ ],
198
+ "dtype": "int32"
199
+ }
200
+ ],
201
+ "attention_caches": [
202
+ {
203
+ "index": 0,
204
+ "decoder_index": 1,
205
+ "layer_index": 0,
206
+ "context": 500,
207
+ "num_heads": 4,
208
+ "head_dim": 64,
209
+ "offset_input_name": "attn_offset_0",
210
+ "offset_output_name": "attn_offset_out_0",
211
+ "cached_keys_input_name": "attn_cached_keys_0",
212
+ "cached_keys_output_name": "attn_cached_keys_out_0",
213
+ "cached_values_input_name": "attn_cached_values_0",
214
+ "cached_values_output_name": "attn_cached_values_out_0",
215
+ "cached_positions_input_name": "attn_cached_positions_0",
216
+ "cached_positions_output_name": "attn_cached_positions_out_0",
217
+ "offset_shape": [
218
+ 1
219
+ ],
220
+ "cache_shape": [
221
+ 1,
222
+ 4,
223
+ 500,
224
+ 64
225
+ ],
226
+ "positions_shape": [
227
+ 1,
228
+ 500
229
+ ],
230
+ "cache_dtype": "float32",
231
+ "positions_dtype": "int32"
232
+ },
233
+ {
234
+ "index": 1,
235
+ "decoder_index": 1,
236
+ "layer_index": 1,
237
+ "context": 500,
238
+ "num_heads": 4,
239
+ "head_dim": 64,
240
+ "offset_input_name": "attn_offset_1",
241
+ "offset_output_name": "attn_offset_out_1",
242
+ "cached_keys_input_name": "attn_cached_keys_1",
243
+ "cached_keys_output_name": "attn_cached_keys_out_1",
244
+ "cached_values_input_name": "attn_cached_values_1",
245
+ "cached_values_output_name": "attn_cached_values_out_1",
246
+ "cached_positions_input_name": "attn_cached_positions_1",
247
+ "cached_positions_output_name": "attn_cached_positions_out_1",
248
+ "offset_shape": [
249
+ 1
250
+ ],
251
+ "cache_shape": [
252
+ 1,
253
+ 4,
254
+ 500,
255
+ 64
256
+ ],
257
+ "positions_shape": [
258
+ 1,
259
+ 500
260
+ ],
261
+ "cache_dtype": "float32",
262
+ "positions_dtype": "int32"
263
+ },
264
+ {
265
+ "index": 2,
266
+ "decoder_index": 1,
267
+ "layer_index": 2,
268
+ "context": 500,
269
+ "num_heads": 4,
270
+ "head_dim": 64,
271
+ "offset_input_name": "attn_offset_2",
272
+ "offset_output_name": "attn_offset_out_2",
273
+ "cached_keys_input_name": "attn_cached_keys_2",
274
+ "cached_keys_output_name": "attn_cached_keys_out_2",
275
+ "cached_values_input_name": "attn_cached_values_2",
276
+ "cached_values_output_name": "attn_cached_values_out_2",
277
+ "cached_positions_input_name": "attn_cached_positions_2",
278
+ "cached_positions_output_name": "attn_cached_positions_out_2",
279
+ "offset_shape": [
280
+ 1
281
+ ],
282
+ "cache_shape": [
283
+ 1,
284
+ 4,
285
+ 500,
286
+ 64
287
+ ],
288
+ "positions_shape": [
289
+ 1,
290
+ 500
291
+ ],
292
+ "cache_dtype": "float32",
293
+ "positions_dtype": "int32"
294
+ },
295
+ {
296
+ "index": 3,
297
+ "decoder_index": 1,
298
+ "layer_index": 3,
299
+ "context": 500,
300
+ "num_heads": 4,
301
+ "head_dim": 64,
302
+ "offset_input_name": "attn_offset_3",
303
+ "offset_output_name": "attn_offset_out_3",
304
+ "cached_keys_input_name": "attn_cached_keys_3",
305
+ "cached_keys_output_name": "attn_cached_keys_out_3",
306
+ "cached_values_input_name": "attn_cached_values_3",
307
+ "cached_values_output_name": "attn_cached_values_out_3",
308
+ "cached_positions_input_name": "attn_cached_positions_3",
309
+ "cached_positions_output_name": "attn_cached_positions_out_3",
310
+ "offset_shape": [
311
+ 1
312
+ ],
313
+ "cache_shape": [
314
+ 1,
315
+ 4,
316
+ 500,
317
+ 64
318
+ ],
319
+ "positions_shape": [
320
+ 1,
321
+ 500
322
+ ],
323
+ "cache_dtype": "float32",
324
+ "positions_dtype": "int32"
325
+ },
326
+ {
327
+ "index": 4,
328
+ "decoder_index": 3,
329
+ "layer_index": 0,
330
+ "context": 800,
331
+ "num_heads": 4,
332
+ "head_dim": 64,
333
+ "offset_input_name": "attn_offset_4",
334
+ "offset_output_name": "attn_offset_out_4",
335
+ "cached_keys_input_name": "attn_cached_keys_4",
336
+ "cached_keys_output_name": "attn_cached_keys_out_4",
337
+ "cached_values_input_name": "attn_cached_values_4",
338
+ "cached_values_output_name": "attn_cached_values_out_4",
339
+ "cached_positions_input_name": "attn_cached_positions_4",
340
+ "cached_positions_output_name": "attn_cached_positions_out_4",
341
+ "offset_shape": [
342
+ 1
343
+ ],
344
+ "cache_shape": [
345
+ 1,
346
+ 4,
347
+ 800,
348
+ 64
349
+ ],
350
+ "positions_shape": [
351
+ 1,
352
+ 800
353
+ ],
354
+ "cache_dtype": "float32",
355
+ "positions_dtype": "int32"
356
+ },
357
+ {
358
+ "index": 5,
359
+ "decoder_index": 3,
360
+ "layer_index": 1,
361
+ "context": 800,
362
+ "num_heads": 4,
363
+ "head_dim": 64,
364
+ "offset_input_name": "attn_offset_5",
365
+ "offset_output_name": "attn_offset_out_5",
366
+ "cached_keys_input_name": "attn_cached_keys_5",
367
+ "cached_keys_output_name": "attn_cached_keys_out_5",
368
+ "cached_values_input_name": "attn_cached_values_5",
369
+ "cached_values_output_name": "attn_cached_values_out_5",
370
+ "cached_positions_input_name": "attn_cached_positions_5",
371
+ "cached_positions_output_name": "attn_cached_positions_out_5",
372
+ "offset_shape": [
373
+ 1
374
+ ],
375
+ "cache_shape": [
376
+ 1,
377
+ 4,
378
+ 800,
379
+ 64
380
+ ],
381
+ "positions_shape": [
382
+ 1,
383
+ 800
384
+ ],
385
+ "cache_dtype": "float32",
386
+ "positions_dtype": "int32"
387
+ },
388
+ {
389
+ "index": 6,
390
+ "decoder_index": 5,
391
+ "layer_index": 0,
392
+ "context": 1200,
393
+ "num_heads": 4,
394
+ "head_dim": 64,
395
+ "offset_input_name": "attn_offset_6",
396
+ "offset_output_name": "attn_offset_out_6",
397
+ "cached_keys_input_name": "attn_cached_keys_6",
398
+ "cached_keys_output_name": "attn_cached_keys_out_6",
399
+ "cached_values_input_name": "attn_cached_values_6",
400
+ "cached_values_output_name": "attn_cached_values_out_6",
401
+ "cached_positions_input_name": "attn_cached_positions_6",
402
+ "cached_positions_output_name": "attn_cached_positions_out_6",
403
+ "offset_shape": [
404
+ 1
405
+ ],
406
+ "cache_shape": [
407
+ 1,
408
+ 4,
409
+ 1200,
410
+ 64
411
+ ],
412
+ "positions_shape": [
413
+ 1,
414
+ 1200
415
+ ],
416
+ "cache_dtype": "float32",
417
+ "positions_dtype": "int32"
418
+ },
419
+ {
420
+ "index": 7,
421
+ "decoder_index": 5,
422
+ "layer_index": 1,
423
+ "context": 1200,
424
+ "num_heads": 4,
425
+ "head_dim": 64,
426
+ "offset_input_name": "attn_offset_7",
427
+ "offset_output_name": "attn_offset_out_7",
428
+ "cached_keys_input_name": "attn_cached_keys_7",
429
+ "cached_keys_output_name": "attn_cached_keys_out_7",
430
+ "cached_values_input_name": "attn_cached_values_7",
431
+ "cached_values_output_name": "attn_cached_values_out_7",
432
+ "cached_positions_input_name": "attn_cached_positions_7",
433
+ "cached_positions_output_name": "attn_cached_positions_out_7",
434
+ "offset_shape": [
435
+ 1
436
+ ],
437
+ "cache_shape": [
438
+ 1,
439
+ 4,
440
+ 1200,
441
+ 64
442
+ ],
443
+ "positions_shape": [
444
+ 1,
445
+ 1200
446
+ ],
447
+ "cache_dtype": "float32",
448
+ "positions_dtype": "int32"
449
+ },
450
+ {
451
+ "index": 8,
452
+ "decoder_index": 7,
453
+ "layer_index": 0,
454
+ "context": 1600,
455
+ "num_heads": 4,
456
+ "head_dim": 64,
457
+ "offset_input_name": "attn_offset_8",
458
+ "offset_output_name": "attn_offset_out_8",
459
+ "cached_keys_input_name": "attn_cached_keys_8",
460
+ "cached_keys_output_name": "attn_cached_keys_out_8",
461
+ "cached_values_input_name": "attn_cached_values_8",
462
+ "cached_values_output_name": "attn_cached_values_out_8",
463
+ "cached_positions_input_name": "attn_cached_positions_8",
464
+ "cached_positions_output_name": "attn_cached_positions_out_8",
465
+ "offset_shape": [
466
+ 1
467
+ ],
468
+ "cache_shape": [
469
+ 1,
470
+ 4,
471
+ 1600,
472
+ 64
473
+ ],
474
+ "positions_shape": [
475
+ 1,
476
+ 1600
477
+ ],
478
+ "cache_dtype": "float32",
479
+ "positions_dtype": "int32"
480
+ },
481
+ {
482
+ "index": 9,
483
+ "decoder_index": 7,
484
+ "layer_index": 1,
485
+ "context": 1600,
486
+ "num_heads": 4,
487
+ "head_dim": 64,
488
+ "offset_input_name": "attn_offset_9",
489
+ "offset_output_name": "attn_offset_out_9",
490
+ "cached_keys_input_name": "attn_cached_keys_9",
491
+ "cached_keys_output_name": "attn_cached_keys_out_9",
492
+ "cached_values_input_name": "attn_cached_values_9",
493
+ "cached_values_output_name": "attn_cached_values_out_9",
494
+ "cached_positions_input_name": "attn_cached_positions_9",
495
+ "cached_positions_output_name": "attn_cached_positions_out_9",
496
+ "offset_shape": [
497
+ 1
498
+ ],
499
+ "cache_shape": [
500
+ 1,
501
+ 4,
502
+ 1600,
503
+ 64
504
+ ],
505
+ "positions_shape": [
506
+ 1,
507
+ 1600
508
+ ],
509
+ "cache_dtype": "float32",
510
+ "positions_dtype": "int32"
511
+ },
512
+ {
513
+ "index": 10,
514
+ "decoder_index": 7,
515
+ "layer_index": 2,
516
+ "context": 1600,
517
+ "num_heads": 4,
518
+ "head_dim": 64,
519
+ "offset_input_name": "attn_offset_10",
520
+ "offset_output_name": "attn_offset_out_10",
521
+ "cached_keys_input_name": "attn_cached_keys_10",
522
+ "cached_keys_output_name": "attn_cached_keys_out_10",
523
+ "cached_values_input_name": "attn_cached_values_10",
524
+ "cached_values_output_name": "attn_cached_values_out_10",
525
+ "cached_positions_input_name": "attn_cached_positions_10",
526
+ "cached_positions_output_name": "attn_cached_positions_out_10",
527
+ "offset_shape": [
528
+ 1
529
+ ],
530
+ "cache_shape": [
531
+ 1,
532
+ 4,
533
+ 1600,
534
+ 64
535
+ ],
536
+ "positions_shape": [
537
+ 1,
538
+ 1600
539
+ ],
540
+ "cache_dtype": "float32",
541
+ "positions_dtype": "int32"
542
+ },
543
+ {
544
+ "index": 11,
545
+ "decoder_index": 7,
546
+ "layer_index": 3,
547
+ "context": 1600,
548
+ "num_heads": 4,
549
+ "head_dim": 64,
550
+ "offset_input_name": "attn_offset_11",
551
+ "offset_output_name": "attn_offset_out_11",
552
+ "cached_keys_input_name": "attn_cached_keys_11",
553
+ "cached_keys_output_name": "attn_cached_keys_out_11",
554
+ "cached_values_input_name": "attn_cached_values_11",
555
+ "cached_values_output_name": "attn_cached_values_out_11",
556
+ "cached_positions_input_name": "attn_cached_positions_11",
557
+ "cached_positions_output_name": "attn_cached_positions_out_11",
558
+ "offset_shape": [
559
+ 1
560
+ ],
561
+ "cache_shape": [
562
+ 1,
563
+ 4,
564
+ 1600,
565
+ 64
566
+ ],
567
+ "positions_shape": [
568
+ 1,
569
+ 1600
570
+ ],
571
+ "cache_dtype": "float32",
572
+ "positions_dtype": "int32"
573
+ }
574
+ ]
575
+ }
576
+ }
app/codec/moss_audio_tokenizer_decode_step.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fe4a66337450ac83e1cea4987a8847c6f016e2dfb209c4907c24834124ea180
3
+ size 22500094
app/codec/moss_audio_tokenizer_encode.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc8109d062700c8ce3fd8d0f887fe929b80cafba3f282a7f7d171635c73fe181
3
+ size 23117167
app/g2p/sea_g2p.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4346e690d0711ebc5231e7a42c5c88aaf6e40377e894b4617c018fd81c6f4096
3
+ size 62829820
app/matcha/date-zh.fst ADDED
app/matcha/lexicon.txt ADDED
The diff for this file is too large to render. See raw diff
 
app/matcha/model-steps-3.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85de262e182fa64a2ba88e88940b5ffb552858902d998382cfa314c620bfd22e
3
+ size 38642571
app/matcha/number-zh.fst ADDED
app/matcha/phone-zh.fst ADDED
app/matcha/tokens.txt ADDED
@@ -0,0 +1,2190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 1
2
+ ; 2
3
+ : 3
4
+ , 4
5
+ . 5
6
+ ! 6
7
+ ? 7
8
+ — 8
9
+ … 9
10
+ " 10
11
+ ( 11
12
+ ) 12
13
+ “ 13
14
+ ” 14
15
+ ̃ 15
16
+ ʣ 16
17
+ ʥ 17
18
+ ʦ 18
19
+ ʨ 19
20
+ ᵝ 20
21
+ ꭧ 21
22
+ A 22
23
+ I 23
24
+ O 24
25
+ Q 25
26
+ S 26
27
+ T 27
28
+ W 28
29
+ Y 29
30
+ ᵊ 30
31
+ a 31
32
+ b 32
33
+ c 33
34
+ d 34
35
+ e 35
36
+ f 36
37
+ h 37
38
+ i 38
39
+ j 39
40
+ k 40
41
+ l 41
42
+ m 42
43
+ n 43
44
+ o 44
45
+ p 45
46
+ q 46
47
+ r 47
48
+ s 48
49
+ t 49
50
+ u 50
51
+ v 51
52
+ w 52
53
+ x 53
54
+ y 54
55
+ z 55
56
+ ɑ 56
57
+ ɐ 57
58
+ ɒ 58
59
+ æ 59
60
+ β 60
61
+ ɔ 61
62
+ ɕ 62
63
+ ç 63
64
+ ɖ 64
65
+ ð 65
66
+ ʤ 66
67
+ ə 67
68
+ ɚ 68
69
+ ɛ 69
70
+ ɜ 70
71
+ ɟ 71
72
+ ɡ 72
73
+ ɥ 73
74
+ ɨ 74
75
+ ɪ 75
76
+ ʝ 76
77
+ ɯ 77
78
+ ɰ 78
79
+ ŋ 79
80
+ ɳ 80
81
+ ɲ 81
82
+ ɴ 82
83
+ ø 83
84
+ ɸ 84
85
+ θ 85
86
+ œ 86
87
+ ɹ 87
88
+ ɾ 88
89
+ ɻ 89
90
+ ʁ 90
91
+ ɽ 91
92
+ ʂ 92
93
+ ʃ 93
94
+ ʈ 94
95
+ ʧ 95
96
+ ʊ 96
97
+ ʋ 97
98
+ ʌ 98
99
+ ɣ 99
100
+ ɤ 100
101
+ χ 101
102
+ ʎ 102
103
+ ʒ 103
104
+ ʔ 104
105
+ ˈ 105
106
+ ˌ 106
107
+ ː 107
108
+ ʰ 108
109
+ ʲ 109
110
+ ↓ 110
111
+ → 111
112
+ ↗ 112
113
+ ↘ 113
114
+ ᵻ 114
115
+ a5 115
116
+ a1 116
117
+ a2 117
118
+ a3 118
119
+ a4 119
120
+ ai5 120
121
+ ai1 121
122
+ ai2 122
123
+ ai3 123
124
+ ai4 124
125
+ an5 125
126
+ an1 126
127
+ an2 127
128
+ an3 128
129
+ an4 129
130
+ ang5 130
131
+ ang1 131
132
+ ang2 132
133
+ ang3 133
134
+ ang4 134
135
+ ao5 135
136
+ ao1 136
137
+ ao2 137
138
+ ao3 138
139
+ ao4 139
140
+ ba5 140
141
+ ba1 141
142
+ ba2 142
143
+ ba3 143
144
+ ba4 144
145
+ bai5 145
146
+ bai1 146
147
+ bai2 147
148
+ bai3 148
149
+ bai4 149
150
+ ban5 150
151
+ ban1 151
152
+ ban2 152
153
+ ban3 153
154
+ ban4 154
155
+ bang5 155
156
+ bang1 156
157
+ bang2 157
158
+ bang3 158
159
+ bang4 159
160
+ bao5 160
161
+ bao1 161
162
+ bao2 162
163
+ bao3 163
164
+ bao4 164
165
+ bei5 165
166
+ bei1 166
167
+ bei2 167
168
+ bei3 168
169
+ bei4 169
170
+ ben5 170
171
+ ben1 171
172
+ ben2 172
173
+ ben3 173
174
+ ben4 174
175
+ beng5 175
176
+ beng1 176
177
+ beng2 177
178
+ beng3 178
179
+ beng4 179
180
+ bi5 180
181
+ bi1 181
182
+ bi2 182
183
+ bi3 183
184
+ bi4 184
185
+ bian5 185
186
+ bian1 186
187
+ bian2 187
188
+ bian3 188
189
+ bian4 189
190
+ biao5 190
191
+ biao1 191
192
+ biao2 192
193
+ biao3 193
194
+ biao4 194
195
+ bie5 195
196
+ bie1 196
197
+ bie2 197
198
+ bie3 198
199
+ bie4 199
200
+ bin5 200
201
+ bin1 201
202
+ bin2 202
203
+ bin3 203
204
+ bin4 204
205
+ bing5 205
206
+ bing1 206
207
+ bing2 207
208
+ bing3 208
209
+ bing4 209
210
+ bo5 210
211
+ bo1 211
212
+ bo2 212
213
+ bo3 213
214
+ bo4 214
215
+ bu5 215
216
+ bu1 216
217
+ bu2 217
218
+ bu3 218
219
+ bu4 219
220
+ ca5 220
221
+ ca1 221
222
+ ca2 222
223
+ ca3 223
224
+ ca4 224
225
+ cai5 225
226
+ cai1 226
227
+ cai2 227
228
+ cai3 228
229
+ cai4 229
230
+ can5 230
231
+ can1 231
232
+ can2 232
233
+ can3 233
234
+ can4 234
235
+ cang5 235
236
+ cang1 236
237
+ cang2 237
238
+ cang3 238
239
+ cang4 239
240
+ cao5 240
241
+ cao1 241
242
+ cao2 242
243
+ cao3 243
244
+ cao4 244
245
+ ce5 245
246
+ ce1 246
247
+ ce2 247
248
+ ce3 248
249
+ ce4 249
250
+ cen5 250
251
+ cen1 251
252
+ cen2 252
253
+ cen3 253
254
+ cen4 254
255
+ ceng5 255
256
+ ceng1 256
257
+ ceng2 257
258
+ ceng3 258
259
+ ceng4 259
260
+ cha5 260
261
+ cha1 261
262
+ cha2 262
263
+ cha3 263
264
+ cha4 264
265
+ chai5 265
266
+ chai1 266
267
+ chai2 267
268
+ chai3 268
269
+ chai4 269
270
+ chan5 270
271
+ chan1 271
272
+ chan2 272
273
+ chan3 273
274
+ chan4 274
275
+ chang5 275
276
+ chang1 276
277
+ chang2 277
278
