shadow-cann commited on
Commit
100694b
·
verified ·
1 Parent(s): d6c1fc8

Add files using upload-large-folder tool

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ 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
+ 1722265270026243_fastspeech2.jpg filter=lfs diff=lfs merge=lfs -text
1722265270026243_fastspeech2.jpg ADDED

Git LFS Details

  • SHA256: ac5f099c378ad73b82939d596beb1e9f1157a3d63c1e9261b6a4d9e364f79031
  • Pointer size: 131 Bytes
  • Size of remote file: 300 kB
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - zh
4
+ tags:
5
+ - hisilicon
6
+ - hispark
7
+ - npu
8
+ - openharmony
9
+ - modelzoo
10
+ - pytorch
11
+ ---
12
+
13
+ # FastSpeech2
14
+
15
+ FastSpeech2 是一种高效的端到端语音合成模型。相比 FastSpeech,FastSpeech2 引入了多尺度时长预测器和能量 / 基频预测分支,优化了时长预测模块并新增韵律特征建模,在合成速度和语音自然度上均有大幅提升。
16
+
17
+ ## Mirror Metadata
18
+
19
+ - Hugging Face repo: shadow-cann/hispark-modelzoo-fastspeech2
20
+ - Portal model id: ie2sc9g1qk00
21
+ - Created at: 2026-01-08 16:18:57
22
+ - Updated at: 2026-01-09 15:08:50
23
+ - Category: 音频
24
+
25
+ ## Framework
26
+
27
+ - PyTorch
28
+
29
+ ## Supported OS
30
+
31
+ - OpenHarmony
32
+ - Linux
33
+
34
+ ## Computing Power
35
+
36
+ - Hi3403V100 SVP_NNN
37
+
38
+ ## Tags
39
+
40
+ - 文本转语音
41
+
42
+ ## Detail Parameters
43
+
44
+ - 输入: 1x40
45
+ - 参数量: 35.266M
46
+ - 计算量: 29.162GFLOPs
47
+
48
+ ## Files In This Repo
49
+
50
+ - fastspeech_hifigan_en.onnx (源模型 / 源模型下载; 源模型 / 源模型元数据; 编译模型 / OM 元数据 / a16w8)
51
+
52
+ ## Upstream Links
53
+
54
+ - Portal card: https://gitbubble.github.io/hisilicon-developer-portal-mirror/model-detail.html?id=ie2sc9g1qk00
55
+ - Upstream repository: https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2/README.md
56
+ - License reference: https://github.com/ming024/FastSpeech2/blob/master/LICENSE
57
+
58
+ ## Notes
59
+
60
+ - This repository was mirrored from the HiSilicon Developer Portal model card and local downloads captured on 2026-03-27.
61
+ - File ownership follows the portal card mapping, not just filename similarity.
62
+ - Cover image: 1722265270026243_fastspeech2.jpg
fastspeech_hifigan_en.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03519020d7e04ba8713780af8da0c4548087f2d3cd19b842efb194ee6e87ca3c
3
+ size 142724651
model-card.json ADDED
@@ -0,0 +1,215 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "FastSpeech2",
3
+ "id": "ie2sc9g1qk00",
4
+ "description": "FastSpeech2 是一种高效的端到端语音合成模型。相比 FastSpeech,FastSpeech2 引入了多尺度时长预测器和能量 / 基频预测分支,优化了时长预测模块并新增韵律特征建模,在合成速度和语音自然度上均有大幅提升。",
5
+ "category": "音频",
6
+ "framework": [
7
+ "PyTorch"
8
+ ],
9
+ "supportOs": [
10
+ "OpenHarmony",
11
+ "Linux"
12
+ ],
13
+ "computingPower": [
14
+ "Hi3403V100 SVP_NNN"
15
+ ],
16
+ "tags": [
17
+ "文本转语音"
18
+ ],
19
+ "repositoryUrl": "https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2/README.md",
20
+ "licenseUrl": "https://github.com/ming024/FastSpeech2/blob/master/LICENSE",
21
+ "downloads": [
22
+ {
23
+ "fileName": "fastspeech_hifigan_en.onnx",
24
+ "variants": [
25
+ "源模型 / 源模型下载",
26
+ "源模型 / 源模型元数据",
27
+ "编译模型 / OM 元数据 / a16w8"
28
+ ]
29
+ }
30
+ ],
31
+ "apiDetail": {
32
+ "createdBy": 139318985286440,
33
+ "creationDate": "2026-01-08 16:18:57",
34
+ "creationUserCN": "sloanqin",
35
+ "lastUpdatedBy": null,
36
+ "lastUpdateDate": "2026-01-09 15:08:50",
37
+ "lastUpdateUserCN": "sloanqin",
38
+ "rowIdx": -1,
39
+ "id": "ie2sc9g1qk00",
40
+ "name": "FastSpeech2",
41
+ "isBeta": 1,
42
+ "betaVersionDesc": "CANN包版本需要适配修改待发布",
43
+ "description": "FastSpeech2 是一种高效的端到端语音合成模型。相比 FastSpeech,FastSpeech2 引入了多尺度时长预测器和能量 / 基频预测分支,优化了时长预测模块并新增韵律特征建模,在合成速度和语音自然度上均有大幅提升。",
44
+ "parentId": "ie2sc9g1qk00",
45
+ "coverImageId": 1722265270026243,
46
+ "coverImageUrl": "https://openxinhuo-board-image.obs.cn-east-3.myhuaweicloud.com/1722265270026243%2Ffastspeech2.jpg",
47
+ "modelEffectId": null,
48
+ "modelEffectUrl": "",
49
+ "computerVersion": [],
50
+ "naturalLanguageProcess": [],
51
+ "multimodal": [],
52
+ "video": [
53
+ "文本转语音"
54
+ ],
55
+ "framework": [
56
+ "PyTorch"
57
+ ],
58
+ "modelRepository": "https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2/README.md",
59
+ "originModel": [
60
+ {
61
+ "id": "1722129101946883",
62
+ "name": "fastspeech_hifigan_en.onnx",
63
+ "url": null,
64
+ "size": "142724651"
65
+ }
66
+ ],
67
+ "originModelLink": null,
68
+ "dataSet": "https://data.keithito.com/data/speech/LJSpeech-1.1.tar.bz2",
69