+ chang3 278
279
+ chang4 279
280
+ chao5 280
281
+ chao1 281
282
+ chao2 282
283
+ chao3 283
284
+ chao4 284
285
+ che5 285
286
+ che1 286
287
+ che2 287
288
+ che3 288
289
+ che4 289
290
+ chen5 290
291
+ chen1 291
292
+ chen2 292
293
+ chen3 293
294
+ chen4 294
295
+ cheng5 295
296
+ cheng1 296
297
+ cheng2 297
298
+ cheng3 298
299
+ cheng4 299
300
+ chi5 300
301
+ chi1 301
302
+ chi2 302
303
+ chi3 303
304
+ chi4 304
305
+ chong5 305
306
+ chong1 306
307
+ chong2 307
308
+ chong3 308
309
+ chong4 309
310
+ chou5 310
311
+ chou1 311
312
+ chou2 312
313
+ chou3 313
314
+ chou4 314
315
+ chu5 315
316
+ chu1 316
317
+ chu2 317
318
+ chu3 318
319
+ chu4 319
320
+ chua5 320
321
+ chua1 321
322
+ chua2 322
323
+ chua3 323
324
+ chua4 324
325
+ chuai5 325
326
+ chuai1 326
327
+ chuai2 327
328
+ chuai3 328
329
+ chuai4 329
330
+ chuan5 330
331
+ chuan1 331
332
+ chuan2 332
333
+ chuan3 333
334
+ chuan4 334
335
+ chuang5 335
336
+ chuang1 336
337
+ chuang2 337
338
+ chuang3 338
339
+ chuang4 339
340
+ chui5 340
341
+ chui1 341
342
+ chui2 342
343
+ chui3 343
344
+ chui4 344
345
+ chun5 345
346
+ chun1 346
347
+ chun2 347
348
+ chun3 348
349
+ chun4 349
350
+ chuo5 350
351
+ chuo1 351
352
+ chuo2 352
353
+ chuo3 353
354
+ chuo4 354
355
+ ci5 355
356
+ ci1 356
357
+ ci2 357
358
+ ci3 358
359
+ ci4 359
360
+ cong5 360
361
+ cong1 361
362
+ cong2 362
363
+ cong3 363
364
+ cong4 364
365
+ cou5 365
366
+ cou1 366
367
+ cou2 367
368
+ cou3 368
369
+ cou4 369
370
+ cu5 370
371
+ cu1 371
372
+ cu2 372
373
+ cu3 373
374
+ cu4 374
375
+ cuan5 375
376
+ cuan1 376
377
+ cuan2 377
378
+ cuan3 378
379
+ cuan4 379
380
+ cui5 380
381
+ cui1 381
382
+ cui2 382
383
+ cui3 383
384
+ cui4 384
385
+ cun5 385
386
+ cun1 386
387
+ cun2 387
388
+ cun3 388
389
+ cun4 389
390
+ cuo5 390
391
+ cuo1 391
392
+ cuo2 392
393
+ cuo3 393
394
+ cuo4 394
395
+ da5 395
396
+ da1 396
397
+ da2 397
398
+ da3 398
399
+ da4 399
400
+ dai5 400
401
+ dai1 401
402
+ dai2 402
403
+ dai3 403
404
+ dai4 404
405
+ dan5 405
406
+ dan1 406
407
+ dan2 407
408
+ dan3 408
409
+ dan4 409
410
+ dang5 410
411
+ dang1 411
412
+ dang2 412
413
+ dang3 413
414
+ dang4 414
415
+ dao5 415
416
+ dao1 416
417
+ dao2 417
418
+ dao3 418
419
+ dao4 419
420
+ de5 420
421
+ de1 421
422
+ de2 422
423
+ de3 423
424
+ de4 424
425
+ dei5 425
426
+ dei1 426
427
+ dei2 427
428
+ dei3 428
429
+ dei4 429
430
+ den5 430
431
+ den1 431
432
+ den2 432
433
+ den3 433
434
+ den4 434
435
+ deng5 435
436
+ deng1 436
437
+ deng2 437
438
+ deng3 438
439
+ deng4 439
440
+ di5 440
441
+ di1 441
442
+ di2 442
443
+ di3 443
444
+ di4 444
445
+ dia5 445
446
+ dia1 446
447
+ dia2 447
448
+ dia3 448
449
+ dia4 449
450
+ dian5 450
451
+ dian1 451
452
+ dian2 452
453
+ dian3 453
454
+ dian4 454
455
+ diao5 455
456
+ diao1 456
457
+ diao2 457
458
+ diao3 458
459
+ diao4 459
460
+ die5 460
461
+ die1 461
462
+ die2 462
463
+ die3 463
464
+ die4 464
465
+ ding5 465
466
+ ding1 466
467
+ ding2 467
468
+ ding3 468
469
+ ding4 469
470
+ diu5 470
471
+ diu1 471
472
+ diu2 472
473
+ diu3 473
474
+ diu4 474
475
+ dong5 475
476
+ dong1 476
477
+ dong2 477
478
+ dong3 478
479
+ dong4 479
480
+ dou5 480
481
+ dou1 481
482
+ dou2 482
483
+ dou3 483
484
+ dou4 484
485
+ du5 485
486
+ du1 486
487
+ du2 487
488
+ du3 488
489
+ du4 489
490
+ duan5 490
491
+ duan1 491
492
+ duan2 492
493
+ duan3 493
494
+ duan4 494
495
+ dui5 495
496
+ dui1 496
497
+ dui2 497
498
+ dui3 498
499
+ dui4 499
500
+ dun5 500
501
+ dun1 501
502
+ dun2 502
503
+ dun3 503
504
+ dun4 504
505
+ duo5 505
506
+ duo1 506
507
+ duo2 507
508
+ duo3 508
509
+ duo4 509
510
+ e5 510
511
+ e1 511
512
+ e2 512
513
+ e3 513
514
+ e4 514
515
+ ei5 515
516
+ ei1 516
517
+ ei2 517
518
+ ei3 518
519
+ ei4 519
520
+ en5 520
521
+ en1 521
522
+ en2 522
523
+ en3 523
524
+ en4 524
525
+ eng5 525
526
+ eng1 526
527
+ eng2 527
528
+ eng3 528
529
+ eng4 529
530
+ er5 530
531
+ er1 531
532
+ er2 532
533
+ er3 533
534
+ er4 534
535
+ fa5 535
536
+ fa1 536
537
+ fa2 537
538
+ fa3 538
539
+ fa4 539
540
+ fan5 540
541
+ fan1 541
542
+ fan2 542
543
+ fan3 543
544
+ fan4 544
545
+ fang5 545
546
+ fang1 546
547
+ fang2 547
548
+ fang3 548
549
+ fang4 549
550
+ fei5 550
551
+ fei1 551
552
+ fei2 552
553
+ fei3 553
554
+ fei4 554
555
+ fen5 555
556
+ fen1 556
557
+ fen2 557
558
+ fen3 558
559
+ fen4 559
560
+ feng5 560
561
+ feng1 561
562
+ feng2 562
563
+ feng3 563
564
+ feng4 564
565
+ fiao5 565
566
+ fiao1 566
567
+ fiao2 567
568
+ fiao3 568
569
+ fiao4 569
570
+ fo5 570
571
+ fo1 571
572
+ fo2 572
573
+ fo3 573
574
+ fo4 574
575
+ fou5 575
576
+ fou1 576
577
+ fou2 577
578
+ fou3 578
579
+ fou4 579
580
+ fu5 580
581
+ fu1 581
582
+ fu2 582
583
+ fu3 583
584
+ fu4 584
585
+ ga5 585
586
+ ga1 586
587
+ ga2 587
588
+ ga3 588
589
+ ga4 589
590
+ gai5 590
591
+ gai1 591
592
+ gai2 592
593
+ gai3 593
594
+ gai4 594
595
+ gan5 595
596
+ gan1 596
597
+ gan2 597
598
+ gan3 598
599
+ gan4 599
600
+ gang5 600
601
+ gang1 601
602
+ gang2 602
603
+ gang3 603
604
+ gang4 604
605
+ gao5 605
606
+ gao1 606
607
+ gao2 607
608
+ gao3 608
609
+ gao4 609
610
+ ge5 610
611
+ ge1 611
612
+ ge2 612
613
+ ge3 613
614
+ ge4 614
615
+ gei5 615
616
+ gei1 616
617
+ gei2 617
618
+ gei3 618
619
+ gei4 619
620
+ gen5 620
621
+ gen1 621
622
+ gen2 622
623
+ gen3 623
624
+ gen4 624
625
+ geng5 625
626
+ geng1 626
627
+ geng2 627
628
+ geng3 628
629
+ geng4 629
630
+ gong5 630
631
+ gong1 631
632
+ gong2 632
633
+ gong3 633
634
+ gong4 634
635
+ gou5 635
636
+ gou1 636
637
+ gou2 637
638
+ gou3 638
639
+ gou4 639
640
+ gu5 640
641
+ gu1 641
642
+ gu2 642
643
+ gu3 643
644
+ gu4 644
645
+ gua5 645
646
+ gua1 646
647
+ gua2 647
648
+ gua3 648
649
+ gua4 649
650
+ guai5 650
651
+ guai1 651
652
+ guai2 652
653
+ guai3 653
654
+ guai4 654
655
+ guan5 655
656
+ guan1 656
657
+ guan2 657
658
+ guan3 658
659
+ guan4 659
660
+ guang5 660
661
+ guang1 661
662
+ guang2 662
663
+ guang3 663
664
+ guang4 664
665
+ gui5 665
666
+ gui1 666
667
+ gui2 667
668
+ gui3 668
669
+ gui4 669
670
+ gun5 670
671
+ gun1 671
672
+ gun2 672
673
+ gun3 673
674
+ gun4 674
675
+ guo5 675
676
+ guo1 676
677
+ guo2 677
678
+ guo3 678
679
+ guo4 679
680
+ ha5 680
681
+ ha1 681
682
+ ha2 682
683
+ ha3 683
684
+ ha4 684
685
+ hai5 685
686
+ hai1 686
687
+ hai2 687
688
+ hai3 688
689
+ hai4 689
690
+ han5 690
691
+ han1 691
692
+ han2 692
693
+ han3 693
694
+ han4 694
695
+ hang5 695
696
+ hang1 696
697
+ hang2 697
698
+ hang3 698
699
+ hang4 699
700
+ hao5 700
701
+ hao1 701
702
+ hao2 702
703
+ hao3 703
704
+ hao4 704
705
+ he5 705
706
+ he1 706
707
+ he2 707
708
+ he3 708
709
+ he4 709
710
+ hei5 710
711
+ hei1 711
712
+ hei2 712
713
+ hei3 713
714
+ hei4 714
715
+ hen5 715
716
+ hen1 716
717
+ hen2 717
718
+ hen3 718
719
+ hen4 719
720
+ heng5 720
721
+ heng1 721
722
+ heng2 722
723
+ heng3 723
724
+ heng4 724
725
+ hm5 725
726
+ hm1 726
727
+ hm2 727
728
+ hm3 728
729
+ hm4 729
730
+ hong5 730
731
+ hong1 731
732
+ hong2 732
733
+ hong3 733
734
+ hong4 734
735
+ hou5 735
736
+ hou1 736
737
+ hou2 737
738
+ hou3 738
739
+ hou4 739
740
+ hu5 740
741
+ hu1 741
742
+ hu2 742
743
+ hu3 743
744
+ hu4 744
745
+ hua5 745
746
+ hua1 746
747
+ hua2 747
748
+ hua3 748
749
+ hua4 749
750
+ huai5 750
751
+ huai1 751
752
+ huai2 752
753
+ huai3 753
754
+ huai4 754
755
+ huan5 755
756
+ huan1 756
757
+ huan2 757
758
+ huan3 758
759
+ huan4 759
760
+ huang5 760
761
+ huang1 761
762
+ huang2 762
763
+ huang3 763
764
+ huang4 764
765
+ hui5 765
766
+ hui1 766
767
+ hui2 767
768
+ hui3 768
769
+ hui4 769
770
+ hun5 770
771
+ hun1 771
772
+ hun2 772
773
+ hun3 773
774
+ hun4 774
775
+ huo5 775
776
+ huo1 776
777
+ huo2 777
778
+ huo3 778
779
+ huo4 779
780
+ ji5 780
781
+ ji1 781
782
+ ji2 782
783
+ ji3 783
784
+ ji4 784
785
+ jia5 785
786
+ jia1 786
787
+ jia2 787
788
+ jia3 788
789
+ jia4 789
790
+ jian5 790
791
+ jian1 791
792
+ jian2 792
793
+ jian3 793
794
+ jian4 794
795
+ jiang5 795
796
+ jiang1 796
797
+ jiang2 797
798
+ jiang3 798
799
+ jiang4 799
800
+ jiao5 800
801
+ jiao1 801
802
+ jiao2 802
803
+ jiao3 803
804
+ jiao4 804
805
+ jie5 805
806
+ jie1 806
807
+ jie2 807
808
+ jie3 808
809
+ jie4 809
810
+ jin5 810
811
+ jin1 811
812
+ jin2 812
813
+ jin3 813
814
+ jin4 814
815
+ jing5 815
816
+ jing1 816
817
+ jing2 817
818
+ jing3 818
819
+ jing4 819
820
+ jiong5 820
821
+ jiong1 821
822
+ jiong2 822
823
+ jiong3 823
824
+ jiong4 824
825
+ jiu5 825
826
+ jiu1 826
827
+ jiu2 827
828
+ jiu3 828
829
+ jiu4 829
830
+ ju5 830
831
+ ju1 831
832
+ ju2 832
833
+ ju3 833
834
+ ju4 834
835
+ juan5 835
836
+ juan1 836
837
+ juan2 837
838
+ juan3 838
839
+ juan4 839
840
+ jue5 840
841
+ jue1 841
842
+ jue2 842
843
+ jue3 843
844
+ jue4 844
845
+ jun5 845
846
+ jun1 846
847
+ jun2 847
848
+ jun3 848
849
+ jun4 849
850
+ ka5 850
851
+ ka1 851
852
+ ka2 852
853
+ ka3 853
854
+ ka4 854
855
+ kai5 855
856
+ kai1 856
857
+ kai2 857
858
+ kai3 858
859
+ kai4 859
860
+ kan5 860
861
+ kan1 861
862
+ kan2 862
863
+ kan3 863
864
+ kan4 864
865
+ kang5 865
866
+ kang1 866
867
+ kang2 867
868
+ kang3 868
869
+ kang4 869
870
+ kao5 870
871
+ kao1 871
872
+ kao2 872
873
+ kao3 873
874
+ kao4 874
875
+ ke5 875
876
+ ke1 876
877
+ ke2 877
878
+ ke3 878
879
+ ke4 879
880
+ ken5 880
881
+ ken1 881
882
+ ken2 882
883
+ ken3 883
884
+ ken4 884
885
+ keng5 885
886
+ keng1 886
887
+ keng2 887
888
+ keng3 888
889
+ keng4 889
890
+ kong5 890
891
+ kong1 891
892
+ kong2 892
893
+ kong3 893
894
+ kong4 894
895
+ kou5 895
896
+ kou1 896
897
+ kou2 897
898
+ kou3 898
899
+ kou4 899
900
+ ku5 900
901
+ ku1 901
902
+ ku2 902
903
+ ku3 903
904
+ ku4 904
905
+ kua5 905
906
+ kua1 906
907
+ kua2 907
908
+ kua3 908
909
+ kua4 909
910
+ kuai5 910
911
+ kuai1 911
912
+ kuai2 912
913
+ kuai3 913
914
+ kuai4 914
915
+ kuan5 915
916
+ kuan1 916
917
+ kuan2 917
918
+ kuan3 918
919
+ kuan4 919
920
+ kuang5 920
921
+ kuang1 921
922
+ kuang2 922
923
+ kuang3 923
924
+ kuang4 924
925
+ kui5 925
926
+ kui1 926
927
+ kui2 927
928
+ kui3 928
929
+ kui4 929
930
+ kun5 930
931
+ kun1 931
932
+ kun2 932
933
+ kun3 933
934
+ kun4 934
935
+ kuo5 935
936
+ kuo1 936
937
+ kuo2 937
938
+ kuo3 938
939
+ kuo4 939
940
+ la5 940
941
+ la1 941
942
+ la2 942
943
+ la3 943
944
+ la4 944
945
+ lai5 945
946
+ lai1 946
947
+ lai2 947
948
+ lai3 948
949
+ lai4 949
950
+ lan5 950
951
+ lan1 951
952
+ lan2 952
953
+ lan3 953
954
+ lan4 954
955
+ lang5 955
956
+ lang1 956
957
+ lang2 957
958
+ lang3 958
959
+ lang4 959
960
+ lao5 960
961
+ lao1 961
962
+ lao2 962
963
+ lao3 963
964
+ lao4 964
965
+ le5 965
966
+ le1 966
967
+ le2 967
968
+ le3 968
969
+ le4 969
970
+ lei5 970
971
+ lei1 971
972
+ lei2 972
973
+ lei3 973
974
+ lei4 974
975
+ leng5 975
976
+ leng1 976
977
+ leng2 977
978
+ leng3 978
979
+ leng4 979
980
+ li5 980
981
+ li1 981
982
+ li2 982
983
+ li3 983
984
+ li4 984
985
+ lia5 985
986
+ lia1 986
987
+ lia2 987
988
+ lia3 988
989
+ lia4 989
990
+ lian5 990
991
+ lian1 991
992
+ lian2 992
993
+ lian3 993
994
+ lian4 994
995
+ liang5 995
996
+ liang1 996
997
+ liang2 997
998
+ liang3 998
999
+ liang4 999
1000
+ liao5 1000
1001
+ liao1 1001
1002
+ liao2 1002
1003
+ liao3 1003
1004
+ liao4 1004
1005
+ lie5 1005
1006
+ lie1 1006
1007
+ lie2 1007
1008
+ lie3 1008
1009
+ lie4 1009
1010
+ lin5 1010
1011
+ lin1 1011
1012
+ lin2 1012
1013
+ lin3 1013
1014
+ lin4 1014
1015
+ ling5 1015
1016
+ ling1 1016
1017
+ ling2 1017
1018
+ ling3 1018
1019
+ ling4 1019
1020
+ liu5 1020
1021
+ liu1 1021
1022
+ liu2 1022
1023
+ liu3 1023
1024
+ liu4 1024
1025
+ lo5 1025
1026
+ lo1 1026
1027
+ lo2 1027
1028
+ lo3 1028
1029
+ lo4 1029
1030
+ long5 1030
1031
+ long1 1031
1032
+ long2 1032
1033
+ long3 1033
1034
+ long4 1034
1035
+ lou5 1035
1036
+ lou1 1036
1037
+ lou2 1037
1038
+ lou3 1038
1039
+ lou4 1039
1040
+ lu5 1040
1041
+ lu1 1041
1042
+ lu2 1042
1043
+ lu3 1043
1044
+ lu4 1044
1045
+ luan5 1045
1046
+ luan1 1046
1047
+ luan2 1047
1048
+ luan3 1048
1049
+ luan4 1049
1050
+ lun5 1050
1051
+ lun1 1051
1052
+ lun2 1052
1053
+ lun3 1053
1054
+ lun4 1054
1055
+ luo5 1055
1056
+ luo1 1056
1057
+ luo2 1057
1058
+ luo3 1058
1059
+ luo4 1059
1060
+ lv5 1060
1061
+ lv1 1061
1062
+ lv2 1062
1063
+ lv3 1063
1064
+ lv4 1064
1065
+ lve5 1065
1066
+ lve1 1066
1067
+ lve2 1067
1068
+ lve3 1068
1069
+ lve4 1069
1070
+ m5 1070
1071
+ m1 1071
1072
+ m2 1072
1073
+ m3 1073
1074
+ m4 1074
1075
+ ma5 1075
1076
+ ma1 1076
1077
+ ma2 1077
1078
+ ma3 1078
1079
+ ma4 1079
1080
+ mai5 1080
1081
+ mai1 1081
1082
+ mai2 1082
1083
+ mai3 1083
1084
+ mai4 1084
1085
+ man5 1085
1086
+ man1 1086
1087
+ man2 1087
1088
+ man3 1088
1089
+ man4 1089
1090
+ mang5 1090
1091
+ mang1 1091
1092
+ mang2 1092
1093
+ mang3 1093
1094
+ mang4 1094
1095
+ mao5 1095
1096
+ mao1 1096
1097
+ mao2 1097
1098
+ mao3 1098
1099
+ mao4 1099
1100
+ me5 1100
1101
+ me1 1101
1102
+ me2 1102
1103
+ me3 1103
1104
+ me4 1104
1105
+ mei5 1105
1106
+ mei1 1106
1107
+ mei2 1107
1108
+ mei3 1108
1109
+ mei4 1109
1110
+ men5 1110
1111
+ men1 1111
1112
+ men2 1112
1113
+ men3 1113
1114
+ men4 1114
1115
+ meng5 1115
1116
+ meng1 1116
1117
+ meng2 1117
1118
+ meng3 1118
1119
+ meng4 1119
1120
+ mi5 1120
1121
+ mi1 1121
1122
+ mi2 1122
1123
+ mi3 1123
1124
+ mi4 1124
1125
+ mian5 1125
1126
+ mian1 1126
1127
+ mian2 1127
1128
+ mian3 1128
1129
+ mian4 1129
1130
+ miao5 1130
1131
+ miao1 1131
1132
+ miao2 1132
1133
+ miao3 1133
1134
+ miao4 1134
1135
+ mie5 1135
1136
+ mie1 1136
1137
+ mie2 1137
1138
+ mie3 1138
1139
+ mie4 1139
1140
+ min5 1140
1141
+ min1 1141
1142
+ min2 1142
1143
+ min3 1143
1144
+ min4 1144
1145
+ ming5 1145
1146
+ ming1 1146
1147
+ ming2 1147
1148
+ ming3 1148
1149
+ ming4 1149
1150
+ miu5 1150
1151
+ miu1 1151
1152
+ miu2 1152
1153
+ miu3 1153
1154
+ miu4 1154
1155
+ mo5 1155
1156
+ mo1 1156
1157
+ mo2 1157
1158
+ mo3 1158
1159
+ mo4 1159
1160
+ mou5 1160
1161
+ mou1 1161
1162
+ mou2 1162
1163
+ mou3 1163
1164
+ mou4 1164
1165
+ mu5 1165
1166
+ mu1 1166
1167
+ mu2 1167
1168
+ mu3 1168
1169
+ mu4 1169
1170
+ n5 1170
1171
+ n1 1171
1172
+ n2 1172
1173
+ n3 1173
1174
+ n4 1174
1175
+ na5 1175
1176
+ na1 1176
1177
+ na2 1177
1178
+ na3 1178
1179
+ na4 1179
1180
+ nai5 1180
1181
+ nai1 1181
1182
+ nai2 1182
1183
+ nai3 1183
1184
+ nai4 1184
1185
+ nan5 1185
1186
+ nan1 1186
1187
+ nan2 1187
1188
+ nan3 1188
1189
+ nan4 1189
1190
+ nang5 1190
1191
+ nang1 1191
1192
+ nang2 1192
1193
+ nang3 1193
1194
+ nang4 1194
1195
+ nao5 1195
1196
+ nao1 1196
1197
+ nao2 1197
1198
+ nao3 1198
1199
+ nao4 1199
1200
+ ne5 1200
1201
+ ne1 1201
1202
+ ne2 1202
1203
+ ne3 1203
1204
+ ne4 1204
1205
+ nei5 1205
1206
+ nei1 1206
1207
+ nei2 1207
1208
+ nei3 1208
1209
+ nei4 1209
1210
+ nen5 1210
1211
+ nen1 1211
1212
+ nen2 1212
1213
+ nen3 1213
1214
+ nen4 1214
1215
+ neng5 1215
1216
+ neng1 1216
1217
+ neng2 1217
1218
+ neng3 1218
1219
+ neng4 1219
1220
+ ni5 1220
1221
+ ni1 1221
1222
+ ni2 1222
1223
+ ni3 1223
1224
+ ni4 1224
1225
+ nian5 1225
1226
+ nian1 1226
1227
+ nian2 1227
1228
+ nian3 1228
1229
+ nian4 1229
1230
+ niang5 1230
1231
+ niang1 1231
1232
+ niang2 1232
1233
+ niang3 1233
1234
+ niang4 1234
1235
+ niao5 1235
1236
+ niao1 1236
1237
+ niao2 1237
1238
+ niao3 1238
1239
+ niao4 1239
1240
+ nie5 1240
1241
+ nie1 1241
1242
+ nie2 1242
1243
+ nie3 1243
1244
+ nie4 1244
1245
+ nin5 1245
1246
+ nin1 1246
1247
+ nin2 1247
1248
+ nin3 1248
1249
+ nin4 1249
1250
+ ning5 1250
1251
+ ning1 1251
1252
+ ning2 1252
1253
+ ning3 1253
1254
+ ning4 1254
1255
+ niu5 1255
1256
+ niu1 1256
1257
+ niu2 1257
1258
+ niu3 1258
1259
+ niu4 1259
1260
+ nong5 1260
1261
+ nong1 1261
1262
+ nong2 1262
1263
+ nong3 1263
1264
+ nong4 1264
1265
+ nou5 1265
1266
+ nou1 1266
1267
+ nou2 1267
1268
+ nou3 1268
1269
+ nou4 1269
1270
+ nu5 1270
1271
+ nu1 1271
1272
+ nu2 1272
1273
+ nu3 1273
1274
+ nu4 1274
1275
+ nuan5 1275
1276
+ nuan1 1276
1277
+ nuan2 1277
1278
+ nuan3 1278
1279
+ nuan4 1279
1280
+ nun5 1280
1281
+ nun1 1281
1282
+ nun2 1282
1283
+ nun3 1283
1284
+ nun4 1284
1285
+ nuo5 1285
1286
+ nuo1 1286
1287
+ nuo2 1287
1288
+ nuo3 1288
1289
+ nuo4 1289
1290
+ nv5 1290
1291
+ nv1 1291
1292
+ nv2 1292
1293
+ nv3 1293
1294
+ nv4 1294
1295
+ nve5 1295
1296
+ nve1 1296
1297
+ nve2 1297
1298
+ nve3 1298
1299
+ nve4 1299
1300
+ o5 1300
1301
+ o1 1301
1302
+ o2 1302
1303
+ o3 1303
1304
+ o4 1304
1305
+ ou5 1305
1306
+ ou1 1306
1307
+ ou2 1307
1308
+ ou3 1308
1309
+ ou4 1309
1310
+ pa5 1310
1311
+ pa1 1311
1312
+ pa2 1312
1313
+ pa3 1313
1314
+ pa4 1314
1315
+ pai5 1315
1316
+ pai1 1316
1317
+ pai2 1317
1318
+ pai3 1318
1319
+ pai4 1319
1320
+ pan5 1320
1321
+ pan1 1321
1322
+ pan2 1322
1323
+ pan3 1323
1324
+ pan4 1324
1325
+ pang5 1325
1326
+ pang1 1326
1327
+ pang2 1327
1328
+ pang3 1328
1329
+ pang4 1329
1330
+ pao5 1330
1331
+ pao1 1331
1332
+ pao2 1332
1333
+ pao3 1333
1334
+ pao4 1334
1335
+ pei5 1335
1336
+ pei1 1336
1337
+ pei2 1337
1338
+ pei3 1338
1339
+ pei4 1339
1340
+ pen5 1340
1341
+ pen1 1341
1342
+ pen2 1342
1343
+ pen3 1343
1344
+ pen4 1344
1345
+ peng5 1345
1346
+ peng1 1346
1347
+ peng2 1347
1348
+ peng3 1348
1349
+ peng4 1349
1350
+ pi5 1350
1351
+ pi1 1351
1352
+ pi2 1352
1353
+ pi3 1353
1354
+ pi4 1354
1355
+ pian5 1355
1356
+ pian1 1356
1357
+ pian2 1357
1358
+ pian3 1358
1359
+ pian4 1359
1360
+ piao5 1360
1361
+ piao1 1361
1362
+ piao2 1362
1363
+ piao3 1363
1364
+ piao4 1364
1365
+ pie5 1365
1366
+ pie1 1366
1367
+ pie2 1367
1368
+ pie3 1368
1369
+ pie4 1369
1370
+ pin5 1370
1371
+ pin1 1371
1372
+ pin2 1372
1373
+ pin3 1373
1374
+ pin4 1374
1375
+ ping5 1375
1376
+ ping1 1376
1377
+ ping2 1377
1378
+ ping3 1378
1379
+ ping4 1379
1380
+ po5 1380
1381
+ po1 1381
1382
+ po2 1382
1383
+ po3 1383
1384
+ po4 1384
1385
+ pou5 1385
1386
+ pou1 1386
1387
+ pou2 1387
1388
+ pou3 1388
1389
+ pou4 1389
1390
+ pu5 1390
1391
+ pu1 1391
1392
+ pu2 1392
1393
+ pu3 1393
1394
+ pu4 1394
1395
+ qi5 1395
1396
+ qi1 1396
1397
+ qi2 1397
1398
+ qi3 1398
1399
+ qi4 1399
1400
+ qia5 1400
1401
+ qia1 1401
1402
+ qia2 1402
1403
+ qia3 1403
1404
+ qia4 1404
1405
+ qian5 1405
1406
+ qian1 1406
1407
+ qian2 1407
1408
+ qian3 1408
1409
+ qian4 1409
1410
+ qiang5 1410
1411
+ qiang1 1411
1412
+ qiang2 1412
1413
+ qiang3 1413
1414
+ qiang4 1414
1415
+ qiao5 1415
1416
+ qiao1 1416
1417
+ qiao2 1417
1418
+ qiao3 1418
1419
+ qiao4 1419
1420
+ qie5 1420
1421
+ qie1 1421
1422
+ qie2 1422
1423
+ qie3 1423
1424
+ qie4 1424
1425
+ qin5 1425
1426
+ qin1 1426
1427
+ qin2 1427
1428
+ qin3 1428
1429
+ qin4 1429
1430
+ qing5 1430
1431
+ qing1 1431
1432
+ qing2 1432
1433
+ qing3 1433
1434
+ qing4 1434
1435
+ qiong5 1435
1436
+ qiong1 1436
1437
+ qiong2 1437
1438
+ qiong3 1438
1439
+ qiong4 1439
1440
+ qiu5 1440
1441
+ qiu1 1441
1442
+ qiu2 1442
1443
+ qiu3 1443
1444
+ qiu4 1444
1445
+ qu5 1445
1446
+ qu1 1446
1447
+ qu2 1447
1448
+ qu3 1448
1449
+ qu4 1449
1450
+ quan5 1450
1451
+ quan1 1451
1452
+ quan2 1452
1453
+ quan3 1453
1454
+ quan4 1454
1455
+ que5 1455
1456
+ que1 1456
1457
+ que2 1457
1458
+ que3 1458
1459
+ que4 1459
1460
+ qun5 1460
1461
+ qun1 1461
1462
+ qun2 1462
1463
+ qun3 1463
1464
+ qun4 1464
1465
+ ran5 1465
1466
+ ran1 1466
1467
+ ran2 1467
1468
+ ran3 1468
1469
+ ran4 1469
1470
+ rang5 1470
1471
+ rang1 1471
1472
+ rang2 1472
1473
+ rang3 1473
1474
+ rang4 1474
1475
+ rao5 1475
1476
+ rao1 1476
1477
+ rao2 1477
1478
+ rao3 1478
1479
+ rao4 1479
1480
+ re5 1480
1481
+ re1 1481
1482
+ re2 1482
1483
+ re3 1483
1484
+ re4 1484
1485
+ ren5 1485
1486
+ ren1 1486
1487
+ ren2 1487
1488
+ ren3 1488
1489
+ ren4 1489
1490
+ reng5 1490
1491
+ reng1 1491
1492
+ reng2 1492
1493
+ reng3 1493
1494
+ reng4 1494
1495
+ ri5 1495
1496
+ ri1 1496
1497
+ ri2 1497
1498
+ ri3 1498
1499
+ ri4 1499
1500
+ rong5 1500
1501
+ rong1 1501
1502
+ rong2 1502
1503
+ rong3 1503
1504
+ rong4 1504
1505
+ rou5 1505
1506
+ rou1 1506
1507
+ rou2 1507
1508
+ rou3 1508
1509
+ rou4 1509
1510
+ ru5 1510
1511
+ ru1 1511
1512
+ ru2 1512
1513
+ ru3 1513
1514
+ ru4 1514
1515
+ rua5 1515
1516
+ rua1 1516
1517
+ rua2 1517
1518
+ rua3 1518
1519
+ rua4 1519
1520
+ ruan5 1520
1521
+ ruan1 1521
1522
+ ruan2 1522
1523
+ ruan3 1523
1524
+ ruan4 1524
1525
+ rui5 1525
1526
+ rui1 1526
1527
+ rui2 1527
1528
+ rui3 1528
1529
+ rui4 1529
1530
+ run5 1530
1531
+ run1 1531
1532
+ run2 1532
1533
+ run3 1533
1534
+ run4 1534
1535
+ ruo5 1535
1536
+ ruo1 1536
1537
+ ruo2 1537
1538
+ ruo3 1538
1539
+ ruo4 1539
1540
+ sa5 1540
1541
+ sa1 1541
1542
+ sa2 1542
1543
+ sa3 1543
1544
+ sa4 1544
1545
+ sai5 1545
1546
+ sai1 1546
1547
+ sai2 1547
1548
+ sai3 1548
1549
+ sai4 1549
1550
+ san5 1550
1551
+ san1 1551
1552
+ san2 1552
1553
+ san3 1553
1554
+ san4 1554
1555
+ sang5 1555
1556
+ sang1 1556
1557
+ sang2 1557
1558
+ sang3 1558
1559
+ sang4 1559
1560
+ sao5 1560
1561
+ sao1 1561
1562
+ sao2 1562
1563
+ sao3 1563
1564
+ sao4 1564
1565
+ se5 1565
1566
+ se1 1566
1567
+ se2 1567
1568
+ se3 1568
1569
+ se4 1569
1570
+ sen5 1570
1571
+ sen1 1571
1572
+ sen2 1572
1573
+ sen3 1573
1574
+ sen4 1574
1575
+ seng5 1575
1576
+ seng1 1576
1577
+ seng2 1577
1578
+ seng3 1578
1579
+ seng4 1579
1580
+ sha5 1580
1581
+ sha1 1581
1582
+ sha2 1582
1583
+ sha3 1583
1584
+ sha4 1584
1585
+ shai5 1585
1586
+ shai1 1586
1587
+ shai2 1587
1588
+ shai3 1588
1589
+ shai4 1589
1590
+ shan5 1590
1591
+ shan1 1591
1592
+ shan2 1592
1593
+ shan3 1593
1594
+ shan4 1594
1595
+ shang5 1595
1596
+ shang1 1596
1597
+ shang2 1597
1598
+ shang3 1598
1599
+ shang4 1599
1600
+ shao5 1600
1601
+ shao1 1601
1602
+ shao2 1602
1603
+ shao3 1603
1604
+ shao4 1604
1605
+ she5 1605
1606
+ she1 1606
1607
+ she2 1607
1608
+ she3 1608
1609
+ she4 1609
1610
+ shen5 1610
1611
+ shen1 1611
1612
+ shen2 1612
1613
+ shen3 1613
1614
+ shen4 1614
1615
+ sheng5 1615
1616
+ sheng1 1616
1617
+ sheng2 1617
1618
+ sheng3 1618
1619
+ sheng4 1619
1620
+ shi5 1620
1621
+ shi1 1621
1622
+ shi2 1622
1623
+ shi3 1623
1624
+ shi4 1624
1625
+ shou5 1625
1626
+ shou1 1626
1627
+ shou2 1627
1628
+ shou3 1628
1629
+ shou4 1629
1630
+ shu5 1630
1631
+ shu1 1631
1632
+ shu2 1632
1633
+ shu3 1633
1634
+ shu4 1634
1635
+ shua5 1635
1636
+ shua1 1636
1637
+ shua2 1637
1638
+ shua3 1638
1639
+ shua4 1639
1640
+ shuai5 1640
1641
+ shuai1 1641
1642
+ shuai2 1642
1643
+ shuai3 1643
1644
+ shuai4 1644
1645
+ shuan5 1645
1646
+ shuan1 1646
1647
+ shuan2 1647
1648
+ shuan3 1648
1649
+ shuan4 1649
1650
+ shuang5 1650
1651
+ shuang1 1651
1652
+ shuang2 1652
1653
+ shuang3 1653
1654
+ shuang4 1654
1655
+ shui5 1655
1656
+ shui1 1656
1657
+ shui2 1657
1658
+ shui3 1658
1659
+ shui4 1659
1660
+ shun5 1660
1661
+ shun1 1661
1662
+ shun2 1662
1663
+ shun3 1663
1664
+ shun4 1664
1665
+ shuo5 1665
1666
+ shuo1 1666
1667
+ shuo2 1667
1668
+ shuo3 1668
1669
+ shuo4 1669
1670
+ si5 1670
1671
+ si1 1671
1672
+ si2 1672
1673
+ si3 1673
1674
+ si4 1674
1675
+ song5 1675
1676
+ song1 1676
1677
+ song2 1677
1678
+ song3 1678
1679
+ song4 1679
1680
+ sou5 1680
1681
+ sou1 1681
1682
+ sou2 1682
1683
+ sou3 1683
1684
+ sou4 1684
1685
+ su5 1685
1686
+ su1 1686
1687
+ su2 1687
1688
+ su3 1688
1689
+ su4 1689
1690
+ suan5 1690
1691
+ suan1 1691
1692
+ suan2 1692
1693
+ suan3 1693
1694
+ suan4 1694
1695
+ sui5 1695
1696
+ sui1 1696
1697
+ sui2 1697
1698
+ sui3 1698
1699
+ sui4 1699
1700
+ sun5 1700
1701
+ sun1 1701
1702
+ sun2 1702
1703
+ sun3 1703
1704
+ sun4 1704
1705
+ suo5 1705
1706
+ suo1 1706
1707
+ suo2 1707
1708
+ suo3 1708
1709
+ suo4 1709
1710
+ ta5 1710
1711
+ ta1 1711
1712
+ ta2 1712
1713
+ ta3 1713
1714
+ ta4 1714
1715
+ tai5 1715
1716
+ tai1 1716
1717
+ tai2 1717
1718
+ tai3 1718
1719
+ tai4 1719
1720
+ tan5 1720
1721
+ tan1 1721
1722
+ tan2 1722
1723
+ tan3 1723
1724
+ tan4 1724
1725
+ tang5 1725
1726
+ tang1 1726
1727
+ tang2 1727
1728
+ tang3 1728
1729
+ tang4 1729
1730
+ tao5 1730
1731
+ tao1 1731
1732
+ tao2 1732
1733
+ tao3 1733
1734
+ tao4 1734
1735
+ te5 1735
1736
+ te1 1736
1737
+ te2 1737
1738
+ te3 1738
1739
+ te4 1739
1740
+ teng5 1740
1741
+ teng1 1741
1742
+ teng2 1742
1743
+ teng3 1743
1744
+ teng4 1744
1745
+ ti5 1745
1746
+ ti1 1746
1747
+ ti2 1747
1748
+ ti3 1748
1749
+ ti4 1749
1750
+ tian5 1750
1751
+ tian1 1751
1752
+ tian2 1752
1753
+ tian3 1753
1754
+ tian4 1754
1755
+ tiao5 1755
1756
+ tiao1 1756
1757
+ tiao2 1757
1758
+ tiao3 1758
1759
+ tiao4 1759
1760
+ tie5 1760
1761
+ tie1 1761
1762
+ tie2 1762
1763
+ tie3 1763
1764
+ tie4 1764
1765
+ ting5 1765
1766
+ ting1 1766
1767
+ ting2 1767
1768
+ ting3 1768
1769
+ ting4 1769
1770
+ tong5 1770
1771
+ tong1 1771
1772
+ tong2 1772
1773
+ tong3 1773
1774
+ tong4 1774
1775
+ tou5 1775
1776
+ tou1 1776
1777
+ tou2 1777
1778
+ tou3 1778
1779
+ tou4 1779
1780
+ tu5 1780
1781
+ tu1 1781
1782
+ tu2 1782
1783
+ tu3 1783
1784
+ tu4 1784
1785
+ tuan5 1785
1786
+ tuan1 1786
1787
+ tuan2 1787
1788
+ tuan3 1788
1789
+ tuan4 1789
1790
+ tui5 1790
1791
+ tui1 1791
1792
+ tui2 1792
1793
+ tui3 1793
1794
+ tui4 1794
1795
+ tun5 1795
1796
+ tun1 1796
1797
+ tun2 1797
1798
+ tun3 1798
1799
+ tun4 1799
1800
+ tuo5 1800
1801
+ tuo1 1801
1802
+ tuo2 1802
1803
+ tuo3 1803
1804
+ tuo4 1804
1805
+ wa5 1805
1806
+ wa1 1806
1807
+ wa2 1807
1808
+ wa3 1808
1809
+ wa4 1809
1810
+ wai5 1810
1811
+ wai1 1811
1812
+ wai2 1812
1813
+ wai3 1813
1814
+ wai4 1814
1815
+ wan5 1815
1816
+ wan1 1816
1817
+ wan2 1817
1818
+ wan3 1818
1819
+ wan4 1819
1820
+ wang5 1820
1821
+ wang1 1821
1822
+ wang2 1822
1823
+ wang3 1823
1824
+ wang4 1824
1825
+ wei5 1825
1826
+ wei1 1826
1827
+ wei2 1827
1828
+ wei3 1828
1829
+ wei4 1829
1830
+ wen5 1830
1831
+ wen1 1831
1832
+ wen2 1832
1833
+ wen3 1833
1834
+ wen4 1834
1835
+ weng5 1835
1836
+ weng1 1836
1837
+ weng2 1837
1838
+ weng3 1838
1839
+ weng4 1839
1840
+ wo5 1840
1841
+ wo1 1841
1842
+ wo2 1842
1843
+ wo3 1843
1844
+ wo4 1844
1845
+ wu5 1845
1846
+ wu1 1846
1847
+ wu2 1847
1848
+ wu3 1848
1849
+ wu4 1849
1850
+ xi5 1850
1851
+ xi1 1851
1852
+ xi2 1852
1853
+ xi3 1853
1854
+ xi4 1854
1855
+ xia5 1855
1856
+ xia1 1856
1857
+ xia2 1857
1858
+ xia3 1858
1859
+ xia4 1859
1860
+ xian5 1860
1861
+ xian1 1861
1862
+ xian2 1862
1863
+ xian3 1863
1864
+ xian4 1864
1865
+ xiang5 1865
1866
+ xiang1 1866
1867
+ xiang2 1867
1868
+ xiang3 1868
1869
+ xiang4 1869
1870
+ xiao5 1870
1871
+ xiao1 1871
1872
+ xiao2 1872
1873
+ xiao3 1873
1874
+ xiao4 1874
1875
+ xie5 1875
1876
+ xie1 1876
1877
+ xie2 1877
1878
+ xie3 1878
1879
+ xie4 1879
1880
+ xin5 1880
1881
+ xin1 1881
1882
+ xin2 1882
1883
+ xin3 1883
1884
+ xin4 1884
1885
+ xing5 1885
1886
+ xing1 1886
1887
+ xing2 1887
1888
+ xing3 1888
1889
+ xing4 1889
1890
+ xiong5 1890
1891
+ xiong1 1891
1892
+ xiong2 1892
1893
+ xiong3 1893
1894
+ xiong4 1894
1895
+ xiu5 1895
1896
+ xiu1 1896
1897
+ xiu2 1897
1898
+ xiu3 1898
1899
+ xiu4 1899
1900
+ xu5 1900
1901
+ xu1 1901
1902
+ xu2 1902
1903
+ xu3 1903
1904
+ xu4 1904
1905
+ xuan5 1905
1906
+ xuan1 1906
1907
+ xuan2 1907
1908
+ xuan3 1908
1909
+ xuan4 1909
1910
+ xue5 1910
1911
+ xue1 1911
1912
+ xue2 1912
1913
+ xue3 1913
1914
+ xue4 1914
1915
+ xun5 1915
1916
+ xun1 1916
1917
+ xun2 1917
1918
+ xun3 1918
1919
+ xun4 1919
1920
+ ya5 1920
1921
+ ya1 1921
1922
+ ya2 1922
1923
+ ya3 1923
1924
+ ya4 1924
1925
+ yan5 1925
1926
+ yan1 1926
1927
+ yan2 1927
1928
+ yan3 1928
1929
+ yan4 1929
1930
+ yang5 1930
1931
+ yang1 1931
1932
+ yang2 1932
1933
+ yang3 1933
1934
+ yang4 1934
1935
+ yao5 1935
1936
+ yao1 1936
1937
+ yao2 1937
1938
+ yao3 1938
1939
+ yao4 1939
1940
+ ye5 1940
1941
+ ye1 1941
1942
+ ye2 1942
1943
+ ye3 1943
1944
+ ye4 1944
1945
+ yi5 1945
1946
+ yi1 1946
1947
+ yi2 1947
1948
+ yi3 1948
1949
+ yi4 1949
1950
+ yin5 1950
1951
+ yin1 1951
1952
+ yin2 1952
1953
+ yin3 1953
1954
+ yin4 1954
1955
+ ying5 1955
1956
+ ying1 1956
1957
+ ying2 1957
1958
+ ying3 1958
1959
+ ying4 1959
1960
+ yo5 1960
1961
+ yo1 1961
1962
+ yo2 1962
1963
+ yo3 1963
1964
+ yo4 1964
1965
+ yong5 1965
1966
+ yong1 1966
1967
+ yong2 1967
1968
+ yong3 1968
1969
+ yong4 1969
1970
+ you5 1970
1971
+ you1 1971
1972
+ you2 1972
1973
+ you3 1973
1974
+ you4 1974
1975
+ yu5 1975
1976
+ yu1 1976
1977
+ yu2 1977
1978
+ yu3 1978
1979
+ yu4 1979
1980
+ yuan5 1980
1981
+ yuan1 1981
1982
+ yuan2 1982
1983
+ yuan3 1983
1984
+ yuan4 1984
1985
+ yue5 1985
1986
+ yue1 1986
1987
+ yue2 1987
1988
+ yue3 1988
1989
+ yue4 1989
1990
+ yun5 1990
1991
+ yun1 1991
1992
+ yun2 1992
1993
+ yun3 1993
1994
+ yun4 1994
1995
+ za5 1995
1996
+ za1 1996
1997
+ za2 1997
1998
+ za3 1998
1999
+ za4 1999
2000
+ zai5 2000
2001
+ zai1 2001
2002
+ zai2 2002
2003
+ zai3 2003
2004
+ zai4 2004
2005
+ zan5 2005
2006
+ zan1 2006
2007
+ zan2 2007
2008
+ zan3 2008
2009
+ zan4 2009
2010
+ zang5 2010
2011
+ zang1 2011
2012
+ zang2 2012
2013
+ zang3 2013
2014
+ zang4 2014
2015
+ zao5 2015
2016
+ zao1 2016
2017
+ zao2 2017
2018
+ zao3 2018
2019
+ zao4 2019
2020
+ ze5 2020
2021
+ ze1 2021
2022
+ ze2 2022
2023
+ ze3 2023
2024
+ ze4 2024
2025
+ zei5 2025
2026
+ zei1 2026
2027
+ zei2 2027
2028
+ zei3 2028
2029
+ zei4 2029
2030
+ zen5 2030
2031
+ zen1 2031
2032
+ zen2 2032
2033
+ zen3 2033
2034
+ zen4 2034
2035
+ zeng5 2035
2036
+ zeng1 2036
2037
+ zeng2 2037
2038
+ zeng3 2038
2039
+ zeng4 2039
2040
+ zha5 2040
2041
+ zha1 2041
2042
+ zha2 2042
2043
+ zha3 2043
2044
+ zha4 2044
2045
+ zhai5 2045
2046
+ zhai1 2046
2047
+ zhai2 2047
2048
+ zhai3 2048
2049
+ zhai4 2049
2050
+ zhan5 2050
2051
+ zhan1 2051
2052
+ zhan2 2052
2053
+ zhan3 2053
2054
+ zhan4 2054
2055
+ zhang5 2055
2056
+ zhang1 2056
2057
+ zhang2 2057
2058
+ zhang3 2058
2059
+ zhang4 2059
2060
+ zhao5 2060
2061
+ zhao1 2061
2062
+ zhao2 2062
2063
+ zhao3 2063
2064
+ zhao4 2064
2065
+ zhe5 2065
2066
+ zhe1 2066
2067
+ zhe2 2067
2068
+ zhe3 2068
2069
+ zhe4 2069
2070
+ zhen5 2070
2071
+ zhen1 2071
2072
+ zhen2 2072
2073
+ zhen3 2073
2074
+ zhen4 2074
2075
+ zheng5 2075
2076
+ zheng1 2076
2077
+ zheng2 2077
2078
+ zheng3 2078
2079
+ zheng4 2079
2080
+ zhi5 2080
2081
+ zhi1 2081
2082
+ zhi2 2082
2083
+ zhi3 2083
2084
+ zhi4 2084
2085
+ zhong5 2085
2086
+ zhong1 2086
2087
+ zhong2 2087
2088
+ zhong3 2088
2089
+ zhong4 2089
2090
+ zhou5 2090
2091
+ zhou1 2091
2092
+ zhou2 2092
2093
+ zhou3 2093
2094
+ zhou4 2094
2095
+ zhu5 2095
2096
+ zhu1 2096
2097
+ zhu2 2097
2098
+ zhu3 2098
2099
+ zhu4 2099
2100
+ zhua5 2100
2101
+ zhua1 2101
2102
+ zhua2 2102
2103
+ zhua3 2103
2104
+ zhua4 2104
2105
+ zhuai5 2105
2106
+ zhuai1 2106
2107
+ zhuai2 2107
2108
+ zhuai3 2108
2109
+ zhuai4 2109
2110
+ zhuan5 2110
2111
+ zhuan1 2111
2112
+ zhuan2 2112
2113
+ zhuan3 2113
2114
+ zhuan4 2114
2115
+ zhuang5 2115
2116
+ zhuang1 2116
2117
+ zhuang2 2117
2118
+ zhuang3 2118
2119
+ zhuang4 2119
2120
+ zhui5 2120
2121
+ zhui1 2121
2122
+ zhui2 2122
2123
+ zhui3 2123
2124
+ zhui4 2124
2125
+ zhun5 2125
2126
+ zhun1 2126
2127
+ zhun2 2127
2128
+ zhun3 2128
2129
+ zhun4 2129
2130
+ zhuo5 2130
2131
+ zhuo1 2131
2132
+ zhuo2 2132
2133
+ zhuo3 2133
2134
+ zhuo4 2134
2135
+ zi5 2135
2136
+ zi1 2136
2137
+ zi2 2137
2138
+ zi3 2138
2139
+ zi4 2139
2140
+ zong5 2140
2141
+ zong1 2141
2142
+ zong2 2142
2143
+ zong3 2143
2144
+ zong4 2144
2145
+ zou5 2145
2146
+ zou1 2146
2147
+ zou2 2147
2148
+ zou3 2148
2149
+ zou4 2149
2150
+ zu5 2150
2151
+ zu1 2151
2152
+ zu2 2152
2153
+ zu3 2153
2154
+ zu4 2154
2155
+ zuan5 2155
2156
+ zuan1 2156
2157
+ zuan2 2157
2158
+ zuan3 2158
2159
+ zuan4 2159
2160
+ zui5 2160
2161
+ zui1 2161
2162
+ zui2 2162
2163
+ zui3 2163
2164
+ zui4 2164
2165
+ zun5 2165
2166
+ zun1 2166
2167
+ zun2 2167
2168
+ zun3 2168
2169
+ zun4 2169
2170
+ zuo5 2170
2171
+ zuo1 2171
2172
+ zuo2 2172
2173
+ zuo3 2173
2174
+ zuo4 2174
2175
+ g 2175
2176
+ <|unused_2176|> 2176
2177
+ <|unused_2177|> 2177
2178
+ <|unused_2178|> 2178
2179
+ <|unused_2179|> 2179
2180
+ <|unused_2180|> 2180
2181
+ <|unused_2181|> 2181
2182
+ <|unused_2182|> 2182
2183
+ <|unused_2183|> 2183
2184
+ <|unused_2184|> 2184
2185
+ <|unused_2185|> 2185
2186
+ <|unused_2186|> 2186
2187
+ <|unused_2187|> 2187
2188
+ <|unused_2188|> 2188
2189
+ <|unused_2189|> 2189
2190
+ <|unused_2190|> 2190
app/matcha/vocos-16khz-univ.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e52c68d588dcae47e31236aea56482db5b1c22b09562ae034081c8465d78e22a
3
+ size 27012981
app/nano/.distill4 ADDED
@@ -0,0 +1 @@
 
 
1
+ hoc-turbo1: 4 bước Euler, không CFG
app/nano/.duration ADDED
@@ -0,0 +1 @@
 
 
1
+ hoc-turbo1
app/nano/.int8 ADDED
@@ -0,0 +1 @@
 
 
1
+ vector_estimator + codec_decoder: int8 dynamic (MatMul/Gemm)
app/nano/.tag ADDED
@@ -0,0 +1 @@
 
 
1
+ -ht1
app/nano/codec_decoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:305dff0e8f9721582384eb9d43c65ed6d5655edc07ca625b4cb053b548428297
3
+ size 36453801
app/nano/config.json ADDED
@@ -0,0 +1,172 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dim": 512,
3
+ "n_blocks": 4,
4
+ "t_scale": 1000.0,
5
+ "text_conv": 6,
6
+ "text_attn": 4,
7
+ "text_ffn": true,
8
+ "text_skip": true,
9
+ "ve_conv": 2,
10
+ "text_mult": 4,
11
+ "dur_attn_pool": true,
12
+ "n_style": 50,
13
+ "style_dim": 256,
14
+ "style_conv": 4,
15
+ "rope": true,
16
+ "soft_attn": true,
17
+ "loss": "l1",
18
+ "step": 10000,
19
+ "weights": "ema",
20
+ "latent_scale": 0.25,
21
+ "latent_dim": 24,
22
+ "group": 6,
23
+ "flow_fps": 15.625,
24
+ "sample_rate": 24000,
25
+ "vocab": {
26
+ "<pad>": 0,
27
+ "<bos>": 1,
28
+ "<eos>": 2,
29
+ " ": 3,
30
+ "!": 4,
31
+ ",": 5,
32
+ "-": 6,
33
+ ".": 7,
34
+ "1": 8,
35
+ "2": 9,
36
+ "4": 10,
37
+ "5": 11,
38
+ "6": 12,
39
+ "7": 13,
40
+ "?": 14,
41
+ "a": 15,
42
+ "b": 16,
43
+ "c": 17,
44
+ "d": 18,
45
+ "e": 19,
46
+ "f": 20,
47
+ "h": 21,
48
+ "i": 22,
49
+ "j": 23,
50
+ "k": 24,
51
+ "l": 25,
52
+ "m": 26,
53
+ "n": 27,
54
+ "o": 28,
55
+ "p": 29,
56
+ "s": 30,
57
+ "t": 31,
58
+ "u": 32,
59
+ "v": 33,
60
+ "w": 34,
61
+ "x": 35,
62
+ "y": 36,
63
+ "z": 37,
64
+ "æ": 38,
65
+ "ð": 39,
66
+ "ŋ": 40,
67
+ "ɐ": 41,
68
+ "ɑ": 42,
69
+ "ɔ": 43,
70
+ "ɗ": 44,
71
+ "ə": 45,
72
+ "ɚ": 46,
73
+ "ɛ": 47,
74
+ "ɜ": 48,
75
+ "ɡ": 49,
76
+ "ɣ": 50,
77
+ "ɪ": 51,
78
+ "ɲ": 52,
79
+ "ɹ": 53,
80
+ "ɾ": 54,
81
+ "ʃ": 55,
82
+ "ʊ": 56,
83
+ "ʌ": 57,
84
+ "ʐ": 58,
85
+ "ʒ": 59,
86
+ "ˈ": 60,
87
+ "ˌ": 61,
88
+ "ː": 62,
89
+ "̪": 63,
90
+ "θ": 64,
91
+ "ᵻ": 65,
92
+ "ʔ": 66,
93
+ "̩": 67,
94
+ "̃": 68,
95
+ "ʲ": 69,
96
+ "ɬ": 70,
97
+ "r": 71,
98
+ "①": 72,
99
+ "②": 73,
100
+ "③": 74,
101
+ "④": 75,
102
+ "⑤": 76,
103
+ "⑥": 77,
104
+ "⑦": 78,
105
+ "⑧": 79,
106
+ "⑨": 80
107
+ },
108
+ "latent_mean": [
109
+ 0.13890727611483136,
110
+ -0.13309574985349126,
111
+ -0.11205816217018973,
112
+ -0.11297748382494684,
113
+ -0.18527600033675304,
114
+ -0.08897385313224515,
115
+ -0.12180636634509173,
116
+ -0.15749172029360317,
117
+ -0.2606409891419003,
118
+ -0.10214384919130223,
119
+ -0.010901057135454117,
120
+ -0.011725050934564085,
121
+ 0.5120446419292695,
122
+ 0.06387069270397264,
123
+ -0.055695457979839894,
124
+ 0.1114036366417339,
125
+ -0.23515650815276756,
126
+ -0.007840761999363875,
127
+ -0.002412323989902976,
128
+ 0.03336932937969914,
129
+ -0.020932621271231463,
130
+ -0.25093378486060836,
131
+ 0.019337150806081855,
132
+ 0.0606351782925217
133
+ ],
134
+ "latent_std": [
135
+ 0.5559272161055647,
136
+ 0.620286225887302,
137
+ 0.5138673716259314,
138
+ 0.544040924845873,
139
+ 0.5967750768416471,
140
+ 0.591528478524644,
141
+ 0.5132562207068648,
142
+ 0.6036245085902558,
143
+ 0.7342680708891451,
144
+ 0.552143540496302,
145
+ 0.5918031044717313,
146
+ 0.493982298226479,
147
+ 0.845329374134451,
148
+ 0.6165397848076107,
149
+ 0.5982130330266255,
150
+ 0.5851740130043556,
151
+ 0.7860731788024539,
152
+ 0.6456386975931636,
153
+ 0.5919169089482502,
154
+ 0.5229639498582862,
155
+ 0.5745385523600368,
156
+ 0.5858547764651196,
157
+ 0.5638392611403258,
158
+ 0.5982619661954102
159
+ ],
160
+ "bos_id": 1,
161
+ "eos_id": 2,
162
+ "pad_id": 0,
163
+ "cfg_default": 3.0,
164
+ "steps_default": 16,
165
+ "ref_max_frames": 140,
166
+ "trained_with_cfg_dropout": true,
167
+ "emotion_tags": {
168
+ "<|emotion_1|>": "①",
169
+ "<|emotion_2|>": "②",
170
+ "<|emotion_3|>": "③"
171
+ }
172
+ }
app/nano/constants.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "null_spk": {
3
+ "file": "null_spk.f32",
4
+ "shape": [
5
+ 192
6
+ ]
7
+ },
8
+ "null_style": {
9
+ "file": "null_style.f32",
10
+ "shape": [
11
+ 50,
12
+ 256
13
+ ]
14
+ }
15
+ }
app/nano/null_spk.f32 ADDED
Binary file (768 Bytes). View file
 
app/nano/null_style.f32 ADDED
Binary file (51.2 kB). View file
 
app/nano/text_encoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bcabb989e98731d54f4c17dd07323cf8484c0481c3a733a7652d8532649c2da
3
+ size 13340211
app/nano/voices.json ADDED
The diff for this file is too large to render. See raw diff
 
app/phonon/bundle.json ADDED
@@ -0,0 +1,332 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "built": 1793354297,
3
+ "conditions": {},
4
+ "ctx": 414,
5
+ "dims": {
6
+ "attn": 512,
7
+ "d": 512,
8
+ "ff": 2048,
9
+ "flow_blocks": 6,
10
+ "flow_d": 512,
11
+ "heads": 8,
12
+ "kv_heads": 8,
13
+ "layers": 6,
14
+ "ldim": 32,
15
+ "z": null
16
+ },
17
+ "eos_threshold": 4.0,
18
+ "files": [
19
+ {
20
+ "path": "flow-c414-t1.mlpackage/Data/com.apple.CoreML/model.mlmodel",
21
+ "sha256": "89d76768cc5cb8d84992fbf71e90b7aaaaaa5e2f57d9df7e8f0588a273a21ef1",
22
+ "size": 240018
23
+ },
24
+ {
25
+ "path": "flow-c414-t1.mlpackage/Data/com.apple.CoreML/weights/weight.bin",
26
+ "sha256": "41ea56cf0daaeeb89280d8162a5a206d39f7048f7039c5f6570d4be79edeb032",
27
+ "size": 55118784
28
+ },
29
+ {
30
+ "path": "flow-c414-t1.mlpackage/Manifest.json",
31
+ "sha256": "ab9cff73bec26e2bc6e9d5ea17ae21fe4a67c7999434bcf300eba0638c1ac99c",
32
+ "size": 534
33
+ },
34
+ {
35
+ "path": "flow-c414-t48.mlpackage/Data/com.apple.CoreML/model.mlmodel",
36
+ "sha256": "02cf6cb2c267e13e4a48ca83405b907c209924087853fdf5fc1bdeba4e9f26a2",
37
+ "size": 240018
38
+ },
39
+ {
40
+ "path": "flow-c414-t48.mlpackage/Data/com.apple.CoreML/weights/weight.bin",
41
+ "sha256": "41ea56cf0daaeeb89280d8162a5a206d39f7048f7039c5f6570d4be79edeb032",
42
+ "size": 55118784
43
+ },
44
+ {
45
+ "path": "flow-c414-t48.mlpackage/Manifest.json",
46
+ "sha256": "3460cba582f1fbeca2330324317fe3c2551e6e91aa46262f667c57f27161a551",
47
+ "size": 534
48
+ },
49
+ {
50
+ "path": "host.safetensors",
51
+ "sha256": "0f12ba46901604a50b79bf8c644f5481a327392d5eb8a254377572dc19b65d65",
52
+ "size": 8262120
53
+ },
54
+ {
55
+ "path": "mimi-c266.mlpackage/Data/com.apple.CoreML/model.mlmodel",
56
+ "sha256": "89d8812cb20639ec1a9101030f94950f2c5f044b0dd1df3121739f149d4e9411",
57
+ "size": 131109
58
+ },
59
+ {
60
+ "path": "mimi-c266.mlpackage/Data/com.apple.CoreML/weights/weight.bin",
61
+ "sha256": "60ec322afab27b8404b02ec0a10a988734b4e1495fbdb3b3d687d9dc9f83ffc3",
62
+ "size": 41182464
63
+ },
64
+ {
65
+ "path": "mimi-c266.mlpackage/Manifest.json",
66
+ "sha256": "b546b5b9e8e0bf340ce5822519c43d92296e6fdfe0f264329b00c8978b6adfd8",
67
+ "size": 534
68
+ },
69
+ {
70
+ "path": "tokenizer.json",
71
+ "sha256": "691a00068fa7b989f2f3e5ae5daa2d1a5bb319f2d5360deafcfb19a8bb6e7936",
72
+ "size": 277967
73
+ },
74
+ {
75
+ "path": "voices/Alfie.safetensors",
76
+ "sha256": "afb0998882edce5ec497bb3811a69b8741223679223d67f1a7108609e17ee98f",
77
+ "size": 256080
78
+ },
79
+ {
80
+ "path": "voices/Azgorath.safetensors",
81
+ "sha256": "1c1aa9f83031abeee0d3de1c2521b5d016b5b0d8399a0db47a53ae47827e2ee0",
82
+ "size": 256080
83
+ },
84
+ {
85
+ "path": "voices/Bryony.safetensors",
86
+ "sha256": "76bd6e832276480fcdd75a258c43b3219ad491728f64f22bce5e7a7a446298a4",
87
+ "size": 256080
88
+ },
89
+ {
90
+ "path": "voices/Captain Silas Blackwater.safetensors",
91
+ "sha256": "6738fcd4a55ab742399aec336c143220ba3707bea525dec6111fdc68ffa5dc00",
92
+ "size": 256080
93
+ },
94
+ {
95
+ "path": "voices/Delia.safetensors",
96
+ "sha256": "e33d74cb00eddf1146afc954280feed1a4678b1d47ee90f932c4c8ce53daf346",
97
+ "size": 256080
98
+ },
99
+ {
100
+ "path": "voices/Easton.safetensors",
101
+ "sha256": "0c4f202676fa9d3a145c7f67fe671f119b6843c90987ef501c31d5d7fdae043f",
102
+ "size": 256080
103
+ },
104
+ {
105
+ "path": "voices/Edwin.safetensors",
106
+ "sha256": "096fe506c612b682f5a370ac38e56d2dcd9e4ae7a698d933a45a723d1d10a9e4",
107
+ "size": 256080
108
+ },
109
+ {
110
+ "path": "voices/Elodie-Rose.safetensors",
111
+ "sha256": "75cc46a6abf7b3dc6bb1895bdfb1914a5e26603b5507672a6910df3e05ff73e4",
112
+ "size": 256080
113
+ },
114
+ {
115
+ "path": "voices/Everly.safetensors",
116
+ "sha256": "5e40608feab857153b76bba00973ab303065aae1d005dcf6b8ffafa9ba9f484c",
117
+ "size": 256080
118
+ },
119
+ {
120
+ "path": "voices/Fairy Godmother.safetensors",
121
+ "sha256": "9818c9e58f07e2199ed2984ca9a11252406272e6f4685c2df9dc52bf0495174d",
122
+ "size": 256080
123
+ },
124
+ {
125
+ "path": "voices/Fennick the Young Elf.safetensors",
126
+ "sha256": "0fa2174b8ba6f4dfc470e617aacbf665551a7f622ac308cecf96447ba77d6d64",
127
+ "size": 256080
128
+ },
129
+ {
130
+ "path": "voices/Freya.safetensors",
131
+ "sha256": "ee09c8b9489ad7531c0f4edb9f6e829200c1ccb2aeac630683f348bdb81fe00a",
132
+ "size": 256080
133
+ },
134
+ {
135
+ "path": "voices/Grub the Ogre.safetensors",
136
+ "sha256": "312da125e167ea5ef68d73e6711068960940dd838e1ad15307a2b8f288548ca7",
137
+ "size": 256080
138
+ },
139
+ {
140
+ "path": "voices/Hadley.safetensors",
141
+ "sha256": "aea4ad06ea632200e92c7742143ae63f2bf8dbf2bb63dbb71f3236793b049d9d",
142
+ "size": 256080
143
+ },
144
+ {
145
+ "path": "voices/Harper.safetensors",
146
+ "sha256": "007d6468bcc85bad68590a567d2bc6ab37b14730ba302eba2b0b6152c263ad77",
147
+ "size": 256080
148
+ },
149
+ {
150
+ "path": "voices/Hugh.safetensors",
151
+ "sha256": "ad1a4c6cdc5a1bed165b6aa6e28ce9d1deb1a9e1ab3d4d6b1aa3442292464d2b",
152
+ "size": 256080
153
+ },
154
+ {
155
+ "path": "voices/Imogen.safetensors",
156
+ "sha256": "4f224d84797a24ad0f77a824f5204d68757826b6e55dd902ba48a5b9b2db9368",
157
+ "size": 256080
158
+ },
159
+ {
160
+ "path": "voices/Juniper.safetensors",
161
+ "sha256": "0cde0e23b22e64f39a8f6fb00444c0cb979a6a268078d0a4830c15dfd11bb491",
162
+ "size": 256080
163
+ },
164
+ {
165
+ "path": "voices/Lottie.safetensors",
166
+ "sha256": "9ba860f38b4cfe55884fc616666ff3272d35689ab36f92e759cd5d2ab27c2556",
167
+ "size": 256080
168
+ },
169
+ {
170
+ "path": "voices/Maeve.safetensors",
171
+ "sha256": "5cf22b2ee74b2033180bb3fc14af79c58bbc02f0f0fa0db10f70172915c21599",
172
+ "size": 256080
173
+ },
174
+ {
175
+ "path": "voices/Master Aldric.safetensors",
176
+ "sha256": "c5f5cbca031859f1ae560455e512c9c44cba89968e1933e85efac3bc93efca56",
177
+ "size": 256080
178
+ },
179
+ {
180
+ "path": "voices/Micah.safetensors",
181
+ "sha256": "9347ca6dfb8c1ee86e3fad57f3d92b64fe2d62f2c3ab30318e8b3c73dd730274",
182
+ "size": 256080
183
+ },
184
+ {
185
+ "path": "voices/Miles.safetensors",
186
+ "sha256": "ec7785c2747f7360797cfdfcf440aa06685dc9de7f7c781491ef46647d6be17f",
187
+ "size": 256080
188
+ },
189
+ {
190
+ "path": "voices/Morwenna of the Deep.safetensors",
191
+ "sha256": "d610fc2dc286b73d34eff32cab1759e0019d8c99a9b6823156343fb1a0ecc99e",
192
+ "size": 256080
193
+ },
194
+ {
195
+ "path": "voices/Old Mortlake.safetensors",
196
+ "sha256": "46465cceddfc814bdf8960954ad599b71de7d77bbafa5ddf81d1d51e3db637a0",
197
+ "size": 256080
198
+ },
199
+ {
200
+ "path": "voices/Pippa.safetensors",
201
+ "sha256": "f6ef4d6fd0fbe303dcad7c668aa64d4d10a729d05ffe8bfdd95cd3b87fc22e67",
202
+ "size": 256080
203
+ },
204
+ {
205
+ "path": "voices/Preston.safetensors",
206
+ "sha256": "a29262436888526ed283292cdc86cd74c53bae7b522df495aeec25907cc9a014",
207
+ "size": 256080
208
+ },
209
+ {
210
+ "path": "voices/Russell.safetensors",
211
+ "sha256": "9477c11856e3be342aef7a95e1ff18c3b20722d51372104e8c61df6fe903e7ca",
212
+ "size": 256080
213
+ },
214
+ {
215
+ "path": "voices/Sir Bramwell.safetensors",
216
+ "sha256": "ed0dca2576306b0791df4519de11a130bfef3b6bc497aad77b834cc9e746175e",
217
+ "size": 256080
218
+ },
219
+ {
220
+ "path": "voices/Spencer.safetensors",
221
+ "sha256": "55167fbbc846c01d0f289feff77e9732be558b94e215a5a1c00fcdc6e7b75b71",
222
+ "size": 256080
223
+ },
224
+ {
225
+ "path": "voices/Teddy.safetensors",
226
+ "sha256": "3ea4baa7c1de20f7bc51254c193bb6cedf1ca3342d20bd504ec381a38c38a5d0",
227
+ "size": 256080
228
+ },
229
+ {
230
+ "path": "voices/The Snow Queen.safetensors",
231
+ "sha256": "95267aa0c9e7c932a7fb9129f98393e1872c8f569e23e36c6360c51323066607",
232
+ "size": 256080
233
+ },
234
+ {
235
+ "path": "voices/The Thing Under the Stairs.safetensors",
236
+ "sha256": "0075203662e0f8fe827412ec6f4b79dfe02917275b97fd1fd39560b0dba4d0e9",
237
+ "size": 256080
238
+ },
239
+ {
240
+ "path": "voices/Toby.safetensors",
241
+ "sha256": "e5bf5a2f2b590f5de19894e65d908263799765a9c63aeaf34cf2c25e86938faf",
242
+ "size": 256080
243
+ },
244
+ {
245
+ "path": "voices/Trevor.safetensors",
246
+ "sha256": "910dc105abef1a46d09ff9124d657e19bac58c1a949065aaa54d09112b6c3420",
247
+ "size": 256080
248
+ },
249
+ {
250
+ "path": "voices/Vyrmathrax the Ancient.safetensors",
251
+ "sha256": "4768e23dcfc3ba5b836bc1eebb1bdea051673d7a5241c74dffc21e7e81f3d9e3",
252
+ "size": 256080
253
+ },
254
+ {
255
+ "path": "voices/Zoey.safetensors",
256
+ "sha256": "6b572207f2de3df786b25053605d4470d183f112477b827da149fadb001cc33d",
257
+ "size": 256080
258
+ },
259
+ {
260
+ "path": "voices/default.safetensors",
261
+ "sha256": "86e524b2bf1e80b350f26cde009cb19dd38b402d202ff77cb4df7ceaec6badeb",
262
+ "size": 256080
263
+ }
264
+ ],
265
+ "languages": [
266
+ "en"
267
+ ],
268
+ "max_frames": 225,
269
+ "mimi": {
270
+ "compress": 2,
271
+ "d_model": 512,
272
+ "dim": 512,
273
+ "ff": 2048,
274
+ "heads": 8,
275
+ "kernel_size": 7,
276
+ "last_kernel_size": 3,
277
+ "layers": 2,
278
+ "ldim": 32,
279
+ "n_filters": 64,
280
+ "out_dim": 512,
281
+ "ratios": [
282
+ 6,
283
+ 5,
284
+ 4
285
+ ],
286
+ "residual_kernel_size": 3,
287
+ "steps": 16
288
+ },
289
+ "mimi_window": 250,
290
+ "prefill_len": 48,
291
+ "temperature": 0.30000001192092896,
292
+ "voices": [
293
+ "Alfie",
294
+ "Azgorath",
295
+ "Bryony",
296
+ "Captain Silas Blackwater",
297
+ "Delia",
298
+ "Easton",
299
+ "Edwin",
300
+ "Elodie-Rose",
301
+ "Everly",
302
+ "Fairy Godmother",
303
+ "Fennick the Young Elf",
304
+ "Freya",
305
+ "Grub the Ogre",
306
+ "Hadley",
307
+ "Harper",
308
+ "Hugh",
309
+ "Imogen",
310
+ "Juniper",
311
+ "Lottie",
312
+ "Maeve",
313
+ "Master Aldric",
314
+ "Micah",
315
+ "Miles",
316
+ "Morwenna of the Deep",
317
+ "Old Mortlake",
318
+ "Pippa",
319
+ "Preston",
320
+ "Russell",
321
+ "Sir Bramwell",
322
+ "Spencer",
323
+ "Teddy",
324
+ "The Snow Queen",
325
+ "The Thing Under the Stairs",
326
+ "Toby",
327
+ "Trevor",
328
+ "Vyrmathrax the Ancient",
329
+ "Zoey",
330
+ "default"
331
+ ]
332
+ }
app/phonon/flow-c414-t1.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a416a60e40a943b35286a170a5f2a9d0d67d91fe57758e22042b46143c960e1
3
+ size 243
app/phonon/flow-c414-t1.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:421857a30423326645767af05dacd7452b5912147d8f456a4a6ae998ad57dbc8
3
+ size 355
app/phonon/flow-c414-t1.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
app/phonon/flow-c414-t1.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30571f3319496a5358306ddcc1b544016ad80ee0d316db179f5418f28da8e272
3
+ size 13964352
app/phonon/flow-c414-t48.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cc5fb02677fdc4027d0a4d8e746b16c806f7223f6173a46700a91df28f21f40
3
+ size 243
app/phonon/flow-c414-t48.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3a317ef22dfea23eb9196c416afd922be915de5feed9e1425ac3ba5645e57a5
3
+ size 355
app/phonon/flow-c414-t48.mlmodelc/model.mil ADDED
The diff for this file is too large to render. See raw diff
 
app/phonon/flow-c414-t48.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30571f3319496a5358306ddcc1b544016ad80ee0d316db179f5418f28da8e272
3
+ size 13964352
app/phonon/host.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f12ba46901604a50b79bf8c644f5481a327392d5eb8a254377572dc19b65d65
3
+ size 8262120
app/phonon/mimi-c266.mlmodelc/analytics/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa74a25d53a49beadedec685d4f0af1bcf172393aba07937261c3d57ae727d52
3
+ size 243
app/phonon/mimi-c266.mlmodelc/coremldata.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e823cb3bd90abab3e9f857319c21aeb76467bc4e1097ea471cca2906bb353e5
3
+ size 901
app/phonon/mimi-c266.mlmodelc/model.mil ADDED
@@ -0,0 +1,478 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ program(1.3)
2
+ [buildInfo = dict<string, string>({{"coremlc-component-MIL", "3605.1.1"}, {"coremlc-version", "3600.25.2"}})]
3
+ {
4
+ func main<ios18>(tensor<fp32, [1, 512, 6]> c0_prev, tensor<fp32, [1, 64, 2]> c11_prev, tensor<fp32, [1, 1, 16, 64]> cos, tensor<fp32, [1, 8, 266, 64]> k0, tensor<fp32, [1, 8, 266, 64]> k1, tensor<fp32, [1, 1, 32]> latent, tensor<fp32, [1, 1, 16, 266]> mask, tensor<fp32, [1, 256, 2]> r3_prev, tensor<fp32, [1, 128, 2]> r6_prev, tensor<fp32, [1, 64, 2]> r9_prev, tensor<fp32, [1, 1, 16, 64]> sin, tensor<fp32, [1, 256, 6]> t2_partial, tensor<fp32, [1, 128, 5]> t5_partial, tensor<fp32, [1, 64, 4]> t8_partial, tensor<fp32, [1, 512, 16]> up_partial, tensor<fp32, [1, 8, 266, 64]> v0, tensor<fp32, [1, 8, 266, 64]> v1) {
5
+ tensor<fp32, [32]> c_1 = const()[name = string("c_1"), val = tensor<fp32, [32]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(64)))];
6
+ tensor<fp32, [32]> c_2 = const()[name = string("c_2"), val = tensor<fp32, [32]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(256)))];
7
+ tensor<fp32, [1, 1, 32]> v_3 = mul(x = latent, y = c_2)[name = string("v_3")];
8
+ tensor<fp32, [1, 1, 32]> v_4 = add(x = v_3, y = c_1)[name = string("v_4")];
9
+ tensor<int32, [3]> c_5 = const()[name = string("c_5"), val = tensor<int32, [3]>([0, 2, 1])];
10
+ tensor<fp32, [512, 32, 1]> c_7_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 32, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(448))), scale = tensor<fp32, [512, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(16896))))[name = string("c_7_quantized")];
11
+ tensor<int32, [1]> c_8 = const()[name = string("c_8"), val = tensor<int32, [1]>([1])];
12
+ tensor<int32, [1]> c_9 = const()[name = string("c_9"), val = tensor<int32, [1]>([1])];
13
+ tensor<int32, [2]> c_10 = const()[name = string("c_10"), val = tensor<int32, [2]>([0, 0])];
14
+ string s_11 = const()[name = string("s_11"), val = string("custom")];
15
+ int32 c_12 = const()[name = string("c_12"), val = int32(1)];
16
+ tensor<fp32, [1, 32, 1]> v_6 = transpose(perm = c_5, x = v_4)[name = string("transpose_6")];
17
+ tensor<fp32, [1, 512, 1]> v_13 = conv(dilations = c_9, groups = c_12, pad = c_10, pad_type = s_11, strides = c_8, weight = c_7_quantized, x = v_6)[name = string("v_13")];
18
+ tensor<fp32, [512, 1, 32]> c_14_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 1, 32]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(19008))), scale = tensor<fp32, [1, 1, 1]>([[[0x1.5e01fp-7]]]))[name = string("c_14_quantized")];
19
+ tensor<int32, [1]> c_15 = const()[name = string("c_15"), val = tensor<int32, [1]>([16])];
20
+ tensor<int32, [1]> c_16 = const()[name = string("c_16"), val = tensor<int32, [1]>([1])];
21
+ tensor<int32, [2]> c_17 = const()[name = string("c_17"), val = tensor<int32, [2]>([0, 0])];
22
+ string s_18 = const()[name = string("s_18"), val = string("custom")];
23
+ int32 c_19 = const()[name = string("c_19"), val = int32(512)];
24
+ tensor<int32, [3]> c_20 = const()[name = string("c_20"), val = tensor<int32, [3]>([1, 512, 32])];
25
+ tensor<fp32, [1, 512, 32]> v_21 = conv_transpose(dilations = c_16, groups = c_19, output_shape = c_20, pad = c_17, pad_type = s_18, strides = c_15, weight = c_14_quantized, x = v_13)[name = string("v_21")];
26
+ tensor<int32, [3]> c_22 = const()[name = string("c_22"), val = tensor<int32, [3]>([0, 0, 0])];
27
+ tensor<int32, [3]> c_23 = const()[name = string("c_23"), val = tensor<int32, [3]>([1, 512, 16])];
28
+ tensor<fp32, [1, 512, 16]> v_24 = slice_by_index(begin = c_22, end = c_23, x = v_21)[name = string("v_24")];
29
+ tensor<fp32, [1, 512, 16]> v_25 = add(x = v_24, y = up_partial)[name = string("v_25")];
30
+ tensor<int32, [3]> c_26 = const()[name = string("c_26"), val = tensor<int32, [3]>([0, 0, 16])];
31
+ tensor<int32, [3]> c_27 = const()[name = string("c_27"), val = tensor<int32, [3]>([1, 512, 32])];
32
+ tensor<fp32, [1, 512, 16]> v_28 = slice_by_index(begin = c_26, end = c_27, x = v_21)[name = string("v_28")];
33
+ int32 c_29 = const()[name = string("c_29"), val = int32(2)];
34
+ bool b_30 = const()[name = string("b_30"), val = bool(false)];
35
+ tensor<fp32, [1, 512, 32]> v_31 = concat(axis = c_29, interleave = b_30, values = (v_25, v_28))[name = string("v_31")];
36
+ tensor<int32, [3]> c_32 = const()[name = string("c_32"), val = tensor<int32, [3]>([0, 0, 0])];
37
+ tensor<int32, [3]> c_33 = const()[name = string("c_33"), val = tensor<int32, [3]>([1, 512, 16])];
38
+ tensor<fp32, [1, 512, 16]> v_34 = slice_by_index(begin = c_32, end = c_33, x = v_31)[name = string("v_34")];
39
+ tensor<int32, [3]> c_35 = const()[name = string("c_35"), val = tensor<int32, [3]>([0, 0, 16])];
40
+ tensor<int32, [3]> c_36 = const()[name = string("c_36"), val = tensor<int32, [3]>([1, 512, 32])];
41
+ tensor<fp32, [1, 512, 16]> v_37 = slice_by_index(begin = c_35, end = c_36, x = v_31)[name = string("v_37")];
42
+ tensor<int32, [3]> c_38 = const()[name = string("c_38"), val = tensor<int32, [3]>([0, 2, 1])];
43
+ tensor<fp32, [512]> c_40 = const()[name = string("c_40"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(35456)))];
44
+ tensor<fp32, [512]> c_41 = const()[name = string("c_41"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(37568)))];
45
+ tensor<int32, [1]> c_42 = const()[name = string("c_42"), val = tensor<int32, [1]>([-1])];
46
+ fp32 c_43 = const()[name = string("c_43"), val = fp32(0x1.4f8b58p-17)];
47
+ tensor<fp32, [1, 16, 512]> v_39 = transpose(perm = c_38, x = v_34)[name = string("transpose_5")];
48
+ tensor<fp32, [1, 16, 512]> v_44 = layer_norm(axes = c_42, beta = c_41, epsilon = c_43, gamma = c_40, x = v_39)[name = string("v_44")];
49
+ tensor<fp32, [1536, 512]> c_45_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [1536, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(39680))), scale = tensor<fp32, [1536, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(826176))))[name = string("c_45_quantized")];
50
+ tensor<fp32, [1536]> c_46 = const()[name = string("c_46"), val = tensor<fp32, [1536]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(832384)))];
51
+ tensor<fp32, [1, 16, 1536]> v_47 = linear(bias = c_46, weight = c_45_quantized, x = v_44)[name = string("v_47")];
52
+ tensor<int32, [5]> c_48 = const()[name = string("c_48"), val = tensor<int32, [5]>([1, 16, 3, 8, 64])];
53
+ tensor<fp32, [1, 16, 3, 8, 64]> v_49 = reshape(shape = c_48, x = v_47)[name = string("v_49")];
54
+ tensor<int32, [5]> c_50 = const()[name = string("c_50"), val = tensor<int32, [5]>([2, 0, 3, 1, 4])];
55
+ tensor<int32, [5]> c_52 = const()[name = string("c_52"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
56
+ tensor<int32, [5]> c_53 = const()[name = string("c_53"), val = tensor<int32, [5]>([1, 1, 8, 16, 64])];
57
+ tensor<fp32, [3, 1, 8, 16, 64]> v_51 = transpose(perm = c_50, x = v_49)[name = string("transpose_4")];
58
+ tensor<fp32, [1, 1, 8, 16, 64]> v_54 = slice_by_index(begin = c_52, end = c_53, x = v_51)[name = string("v_54")];
59
+ tensor<int32, [4]> c_55 = const()[name = string("c_55"), val = tensor<int32, [4]>([1, 8, 16, 64])];
60
+ tensor<fp32, [1, 8, 16, 64]> v_56 = reshape(shape = c_55, x = v_54)[name = string("v_56")];
61
+ tensor<int32, [5]> c_57 = const()[name = string("c_57"), val = tensor<int32, [5]>([1, 0, 0, 0, 0])];
62
+ tensor<int32, [5]> c_58 = const()[name = string("c_58"), val = tensor<int32, [5]>([2, 1, 8, 16, 64])];
63
+ tensor<fp32, [1, 1, 8, 16, 64]> v_59 = slice_by_index(begin = c_57, end = c_58, x = v_51)[name = string("v_59")];
64
+ tensor<int32, [4]> c_60 = const()[name = string("c_60"), val = tensor<int32, [4]>([1, 8, 16, 64])];
65
+ tensor<fp32, [1, 8, 16, 64]> v_61 = reshape(shape = c_60, x = v_59)[name = string("v_61")];
66
+ tensor<int32, [5]> c_62 = const()[name = string("c_62"), val = tensor<int32, [5]>([2, 0, 0, 0, 0])];
67
+ tensor<int32, [5]> c_63 = const()[name = string("c_63"), val = tensor<int32, [5]>([3, 1, 8, 16, 64])];
68
+ tensor<fp32, [1, 1, 8, 16, 64]> v_64 = slice_by_index(begin = c_62, end = c_63, x = v_51)[name = string("v_64")];
69
+ tensor<int32, [4]> c_65 = const()[name = string("c_65"), val = tensor<int32, [4]>([1, 8, 16, 64])];
70
+ tensor<fp32, [1, 8, 16, 64]> v_66 = reshape(shape = c_65, x = v_64)[name = string("v_66")];
71
+ tensor<int32, [5]> c_67 = const()[name = string("c_67"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
72
+ tensor<fp32, [1, 8, 16, 32, 2]> v_68 = reshape(shape = c_67, x = v_56)[name = string("v_68")];
73
+ tensor<int32, [5]> c_69 = const()[name = string("c_69"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
74
+ tensor<int32, [5]> c_70 = const()[name = string("c_70"), val = tensor<int32, [5]>([1, 8, 16, 32, 1])];
75
+ tensor<fp32, [1, 8, 16, 32, 1]> v_71 = slice_by_index(begin = c_69, end = c_70, x = v_68)[name = string("v_71")];
76
+ tensor<int32, [5]> c_72 = const()[name = string("c_72"), val = tensor<int32, [5]>([0, 0, 0, 0, 1])];
77
+ tensor<int32, [5]> c_73 = const()[name = string("c_73"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
78
+ tensor<fp32, [1, 8, 16, 32, 1]> v_74 = slice_by_index(begin = c_72, end = c_73, x = v_68)[name = string("v_74")];
79
+ fp32 c_75 = const()[name = string("c_75"), val = fp32(-0x1p+0)];
80
+ tensor<fp32, [1, 8, 16, 32, 1]> v_76 = mul(x = v_74, y = c_75)[name = string("v_76")];
81
+ int32 c_77 = const()[name = string("c_77"), val = int32(4)];
82
+ bool b_78 = const()[name = string("b_78"), val = bool(false)];
83
+ tensor<fp32, [1, 8, 16, 32, 2]> v_79 = concat(axis = c_77, interleave = b_78, values = (v_76, v_71))[name = string("v_79")];
84
+ tensor<int32, [4]> c_80 = const()[name = string("c_80"), val = tensor<int32, [4]>([1, 8, 16, 64])];
85
+ tensor<fp32, [1, 8, 16, 64]> v_81 = reshape(shape = c_80, x = v_79)[name = string("v_81")];
86
+ tensor<fp32, [1, 8, 16, 64]> v_82 = mul(x = v_56, y = cos)[name = string("v_82")];
87
+ tensor<fp32, [1, 8, 16, 64]> v_83 = mul(x = v_81, y = sin)[name = string("v_83")];
88
+ tensor<fp32, [1, 8, 16, 64]> v_84 = add(x = v_82, y = v_83)[name = string("v_84")];
89
+ tensor<int32, [5]> c_85 = const()[name = string("c_85"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
90
+ tensor<fp32, [1, 8, 16, 32, 2]> v_86 = reshape(shape = c_85, x = v_61)[name = string("v_86")];
91
+ tensor<int32, [5]> c_87 = const()[name = string("c_87"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
92
+ tensor<int32, [5]> c_88 = const()[name = string("c_88"), val = tensor<int32, [5]>([1, 8, 16, 32, 1])];
93
+ tensor<fp32, [1, 8, 16, 32, 1]> v_89 = slice_by_index(begin = c_87, end = c_88, x = v_86)[name = string("v_89")];
94
+ tensor<int32, [5]> c_90 = const()[name = string("c_90"), val = tensor<int32, [5]>([0, 0, 0, 0, 1])];
95
+ tensor<int32, [5]> c_91 = const()[name = string("c_91"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
96
+ tensor<fp32, [1, 8, 16, 32, 1]> v_92 = slice_by_index(begin = c_90, end = c_91, x = v_86)[name = string("v_92")];
97
+ fp32 c_93 = const()[name = string("c_93"), val = fp32(-0x1p+0)];
98
+ tensor<fp32, [1, 8, 16, 32, 1]> v_94 = mul(x = v_92, y = c_93)[name = string("v_94")];
99
+ int32 c_95 = const()[name = string("c_95"), val = int32(4)];
100
+ bool b_96 = const()[name = string("b_96"), val = bool(false)];
101
+ tensor<fp32, [1, 8, 16, 32, 2]> v_97 = concat(axis = c_95, interleave = b_96, values = (v_94, v_89))[name = string("v_97")];
102
+ tensor<int32, [4]> c_98 = const()[name = string("c_98"), val = tensor<int32, [4]>([1, 8, 16, 64])];
103
+ tensor<fp32, [1, 8, 16, 64]> v_99 = reshape(shape = c_98, x = v_97)[name = string("v_99")];
104
+ tensor<fp32, [1, 8, 16, 64]> v_100 = mul(x = v_61, y = cos)[name = string("v_100")];
105
+ tensor<fp32, [1, 8, 16, 64]> v_101 = mul(x = v_99, y = sin)[name = string("v_101")];
106
+ tensor<fp32, [1, 8, 16, 64]> v_102 = add(x = v_100, y = v_101)[name = string("v_102")];
107
+ tensor<int32, [4]> c_103 = const()[name = string("c_103"), val = tensor<int32, [4]>([0, 0, 16, 0])];
108
+ tensor<int32, [4]> c_104 = const()[name = string("c_104"), val = tensor<int32, [4]>([1, 8, 266, 64])];
109
+ tensor<fp32, [1, 8, 250, 64]> v_105 = slice_by_index(begin = c_103, end = c_104, x = k0)[name = string("v_105")];
110
+ tensor<int32, [4]> c_106 = const()[name = string("c_106"), val = tensor<int32, [4]>([0, 0, 16, 0])];
111
+ tensor<int32, [4]> c_107 = const()[name = string("c_107"), val = tensor<int32, [4]>([1, 8, 266, 64])];
112
+ tensor<fp32, [1, 8, 250, 64]> v_108 = slice_by_index(begin = c_106, end = c_107, x = v0)[name = string("v_108")];
113
+ int32 c_109 = const()[name = string("c_109"), val = int32(2)];
114
+ bool b_110 = const()[name = string("b_110"), val = bool(false)];
115
+ tensor<fp32, [1, 8, 266, 64]> v_111 = concat(axis = c_109, interleave = b_110, values = (v_105, v_102))[name = string("v_111")];
116
+ int32 c_112 = const()[name = string("c_112"), val = int32(2)];
117
+ bool b_113 = const()[name = string("b_113"), val = bool(false)];
118
+ tensor<fp32, [1, 8, 266, 64]> v_114 = concat(axis = c_112, interleave = b_113, values = (v_108, v_66))[name = string("v_114")];
119
+ bool b_115 = const()[name = string("b_115"), val = bool(false)];
120
+ bool b_116 = const()[name = string("b_116"), val = bool(true)];
121
+ tensor<fp32, [1, 8, 16, 266]> v_117 = matmul(transpose_x = b_115, transpose_y = b_116, x = v_84, y = v_111)[name = string("v_117")];
122
+ fp32 c_118 = const()[name = string("c_118"), val = fp32(0x1p-3)];
123
+ tensor<fp32, [1, 8, 16, 266]> v_119 = mul(x = v_117, y = c_118)[name = string("v_119")];
124
+ tensor<fp32, [1, 8, 16, 266]> v_120 = add(x = v_119, y = mask)[name = string("v_120")];
125
+ int32 c_121 = const()[name = string("c_121"), val = int32(-1)];
126
+ tensor<fp32, [1, 8, 16, 266]> v_122 = softmax(axis = c_121, x = v_120)[name = string("v_122")];
127
+ bool b_123 = const()[name = string("b_123"), val = bool(false)];
128
+ bool b_124 = const()[name = string("b_124"), val = bool(false)];
129
+ tensor<fp32, [1, 8, 16, 64]> v_125 = matmul(transpose_x = b_123, transpose_y = b_124, x = v_122, y = v_114)[name = string("v_125")];
130
+ tensor<int32, [4]> c_126 = const()[name = string("c_126"), val = tensor<int32, [4]>([0, 2, 1, 3])];
131
+ tensor<int32, [3]> c_128 = const()[name = string("c_128"), val = tensor<int32, [3]>([1, 16, 512])];
132
+ tensor<fp32, [1, 16, 8, 64]> v_127 = transpose(perm = c_126, x = v_125)[name = string("transpose_3")];
133
+ tensor<fp32, [1, 16, 512]> v_129 = reshape(shape = c_128, x = v_127)[name = string("v_129")];
134
+ tensor<fp32, [512, 512]> c_130_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(838592))), scale = tensor<fp32, [512, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1100800))))[name = string("c_130_quantized")];
135
+ tensor<fp32, [512]> c_131 = const()[name = string("c_131"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1102912)))];
136
+ tensor<fp32, [1, 16, 512]> v_132 = linear(bias = c_131, weight = c_130_quantized, x = v_129)[name = string("v_132")];
137
+ tensor<fp32, [512]> c_133 = const()[name = string("c_133"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1105024)))];
138
+ tensor<fp32, [1, 16, 512]> v_134 = mul(x = v_132, y = c_133)[name = string("v_134")];
139
+ tensor<fp32, [1, 16, 512]> v_135 = add(x = v_39, y = v_134)[name = string("v_135")];
140
+ tensor<fp32, [512]> c_136 = const()[name = string("c_136"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1107136)))];
141
+ tensor<fp32, [512]> c_137 = const()[name = string("c_137"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1109248)))];
142
+ tensor<int32, [1]> c_138 = const()[name = string("c_138"), val = tensor<int32, [1]>([-1])];
143
+ fp32 c_139 = const()[name = string("c_139"), val = fp32(0x1.4f8b58p-17)];
144
+ tensor<fp32, [1, 16, 512]> v_140 = layer_norm(axes = c_138, beta = c_137, epsilon = c_139, gamma = c_136, x = v_135)[name = string("v_140")];
145
+ tensor<fp32, [2048, 512]> c_141_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [2048, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(1111360))), scale = tensor<fp32, [2048, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(2160000))))[name = string("c_141_quantized")];
146
+ tensor<fp32, [512, 2048]> c_142_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 2048]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(2168256))), scale = tensor<fp32, [512, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3216896))))[name = string("c_142_quantized")];
147
+ tensor<fp32, [2048]> c_143 = const()[name = string("c_143"), val = tensor<fp32, [2048]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3219008)))];
148
+ tensor<fp32, [1, 16, 2048]> v_144 = linear(bias = c_143, weight = c_141_quantized, x = v_140)[name = string("v_144")];
149
+ string s_145 = const()[name = string("s_145"), val = string("EXACT")];
150
+ tensor<fp32, [1, 16, 2048]> v_146 = gelu(mode = s_145, x = v_144)[name = string("v_146")];
151
+ tensor<fp32, [1, 16, 512]> v_148 = linear(bias = c_131, weight = c_142_quantized, x = v_146)[name = string("v_148")];
152
+ tensor<fp32, [512]> c_149 = const()[name = string("c_149"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3227264)))];
153
+ tensor<fp32, [1, 16, 512]> v_150 = mul(x = v_148, y = c_149)[name = string("v_150")];
154
+ tensor<fp32, [1, 16, 512]> v_151 = add(x = v_135, y = v_150)[name = string("v_151")];
155
+ tensor<fp32, [512]> c_152 = const()[name = string("c_152"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3229376)))];
156
+ tensor<fp32, [512]> c_153 = const()[name = string("c_153"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3231488)))];
157
+ tensor<int32, [1]> c_154 = const()[name = string("c_154"), val = tensor<int32, [1]>([-1])];
158
+ fp32 c_155 = const()[name = string("c_155"), val = fp32(0x1.4f8b58p-17)];
159
+ tensor<fp32, [1, 16, 512]> v_156 = layer_norm(axes = c_154, beta = c_153, epsilon = c_155, gamma = c_152, x = v_151)[name = string("v_156")];
160
+ tensor<fp32, [1536, 512]> c_157_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [1536, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(3233600))), scale = tensor<fp32, [1536, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4020096))))[name = string("c_157_quantized")];
161
+ tensor<fp32, [1, 16, 1536]> v_159 = linear(bias = c_46, weight = c_157_quantized, x = v_156)[name = string("v_159")];
162
+ tensor<int32, [5]> c_160 = const()[name = string("c_160"), val = tensor<int32, [5]>([1, 16, 3, 8, 64])];
163
+ tensor<fp32, [1, 16, 3, 8, 64]> v_161 = reshape(shape = c_160, x = v_159)[name = string("v_161")];
164
+ tensor<int32, [5]> c_162 = const()[name = string("c_162"), val = tensor<int32, [5]>([2, 0, 3, 1, 4])];
165
+ tensor<int32, [5]> c_164 = const()[name = string("c_164"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
166
+ tensor<int32, [5]> c_165 = const()[name = string("c_165"), val = tensor<int32, [5]>([1, 1, 8, 16, 64])];
167
+ tensor<fp32, [3, 1, 8, 16, 64]> v_163 = transpose(perm = c_162, x = v_161)[name = string("transpose_2")];
168
+ tensor<fp32, [1, 1, 8, 16, 64]> v_166 = slice_by_index(begin = c_164, end = c_165, x = v_163)[name = string("v_166")];
169
+ tensor<int32, [4]> c_167 = const()[name = string("c_167"), val = tensor<int32, [4]>([1, 8, 16, 64])];
170
+ tensor<fp32, [1, 8, 16, 64]> v_168 = reshape(shape = c_167, x = v_166)[name = string("v_168")];
171
+ tensor<int32, [5]> c_169 = const()[name = string("c_169"), val = tensor<int32, [5]>([1, 0, 0, 0, 0])];
172
+ tensor<int32, [5]> c_170 = const()[name = string("c_170"), val = tensor<int32, [5]>([2, 1, 8, 16, 64])];
173
+ tensor<fp32, [1, 1, 8, 16, 64]> v_171 = slice_by_index(begin = c_169, end = c_170, x = v_163)[name = string("v_171")];
174
+ tensor<int32, [4]> c_172 = const()[name = string("c_172"), val = tensor<int32, [4]>([1, 8, 16, 64])];
175
+ tensor<fp32, [1, 8, 16, 64]> v_173 = reshape(shape = c_172, x = v_171)[name = string("v_173")];
176
+ tensor<int32, [5]> c_174 = const()[name = string("c_174"), val = tensor<int32, [5]>([2, 0, 0, 0, 0])];
177
+ tensor<int32, [5]> c_175 = const()[name = string("c_175"), val = tensor<int32, [5]>([3, 1, 8, 16, 64])];
178
+ tensor<fp32, [1, 1, 8, 16, 64]> v_176 = slice_by_index(begin = c_174, end = c_175, x = v_163)[name = string("v_176")];
179
+ tensor<int32, [4]> c_177 = const()[name = string("c_177"), val = tensor<int32, [4]>([1, 8, 16, 64])];
180
+ tensor<fp32, [1, 8, 16, 64]> v_178 = reshape(shape = c_177, x = v_176)[name = string("v_178")];
181
+ tensor<int32, [5]> c_179 = const()[name = string("c_179"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
182
+ tensor<fp32, [1, 8, 16, 32, 2]> v_180 = reshape(shape = c_179, x = v_168)[name = string("v_180")];
183
+ tensor<int32, [5]> c_181 = const()[name = string("c_181"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
184
+ tensor<int32, [5]> c_182 = const()[name = string("c_182"), val = tensor<int32, [5]>([1, 8, 16, 32, 1])];
185
+ tensor<fp32, [1, 8, 16, 32, 1]> v_183 = slice_by_index(begin = c_181, end = c_182, x = v_180)[name = string("v_183")];
186
+ tensor<int32, [5]> c_184 = const()[name = string("c_184"), val = tensor<int32, [5]>([0, 0, 0, 0, 1])];
187
+ tensor<int32, [5]> c_185 = const()[name = string("c_185"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
188
+ tensor<fp32, [1, 8, 16, 32, 1]> v_186 = slice_by_index(begin = c_184, end = c_185, x = v_180)[name = string("v_186")];
189
+ fp32 c_187 = const()[name = string("c_187"), val = fp32(-0x1p+0)];
190
+ tensor<fp32, [1, 8, 16, 32, 1]> v_188 = mul(x = v_186, y = c_187)[name = string("v_188")];
191
+ int32 c_189 = const()[name = string("c_189"), val = int32(4)];
192
+ bool b_190 = const()[name = string("b_190"), val = bool(false)];
193
+ tensor<fp32, [1, 8, 16, 32, 2]> v_191 = concat(axis = c_189, interleave = b_190, values = (v_188, v_183))[name = string("v_191")];
194
+ tensor<int32, [4]> c_192 = const()[name = string("c_192"), val = tensor<int32, [4]>([1, 8, 16, 64])];
195
+ tensor<fp32, [1, 8, 16, 64]> v_193 = reshape(shape = c_192, x = v_191)[name = string("v_193")];
196
+ tensor<fp32, [1, 8, 16, 64]> v_194 = mul(x = v_168, y = cos)[name = string("v_194")];
197
+ tensor<fp32, [1, 8, 16, 64]> v_195 = mul(x = v_193, y = sin)[name = string("v_195")];
198
+ tensor<fp32, [1, 8, 16, 64]> v_196 = add(x = v_194, y = v_195)[name = string("v_196")];
199
+ tensor<int32, [5]> c_197 = const()[name = string("c_197"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
200
+ tensor<fp32, [1, 8, 16, 32, 2]> v_198 = reshape(shape = c_197, x = v_173)[name = string("v_198")];
201
+ tensor<int32, [5]> c_199 = const()[name = string("c_199"), val = tensor<int32, [5]>([0, 0, 0, 0, 0])];
202
+ tensor<int32, [5]> c_200 = const()[name = string("c_200"), val = tensor<int32, [5]>([1, 8, 16, 32, 1])];
203
+ tensor<fp32, [1, 8, 16, 32, 1]> v_201 = slice_by_index(begin = c_199, end = c_200, x = v_198)[name = string("v_201")];
204
+ tensor<int32, [5]> c_202 = const()[name = string("c_202"), val = tensor<int32, [5]>([0, 0, 0, 0, 1])];
205
+ tensor<int32, [5]> c_203 = const()[name = string("c_203"), val = tensor<int32, [5]>([1, 8, 16, 32, 2])];
206
+ tensor<fp32, [1, 8, 16, 32, 1]> v_204 = slice_by_index(begin = c_202, end = c_203, x = v_198)[name = string("v_204")];
207
+ fp32 c_205 = const()[name = string("c_205"), val = fp32(-0x1p+0)];
208
+ tensor<fp32, [1, 8, 16, 32, 1]> v_206 = mul(x = v_204, y = c_205)[name = string("v_206")];
209
+ int32 c_207 = const()[name = string("c_207"), val = int32(4)];
210
+ bool b_208 = const()[name = string("b_208"), val = bool(false)];
211
+ tensor<fp32, [1, 8, 16, 32, 2]> v_209 = concat(axis = c_207, interleave = b_208, values = (v_206, v_201))[name = string("v_209")];
212
+ tensor<int32, [4]> c_210 = const()[name = string("c_210"), val = tensor<int32, [4]>([1, 8, 16, 64])];
213
+ tensor<fp32, [1, 8, 16, 64]> v_211 = reshape(shape = c_210, x = v_209)[name = string("v_211")];
214
+ tensor<fp32, [1, 8, 16, 64]> v_212 = mul(x = v_173, y = cos)[name = string("v_212")];
215
+ tensor<fp32, [1, 8, 16, 64]> v_213 = mul(x = v_211, y = sin)[name = string("v_213")];
216
+ tensor<fp32, [1, 8, 16, 64]> v_214 = add(x = v_212, y = v_213)[name = string("v_214")];
217
+ tensor<int32, [4]> c_215 = const()[name = string("c_215"), val = tensor<int32, [4]>([0, 0, 16, 0])];
218
+ tensor<int32, [4]> c_216 = const()[name = string("c_216"), val = tensor<int32, [4]>([1, 8, 266, 64])];
219
+ tensor<fp32, [1, 8, 250, 64]> v_217 = slice_by_index(begin = c_215, end = c_216, x = k1)[name = string("v_217")];
220
+ tensor<int32, [4]> c_218 = const()[name = string("c_218"), val = tensor<int32, [4]>([0, 0, 16, 0])];
221
+ tensor<int32, [4]> c_219 = const()[name = string("c_219"), val = tensor<int32, [4]>([1, 8, 266, 64])];
222
+ tensor<fp32, [1, 8, 250, 64]> v_220 = slice_by_index(begin = c_218, end = c_219, x = v1)[name = string("v_220")];
223
+ int32 c_221 = const()[name = string("c_221"), val = int32(2)];
224
+ bool b_222 = const()[name = string("b_222"), val = bool(false)];
225
+ tensor<fp32, [1, 8, 266, 64]> v_223 = concat(axis = c_221, interleave = b_222, values = (v_217, v_214))[name = string("v_223")];
226
+ int32 c_224 = const()[name = string("c_224"), val = int32(2)];
227
+ bool b_225 = const()[name = string("b_225"), val = bool(false)];
228
+ tensor<fp32, [1, 8, 266, 64]> v_226 = concat(axis = c_224, interleave = b_225, values = (v_220, v_178))[name = string("v_226")];
229
+ bool b_227 = const()[name = string("b_227"), val = bool(false)];
230
+ bool b_228 = const()[name = string("b_228"), val = bool(true)];
231
+ tensor<fp32, [1, 8, 16, 266]> v_229 = matmul(transpose_x = b_227, transpose_y = b_228, x = v_196, y = v_223)[name = string("v_229")];
232
+ fp32 c_230 = const()[name = string("c_230"), val = fp32(0x1p-3)];
233
+ tensor<fp32, [1, 8, 16, 266]> v_231 = mul(x = v_229, y = c_230)[name = string("v_231")];
234
+ tensor<fp32, [1, 8, 16, 266]> v_232 = add(x = v_231, y = mask)[name = string("v_232")];
235
+ int32 c_233 = const()[name = string("c_233"), val = int32(-1)];
236
+ tensor<fp32, [1, 8, 16, 266]> v_234 = softmax(axis = c_233, x = v_232)[name = string("v_234")];
237
+ bool b_235 = const()[name = string("b_235"), val = bool(false)];
238
+ bool b_236 = const()[name = string("b_236"), val = bool(false)];
239
+ tensor<fp32, [1, 8, 16, 64]> v_237 = matmul(transpose_x = b_235, transpose_y = b_236, x = v_234, y = v_226)[name = string("v_237")];
240
+ tensor<int32, [4]> c_238 = const()[name = string("c_238"), val = tensor<int32, [4]>([0, 2, 1, 3])];
241
+ tensor<int32, [3]> c_240 = const()[name = string("c_240"), val = tensor<int32, [3]>([1, 16, 512])];
242
+ tensor<fp32, [1, 16, 8, 64]> v_239 = transpose(perm = c_238, x = v_237)[name = string("transpose_1")];
243
+ tensor<fp32, [1, 16, 512]> v_241 = reshape(shape = c_240, x = v_239)[name = string("v_241")];
244
+ tensor<fp32, [512, 512]> c_242_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4026304))), scale = tensor<fp32, [512, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4288512))))[name = string("c_242_quantized")];
245
+ tensor<fp32, [1, 16, 512]> v_244 = linear(bias = c_131, weight = c_242_quantized, x = v_241)[name = string("v_244")];
246
+ tensor<fp32, [512]> c_245 = const()[name = string("c_245"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4290624)))];
247
+ tensor<fp32, [1, 16, 512]> v_246 = mul(x = v_244, y = c_245)[name = string("v_246")];
248
+ tensor<fp32, [1, 16, 512]> v_247 = add(x = v_151, y = v_246)[name = string("v_247")];
249
+ tensor<fp32, [512]> c_248 = const()[name = string("c_248"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4292736)))];
250
+ tensor<fp32, [512]> c_249 = const()[name = string("c_249"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4294848)))];
251
+ tensor<int32, [1]> c_250 = const()[name = string("c_250"), val = tensor<int32, [1]>([-1])];
252
+ fp32 c_251 = const()[name = string("c_251"), val = fp32(0x1.4f8b58p-17)];
253
+ tensor<fp32, [1, 16, 512]> v_252 = layer_norm(axes = c_250, beta = c_249, epsilon = c_251, gamma = c_248, x = v_247)[name = string("v_252")];
254
+ tensor<fp32, [2048, 512]> c_253_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [2048, 512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(4296960))), scale = tensor<fp32, [2048, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(5345600))))[name = string("c_253_quantized")];
255
+ tensor<fp32, [512, 2048]> c_254_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 2048]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(5353856))), scale = tensor<fp32, [512, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(6402496))))[name = string("c_254_quantized")];
256
+ tensor<fp32, [1, 16, 2048]> v_256 = linear(bias = c_143, weight = c_253_quantized, x = v_252)[name = string("v_256")];
257
+ string s_257 = const()[name = string("s_257"), val = string("EXACT")];
258
+ tensor<fp32, [1, 16, 2048]> v_258 = gelu(mode = s_257, x = v_256)[name = string("v_258")];
259
+ tensor<fp32, [1, 16, 512]> v_260 = linear(bias = c_131, weight = c_254_quantized, x = v_258)[name = string("v_260")];
260
+ tensor<fp32, [512]> c_261 = const()[name = string("c_261"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(6404608)))];
261
+ tensor<fp32, [1, 16, 512]> v_262 = mul(x = v_260, y = c_261)[name = string("v_262")];
262
+ tensor<fp32, [1, 16, 512]> v_263 = add(x = v_247, y = v_262)[name = string("v_263")];
263
+ tensor<int32, [3]> c_264 = const()[name = string("c_264"), val = tensor<int32, [3]>([0, 2, 1])];
264
+ tensor<fp32, [512, 512, 7]> c_266_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 512, 7]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(6406720))), scale = tensor<fp32, [512, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(8241792))))[name = string("c_266_quantized")];
265
+ tensor<fp32, [512]> c_267 = const()[name = string("c_267"), val = tensor<fp32, [512]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(8243904)))];
266
+ int32 c_268 = const()[name = string("c_268"), val = int32(2)];
267
+ bool b_269 = const()[name = string("b_269"), val = bool(false)];
268
+ tensor<fp32, [1, 512, 16]> v_265 = transpose(perm = c_264, x = v_263)[name = string("transpose_0")];
269
+ tensor<fp32, [1, 512, 22]> v_270 = concat(axis = c_268, interleave = b_269, values = (c0_prev, v_265))[name = string("v_270")];
270
+ tensor<int32, [1]> c_271 = const()[name = string("c_271"), val = tensor<int32, [1]>([1])];
271
+ tensor<int32, [1]> c_272 = const()[name = string("c_272"), val = tensor<int32, [1]>([1])];
272
+ tensor<int32, [2]> c_273 = const()[name = string("c_273"), val = tensor<int32, [2]>([0, 0])];
273
+ string s_274 = const()[name = string("s_274"), val = string("custom")];
274
+ int32 c_275 = const()[name = string("c_275"), val = int32(1)];
275
+ tensor<fp32, [1, 512, 16]> v_276 = conv(bias = c_267, dilations = c_272, groups = c_275, pad = c_273, pad_type = s_274, strides = c_271, weight = c_266_quantized, x = v_270)[name = string("v_276")];
276
+ tensor<int32, [3]> c_277 = const()[name = string("c_277"), val = tensor<int32, [3]>([0, 0, 16])];
277
+ tensor<int32, [3]> c_278 = const()[name = string("c_278"), val = tensor<int32, [3]>([1, 512, 22])];
278
+ tensor<fp32, [1, 512, 6]> v_279 = slice_by_index(begin = c_277, end = c_278, x = v_270)[name = string("v_279")];
279
+ tensor<fp32, [512, 256, 12]> c_280_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [512, 256, 12]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(8246016))), scale = tensor<fp32, [1, 256, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9818944))))[name = string("c_280_quantized")];
280
+ tensor<fp32, [256]> c_281 = const()[name = string("c_281"), val = tensor<fp32, [256]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9820032)))];
281
+ fp32 c_282 = const()[name = string("c_282"), val = fp32(0x1p+0)];
282
+ tensor<fp32, [1, 512, 16]> v_283 = elu(alpha = c_282, x = v_276)[name = string("v_283")];
283
+ tensor<int32, [1]> c_284 = const()[name = string("c_284"), val = tensor<int32, [1]>([6])];
284
+ tensor<int32, [1]> c_285 = const()[name = string("c_285"), val = tensor<int32, [1]>([1])];
285
+ tensor<int32, [2]> c_286 = const()[name = string("c_286"), val = tensor<int32, [2]>([0, 0])];
286
+ string s_287 = const()[name = string("s_287"), val = string("custom")];
287
+ int32 c_288 = const()[name = string("c_288"), val = int32(1)];
288
+ tensor<int32, [3]> c_289 = const()[name = string("c_289"), val = tensor<int32, [3]>([1, 256, 102])];
289
+ tensor<fp32, [1, 256, 102]> v_290 = conv_transpose(bias = c_281, dilations = c_285, groups = c_288, output_shape = c_289, pad = c_286, pad_type = s_287, strides = c_284, weight = c_280_quantized, x = v_283)[name = string("v_290")];
290
+ tensor<int32, [3]> c_291 = const()[name = string("c_291"), val = tensor<int32, [3]>([0, 0, 0])];
291
+ tensor<int32, [3]> c_292 = const()[name = string("c_292"), val = tensor<int32, [3]>([1, 256, 6])];
292
+ tensor<fp32, [1, 256, 6]> v_293 = slice_by_index(begin = c_291, end = c_292, x = v_290)[name = string("v_293")];
293
+ tensor<fp32, [1, 256, 6]> v_294 = add(x = v_293, y = t2_partial)[name = string("v_294")];
294
+ tensor<int32, [3]> c_295 = const()[name = string("c_295"), val = tensor<int32, [3]>([0, 0, 6])];
295
+ tensor<int32, [3]> c_296 = const()[name = string("c_296"), val = tensor<int32, [3]>([1, 256, 102])];
296
+ tensor<fp32, [1, 256, 96]> v_297 = slice_by_index(begin = c_295, end = c_296, x = v_290)[name = string("v_297")];
297
+ int32 c_298 = const()[name = string("c_298"), val = int32(2)];
298
+ bool b_299 = const()[name = string("b_299"), val = bool(false)];
299
+ tensor<fp32, [1, 256, 102]> v_300 = concat(axis = c_298, interleave = b_299, values = (v_294, v_297))[name = string("v_300")];
300
+ tensor<int32, [3]> c_301 = const()[name = string("c_301"), val = tensor<int32, [3]>([0, 0, 0])];
301
+ tensor<int32, [3]> c_302 = const()[name = string("c_302"), val = tensor<int32, [3]>([1, 256, 96])];
302
+ tensor<fp32, [1, 256, 96]> v_303 = slice_by_index(begin = c_301, end = c_302, x = v_300)[name = string("v_303")];
303
+ tensor<int32, [3]> c_304 = const()[name = string("c_304"), val = tensor<int32, [3]>([0, 0, 96])];
304
+ tensor<int32, [3]> c_305 = const()[name = string("c_305"), val = tensor<int32, [3]>([1, 256, 102])];
305
+ tensor<fp32, [1, 256, 6]> v_306 = slice_by_index(begin = c_304, end = c_305, x = v_300)[name = string("v_306")];
306
+ tensor<fp32, [1, 256, 1]> v_308 = const()[name = string("v_308"), val = tensor<fp32, [1, 256, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9821120)))];
307
+ tensor<fp32, [1, 256, 6]> v_309 = sub(x = v_306, y = v_308)[name = string("v_309")];
308
+ tensor<fp32, [128, 256, 3]> c_310_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [128, 256, 3]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9822208))), scale = tensor<fp32, [128, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9920576))))[name = string("c_310_quantized")];
309
+ tensor<fp32, [128]> c_311 = const()[name = string("c_311"), val = tensor<fp32, [128]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9921152)))];
310
+ tensor<fp32, [256, 128, 1]> c_312_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [256, 128, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9921728))), scale = tensor<fp32, [256, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9954560))))[name = string("c_312_quantized")];
311
+ tensor<fp32, [256]> c_313 = const()[name = string("c_313"), val = tensor<fp32, [256]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9955648)))];
312
+ fp32 c_314 = const()[name = string("c_314"), val = fp32(0x1p+0)];
313
+ tensor<fp32, [1, 256, 96]> v_315 = elu(alpha = c_314, x = v_303)[name = string("v_315")];
314
+ int32 c_316 = const()[name = string("c_316"), val = int32(2)];
315
+ bool b_317 = const()[name = string("b_317"), val = bool(false)];
316
+ tensor<fp32, [1, 256, 98]> v_318 = concat(axis = c_316, interleave = b_317, values = (r3_prev, v_315))[name = string("v_318")];
317
+ tensor<int32, [1]> c_319 = const()[name = string("c_319"), val = tensor<int32, [1]>([1])];
318
+ tensor<int32, [1]> c_320 = const()[name = string("c_320"), val = tensor<int32, [1]>([1])];
319
+ tensor<int32, [2]> c_321 = const()[name = string("c_321"), val = tensor<int32, [2]>([0, 0])];
320
+ string s_322 = const()[name = string("s_322"), val = string("custom")];
321
+ int32 c_323 = const()[name = string("c_323"), val = int32(1)];
322
+ tensor<fp32, [1, 128, 96]> v_324 = conv(bias = c_311, dilations = c_320, groups = c_323, pad = c_321, pad_type = s_322, strides = c_319, weight = c_310_quantized, x = v_318)[name = string("v_324")];
323
+ tensor<int32, [3]> c_325 = const()[name = string("c_325"), val = tensor<int32, [3]>([0, 0, 96])];
324
+ tensor<int32, [3]> c_326 = const()[name = string("c_326"), val = tensor<int32, [3]>([1, 256, 98])];
325
+ tensor<fp32, [1, 256, 2]> v_327 = slice_by_index(begin = c_325, end = c_326, x = v_318)[name = string("v_327")];
326
+ fp32 c_328 = const()[name = string("c_328"), val = fp32(0x1p+0)];
327
+ tensor<fp32, [1, 128, 96]> v_329 = elu(alpha = c_328, x = v_324)[name = string("v_329")];
328
+ tensor<int32, [1]> c_330 = const()[name = string("c_330"), val = tensor<int32, [1]>([1])];
329
+ tensor<int32, [1]> c_331 = const()[name = string("c_331"), val = tensor<int32, [1]>([1])];
330
+ tensor<int32, [2]> c_332 = const()[name = string("c_332"), val = tensor<int32, [2]>([0, 0])];
331
+ string s_333 = const()[name = string("s_333"), val = string("custom")];
332
+ int32 c_334 = const()[name = string("c_334"), val = int32(1)];
333
+ tensor<fp32, [1, 256, 96]> v_335 = conv(bias = c_313, dilations = c_331, groups = c_334, pad = c_332, pad_type = s_333, strides = c_330, weight = c_312_quantized, x = v_329)[name = string("v_335")];
334
+ tensor<fp32, [1, 256, 96]> v_336 = add(x = v_303, y = v_335)[name = string("v_336")];
335
+ tensor<fp32, [256, 128, 10]> c_337_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [256, 128, 10]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(9956736))), scale = tensor<fp32, [1, 128, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10284480))))[name = string("c_337_quantized")];
336
+ tensor<fp32, [128]> c_338 = const()[name = string("c_338"), val = tensor<fp32, [128]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10285056)))];
337
+ fp32 c_339 = const()[name = string("c_339"), val = fp32(0x1p+0)];
338
+ tensor<fp32, [1, 256, 96]> v_340 = elu(alpha = c_339, x = v_336)[name = string("v_340")];
339
+ tensor<int32, [1]> c_341 = const()[name = string("c_341"), val = tensor<int32, [1]>([5])];
340
+ tensor<int32, [1]> c_342 = const()[name = string("c_342"), val = tensor<int32, [1]>([1])];
341
+ tensor<int32, [2]> c_343 = const()[name = string("c_343"), val = tensor<int32, [2]>([0, 0])];
342
+ string s_344 = const()[name = string("s_344"), val = string("custom")];
343
+ int32 c_345 = const()[name = string("c_345"), val = int32(1)];
344
+ tensor<int32, [3]> c_346 = const()[name = string("c_346"), val = tensor<int32, [3]>([1, 128, 485])];
345
+ tensor<fp32, [1, 128, 485]> v_347 = conv_transpose(bias = c_338, dilations = c_342, groups = c_345, output_shape = c_346, pad = c_343, pad_type = s_344, strides = c_341, weight = c_337_quantized, x = v_340)[name = string("v_347")];
346
+ tensor<int32, [3]> c_348 = const()[name = string("c_348"), val = tensor<int32, [3]>([0, 0, 0])];
347
+ tensor<int32, [3]> c_349 = const()[name = string("c_349"), val = tensor<int32, [3]>([1, 128, 5])];
348
+ tensor<fp32, [1, 128, 5]> v_350 = slice_by_index(begin = c_348, end = c_349, x = v_347)[name = string("v_350")];
349
+ tensor<fp32, [1, 128, 5]> v_351 = add(x = v_350, y = t5_partial)[name = string("v_351")];
350
+ tensor<int32, [3]> c_352 = const()[name = string("c_352"), val = tensor<int32, [3]>([0, 0, 5])];
351
+ tensor<int32, [3]> c_353 = const()[name = string("c_353"), val = tensor<int32, [3]>([1, 128, 485])];
352
+ tensor<fp32, [1, 128, 480]> v_354 = slice_by_index(begin = c_352, end = c_353, x = v_347)[name = string("v_354")];
353
+ int32 c_355 = const()[name = string("c_355"), val = int32(2)];
354
+ bool b_356 = const()[name = string("b_356"), val = bool(false)];
355
+ tensor<fp32, [1, 128, 485]> v_357 = concat(axis = c_355, interleave = b_356, values = (v_351, v_354))[name = string("v_357")];
356
+ tensor<int32, [3]> c_358 = const()[name = string("c_358"), val = tensor<int32, [3]>([0, 0, 0])];
357
+ tensor<int32, [3]> c_359 = const()[name = string("c_359"), val = tensor<int32, [3]>([1, 128, 480])];
358
+ tensor<fp32, [1, 128, 480]> v_360 = slice_by_index(begin = c_358, end = c_359, x = v_357)[name = string("v_360")];
359
+ tensor<int32, [3]> c_361 = const()[name = string("c_361"), val = tensor<int32, [3]>([0, 0, 480])];
360
+ tensor<int32, [3]> c_362 = const()[name = string("c_362"), val = tensor<int32, [3]>([1, 128, 485])];
361
+ tensor<fp32, [1, 128, 5]> v_363 = slice_by_index(begin = c_361, end = c_362, x = v_357)[name = string("v_363")];
362
+ tensor<fp32, [1, 128, 1]> v_365 = const()[name = string("v_365"), val = tensor<fp32, [1, 128, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10285632)))];
363
+ tensor<fp32, [1, 128, 5]> v_366 = sub(x = v_363, y = v_365)[name = string("v_366")];
364
+ tensor<fp32, [64, 128, 3]> c_367_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [64, 128, 3]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10286208))), scale = tensor<fp32, [64, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10310848))))[name = string("c_367_quantized")];
365
+ tensor<fp32, [64]> c_368 = const()[name = string("c_368"), val = tensor<fp32, [64]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10311168)))];
366
+ tensor<fp32, [128, 64, 1]> c_369_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [128, 64, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10311488))), scale = tensor<fp32, [128, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10319744))))[name = string("c_369_quantized")];
367
+ tensor<fp32, [128]> c_370 = const()[name = string("c_370"), val = tensor<fp32, [128]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10320320)))];
368
+ fp32 c_371 = const()[name = string("c_371"), val = fp32(0x1p+0)];
369
+ tensor<fp32, [1, 128, 480]> v_372 = elu(alpha = c_371, x = v_360)[name = string("v_372")];
370
+ int32 c_373 = const()[name = string("c_373"), val = int32(2)];
371
+ bool b_374 = const()[name = string("b_374"), val = bool(false)];
372
+ tensor<fp32, [1, 128, 482]> v_375 = concat(axis = c_373, interleave = b_374, values = (r6_prev, v_372))[name = string("v_375")];
373
+ tensor<int32, [1]> c_376 = const()[name = string("c_376"), val = tensor<int32, [1]>([1])];
374
+ tensor<int32, [1]> c_377 = const()[name = string("c_377"), val = tensor<int32, [1]>([1])];
375
+ tensor<int32, [2]> c_378 = const()[name = string("c_378"), val = tensor<int32, [2]>([0, 0])];
376
+ string s_379 = const()[name = string("s_379"), val = string("custom")];
377
+ int32 c_380 = const()[name = string("c_380"), val = int32(1)];
378
+ tensor<fp32, [1, 64, 480]> v_381 = conv(bias = c_368, dilations = c_377, groups = c_380, pad = c_378, pad_type = s_379, strides = c_376, weight = c_367_quantized, x = v_375)[name = string("v_381")];
379
+ tensor<int32, [3]> c_382 = const()[name = string("c_382"), val = tensor<int32, [3]>([0, 0, 480])];
380
+ tensor<int32, [3]> c_383 = const()[name = string("c_383"), val = tensor<int32, [3]>([1, 128, 482])];
381
+ tensor<fp32, [1, 128, 2]> v_384 = slice_by_index(begin = c_382, end = c_383, x = v_375)[name = string("v_384")];
382
+ fp32 c_385 = const()[name = string("c_385"), val = fp32(0x1p+0)];
383
+ tensor<fp32, [1, 64, 480]> v_386 = elu(alpha = c_385, x = v_381)[name = string("v_386")];
384
+ tensor<int32, [1]> c_387 = const()[name = string("c_387"), val = tensor<int32, [1]>([1])];
385
+ tensor<int32, [1]> c_388 = const()[name = string("c_388"), val = tensor<int32, [1]>([1])];
386
+ tensor<int32, [2]> c_389 = const()[name = string("c_389"), val = tensor<int32, [2]>([0, 0])];
387
+ string s_390 = const()[name = string("s_390"), val = string("custom")];
388
+ int32 c_391 = const()[name = string("c_391"), val = int32(1)];
389
+ tensor<fp32, [1, 128, 480]> v_392 = conv(bias = c_370, dilations = c_388, groups = c_391, pad = c_389, pad_type = s_390, strides = c_387, weight = c_369_quantized, x = v_386)[name = string("v_392")];
390
+ tensor<fp32, [1, 128, 480]> v_393 = add(x = v_360, y = v_392)[name = string("v_393")];
391
+ tensor<fp32, [128, 64, 8]> c_394_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [128, 64, 8]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10320896))), scale = tensor<fp32, [1, 64, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10386496))))[name = string("c_394_quantized")];
392
+ tensor<fp32, [64]> c_395 = const()[name = string("c_395"), val = tensor<fp32, [64]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10386816)))];
393
+ fp32 c_396 = const()[name = string("c_396"), val = fp32(0x1p+0)];
394
+ tensor<fp32, [1, 128, 480]> v_397 = elu(alpha = c_396, x = v_393)[name = string("v_397")];
395
+ tensor<int32, [1]> c_398 = const()[name = string("c_398"), val = tensor<int32, [1]>([4])];
396
+ tensor<int32, [1]> c_399 = const()[name = string("c_399"), val = tensor<int32, [1]>([1])];
397
+ tensor<int32, [2]> c_400 = const()[name = string("c_400"), val = tensor<int32, [2]>([0, 0])];
398
+ string s_401 = const()[name = string("s_401"), val = string("custom")];
399
+ int32 c_402 = const()[name = string("c_402"), val = int32(1)];
400
+ tensor<int32, [3]> c_403 = const()[name = string("c_403"), val = tensor<int32, [3]>([1, 64, 1924])];
401
+ tensor<fp32, [1, 64, 1924]> v_404 = conv_transpose(bias = c_395, dilations = c_399, groups = c_402, output_shape = c_403, pad = c_400, pad_type = s_401, strides = c_398, weight = c_394_quantized, x = v_397)[name = string("v_404")];
402
+ tensor<int32, [3]> c_405 = const()[name = string("c_405"), val = tensor<int32, [3]>([0, 0, 0])];
403
+ tensor<int32, [3]> c_406 = const()[name = string("c_406"), val = tensor<int32, [3]>([1, 64, 4])];
404
+ tensor<fp32, [1, 64, 4]> v_407 = slice_by_index(begin = c_405, end = c_406, x = v_404)[name = string("v_407")];
405
+ tensor<fp32, [1, 64, 4]> v_408 = add(x = v_407, y = t8_partial)[name = string("v_408")];
406
+ tensor<int32, [3]> c_409 = const()[name = string("c_409"), val = tensor<int32, [3]>([0, 0, 4])];
407
+ tensor<int32, [3]> c_410 = const()[name = string("c_410"), val = tensor<int32, [3]>([1, 64, 1924])];
408
+ tensor<fp32, [1, 64, 1920]> v_411 = slice_by_index(begin = c_409, end = c_410, x = v_404)[name = string("v_411")];
409
+ int32 c_412 = const()[name = string("c_412"), val = int32(2)];
410
+ bool b_413 = const()[name = string("b_413"), val = bool(false)];
411
+ tensor<fp32, [1, 64, 1924]> v_414 = concat(axis = c_412, interleave = b_413, values = (v_408, v_411))[name = string("v_414")];
412
+ tensor<int32, [3]> c_415 = const()[name = string("c_415"), val = tensor<int32, [3]>([0, 0, 0])];
413
+ tensor<int32, [3]> c_416 = const()[name = string("c_416"), val = tensor<int32, [3]>([1, 64, 1920])];
414
+ tensor<fp32, [1, 64, 1920]> v_417 = slice_by_index(begin = c_415, end = c_416, x = v_414)[name = string("v_417")];
415
+ tensor<int32, [3]> c_418 = const()[name = string("c_418"), val = tensor<int32, [3]>([0, 0, 1920])];
416
+ tensor<int32, [3]> c_419 = const()[name = string("c_419"), val = tensor<int32, [3]>([1, 64, 1924])];
417
+ tensor<fp32, [1, 64, 4]> v_420 = slice_by_index(begin = c_418, end = c_419, x = v_414)[name = string("v_420")];
418
+ tensor<fp32, [1, 64, 1]> v_422 = const()[name = string("v_422"), val = tensor<fp32, [1, 64, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10387136)))];
419
+ tensor<fp32, [1, 64, 4]> v_423 = sub(x = v_420, y = v_422)[name = string("v_423")];
420
+ tensor<fp32, [32, 64, 3]> c_424_quantized = constexpr_blockwise_shift_scale(data = tensor<int8, [32, 64, 3]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10387456))), scale = tensor<fp32, [32, 1, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10393664))))[name = string("c_424_quantized")];
421
+ tensor<fp32, [32]> c_425 = const()[name = string("c_425"), val = tensor<fp32, [32]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10393856)))];
422
+ tensor<fp32, [64, 32, 1]> c_426 = const()[name = string("c_426"), val = tensor<fp32, [64, 32, 1]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10394048)))];
423
+ tensor<fp32, [64]> c_427 = const()[name = string("c_427"), val = tensor<fp32, [64]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10402304)))];
424
+ fp32 c_428 = const()[name = string("c_428"), val = fp32(0x1p+0)];
425
+ tensor<fp32, [1, 64, 1920]> v_429 = elu(alpha = c_428, x = v_417)[name = string("v_429")];
426
+ int32 c_430 = const()[name = string("c_430"), val = int32(2)];
427
+ bool b_431 = const()[name = string("b_431"), val = bool(false)];
428
+ tensor<fp32, [1, 64, 1922]> v_432 = concat(axis = c_430, interleave = b_431, values = (r9_prev, v_429))[name = string("v_432")];
429
+ tensor<int32, [1]> c_433 = const()[name = string("c_433"), val = tensor<int32, [1]>([1])];
430
+ tensor<int32, [1]> c_434 = const()[name = string("c_434"), val = tensor<int32, [1]>([1])];
431
+ tensor<int32, [2]> c_435 = const()[name = string("c_435"), val = tensor<int32, [2]>([0, 0])];
432
+ string s_436 = const()[name = string("s_436"), val = string("custom")];
433
+ int32 c_437 = const()[name = string("c_437"), val = int32(1)];
434
+ tensor<fp32, [1, 32, 1920]> v_438 = conv(bias = c_425, dilations = c_434, groups = c_437, pad = c_435, pad_type = s_436, strides = c_433, weight = c_424_quantized, x = v_432)[name = string("v_438")];
435
+ tensor<int32, [3]> c_439 = const()[name = string("c_439"), val = tensor<int32, [3]>([0, 0, 1920])];
436
+ tensor<int32, [3]> c_440 = const()[name = string("c_440"), val = tensor<int32, [3]>([1, 64, 1922])];
437
+ tensor<fp32, [1, 64, 2]> v_441 = slice_by_index(begin = c_439, end = c_440, x = v_432)[name = string("v_441")];
438
+ fp32 c_442 = const()[name = string("c_442"), val = fp32(0x1p+0)];
439
+ tensor<fp32, [1, 32, 1920]> v_443 = elu(alpha = c_442, x = v_438)[name = string("v_443")];
440
+ tensor<int32, [1]> c_444 = const()[name = string("c_444"), val = tensor<int32, [1]>([1])];
441
+ tensor<int32, [1]> c_445 = const()[name = string("c_445"), val = tensor<int32, [1]>([1])];
442
+ tensor<int32, [2]> c_446 = const()[name = string("c_446"), val = tensor<int32, [2]>([0, 0])];
443
+ string s_447 = const()[name = string("s_447"), val = string("custom")];
444
+ int32 c_448 = const()[name = string("c_448"), val = int32(1)];
445
+ tensor<fp32, [1, 64, 1920]> v_449 = conv(bias = c_427, dilations = c_445, groups = c_448, pad = c_446, pad_type = s_447, strides = c_444, weight = c_426, x = v_443)[name = string("v_449")];
446
+ tensor<fp32, [1, 64, 1920]> v_450 = add(x = v_417, y = v_449)[name = string("v_450")];
447
+ fp32 c_451 = const()[name = string("c_451"), val = fp32(0x1p+0)];
448
+ tensor<fp32, [1, 64, 1920]> v_452 = elu(alpha = c_451, x = v_450)[name = string("v_452")];
449
+ tensor<fp32, [1, 64, 3]> c_453 = const()[name = string("c_453"), val = tensor<fp32, [1, 64, 3]>(BLOBFILE(path = string("@model_path/weights/weight.bin"), offset = uint64(10402624)))];
450
+ tensor<fp32, [1]> c_454 = const()[name = string("c_454"), val = tensor<fp32, [1]>([0x1.0becc4p-12])];
451
+ int32 c_455 = const()[name = string("c_455"), val = int32(2)];
452
+ bool b_456 = const()[name = string("b_456"), val = bool(false)];
453
+ tensor<fp32, [1, 64, 1922]> v_457 = concat(axis = c_455, interleave = b_456, values = (c11_prev, v_452))[name = string("v_457")];
454
+ tensor<int32, [1]> c_458 = const()[name = string("c_458"), val = tensor<int32, [1]>([1])];
455
+ tensor<int32, [1]> c_459 = const()[name = string("c_459"), val = tensor<int32, [1]>([1])];
456
+ tensor<int32, [2]> c_460 = const()[name = string("c_460"), val = tensor<int32, [2]>([0, 0])];
457
+ string s_461 = const()[name = string("s_461"), val = string("custom")];
458
+ int32 c_462 = const()[name = string("c_462"), val = int32(1)];
459
+ tensor<fp32, [1, 1, 1920]> v_463 = conv(bias = c_454, dilations = c_459, groups = c_462, pad = c_460, pad_type = s_461, strides = c_458, weight = c_453, x = v_457)[name = string("v_463")];
460
+ tensor<int32, [3]> c_464 = const()[name = string("c_464"), val = tensor<int32, [3]>([0, 0, 1920])];
461
+ tensor<int32, [3]> c_465 = const()[name = string("c_465"), val = tensor<int32, [3]>([1, 64, 1922])];
462
+ tensor<fp32, [1, 64, 2]> v_466 = slice_by_index(begin = c_464, end = c_465, x = v_457)[name = string("v_466")];
463
+ tensor<fp32, [1, 1, 1920]> pcm = identity(x = v_463)[name = string("pcm")];
464
+ tensor<fp32, [1, 512, 16]> up_partial_out = identity(x = v_37)[name = string("up_partial_out")];
465
+ tensor<fp32, [1, 512, 6]> c0_prev_out = identity(x = v_279)[name = string("c0_prev_out")];
466
+ tensor<fp32, [1, 256, 6]> t2_partial_out = identity(x = v_309)[name = string("t2_partial_out")];
467
+ tensor<fp32, [1, 256, 2]> r3_prev_out = identity(x = v_327)[name = string("r3_prev_out")];
468
+ tensor<fp32, [1, 128, 5]> t5_partial_out = identity(x = v_366)[name = string("t5_partial_out")];
469
+ tensor<fp32, [1, 128, 2]> r6_prev_out = identity(x = v_384)[name = string("r6_prev_out")];
470
+ tensor<fp32, [1, 64, 4]> t8_partial_out = identity(x = v_423)[name = string("t8_partial_out")];
471
+ tensor<fp32, [1, 64, 2]> r9_prev_out = identity(x = v_441)[name = string("r9_prev_out")];
472
+ tensor<fp32, [1, 64, 2]> c11_prev_out = identity(x = v_466)[name = string("c11_prev_out")];
473
+ tensor<fp32, [1, 8, 266, 64]> k0_out = identity(x = v_111)[name = string("k0_out")];
474
+ tensor<fp32, [1, 8, 266, 64]> v0_out = identity(x = v_114)[name = string("v0_out")];
475
+ tensor<fp32, [1, 8, 266, 64]> k1_out = identity(x = v_223)[name = string("k1_out")];
476
+ tensor<fp32, [1, 8, 266, 64]> v1_out = identity(x = v_226)[name = string("v1_out")];
477
+ } -> (pcm, up_partial_out, c0_prev_out, t2_partial_out, r3_prev_out, t5_partial_out, r6_prev_out, t8_partial_out, r9_prev_out, c11_prev_out, k0_out, v0_out, k1_out, v1_out);
478
+ }
app/phonon/mimi-c266.mlmodelc/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72572d8a87531123c5e17bff3a82cc65ab6377348965ccc5fc973fc92b374125
3
+ size 10403456
app/phonon/source.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "model": "hf://gradium/phonon-31m-en@8c64b55ec301728e2d1695bfea87a08a267d0c2d",
3
+ "weights": "model.safetensors",
4
+ "voices": "hf://gradium/phonon-voices@da6c9ec0d47be8df5c87c730a3aa97187f3229c3"
5
+ }
app/phonon/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
app/phonon/voices/Alfie.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afb0998882edce5ec497bb3811a69b8741223679223d67f1a7108609e17ee98f
3
+ size 256080
app/phonon/voices/Azgorath.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c1aa9f83031abeee0d3de1c2521b5d016b5b0d8399a0db47a53ae47827e2ee0
3
+ size 256080
app/phonon/voices/Bryony.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76bd6e832276480fcdd75a258c43b3219ad491728f64f22bce5e7a7a446298a4
3
+ size 256080
app/phonon/voices/Captain Silas Blackwater.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6738fcd4a55ab742399aec336c143220ba3707bea525dec6111fdc68ffa5dc00
3
+ size 256080