+ "modelLicense": "https://github.com/ming024/FastSpeech2/blob/master/LICENSE",
70
+ "detailParams": [
71
+ {
72
+ "name": "输入",
73
+ "value": "1x40"
74
+ },
75
+ {
76
+ "name": "参数量",
77
+ "value": "35.266M"
78
+ },
79
+ {
80
+ "name": "计算量",
81
+ "value": "29.162GFLOPs"
82
+ }
83
+ ],
84
+ "quickStart": {
85
+ "url": "https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2",
86
+ "markDownUrl": "",
87
+ "developLanguage": [
88
+ {
89
+ "language": "C++",
90
+ "context": "{\"ops\":[{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#40485b\",\"size\":\"16px\",\"background\":\"#ffffff\"},\"insert\":\"模型可以通过以下代码完成快速推理\"},{\"attributes\":{\"text-indent\":\"0px\"},\"insert\":\"\\n\"},{\"insert\":\"#include \\\"model.h\\\"\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"#include \\\"log.h\\\"\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\\n\"},{\"insert\":\"using namespace Infer;\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\\n\"},{\"insert\":\"int main()\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"{\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    EnvInit();\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    std::string omModelPath = \\\"/path/to/model.om\\\"; // 模型文件路径 \"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    std::string filePath= \\\"/path/to/file_list.json\\\"; // 输入文本文件路径\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    std::unique_ptr<Model> model = std::make_unique<Model>();\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    if (model->Load(omModelPath, ModelType::FastSpeech2) != 0) {\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        LOG(ERROR) << \\\"fail to load model\\\";\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        return -1;\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    }\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    auto ret = model->Infer(filePath, FileType::JsonFile);\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    if (ret.size() == 0) {\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        LOG(ERROR) << \\\"fail to infer model\\\";\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        model->Unload();\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        return -1;\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    }\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    if (model->Unload() != 0) {\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        LOG(ERROR) << \\\"fail to unload model\\\";\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"        return -1;\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    }\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    EnvDeinit();\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"    return 0;\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\"}\"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#40485b\",\"size\":\"16px\",\"background\":\"#ffffff\"},\"insert\":\"备注:上述C++代码依赖的动态库与头文件位于\"},{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#095eab\",\"size\":\"16px\",\"background\":\"transparent\",\"link\":\"https://gitee.com/HiSpark/modelzoo/tree/master/samples/common\"},\"insert\":\"/samples/common\"},{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#40485b\",\"size\":\"16px\",\"background\":\"#ffffff\"},\"insert\":\"目录下,编译相关配置参考\"},{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#095eab\",\"size\":\"16px\",\"background\":\"transparent\",\"link\":\"https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2/src/CMakeLists.txt\"},\"insert\":\"CMakeLists.txt\"},{\"attributes\":{\"line-height\":\"1.6\",\"color\":\"#40485b\",\"size\":\"16px\",\"background\":\"#ffffff\"},\"insert\":\"。\"},{\"attributes\":{\"text-indent\":\"0px\"},\"insert\":\"\\n\"},{\"insert\":\"\\n\"}]}"
91
+ }
92
+ ]
93
+ },
94
+ "status": "released",
95
+ "currentHandler": "",
96
+ "currentHandlerName": "",
97
+ "jsonPath": "https://gitee.com/sloanqin/modelzoo-dev/blob/fastspeech2/samples/built-in/audio/FastSpeech2/fastspeech2.json",
98
+ "modelAdaptor": [
99
+ {
100
+ "createdBy": null,
101
+ "creationDate": null,
102
+ "creationUserCN": null,
103
+ "lastUpdatedBy": null,
104
+ "lastUpdateDate": null,
105
+ "lastUpdateUserCN": null,
106
+ "rowIdx": -1,
107
+ "id": "i8ttm5k1tc00",
108
+ "name": "Hi3403V100 SVP_NNN",
109
+ "modelId": "ie2sc9g1qk00",
110
+ "modelName": "FastSpeech2",
111
+ "supportNames": [
112
+ "a16w8"
113
+ ],
114
+ "toolkit": [
115
+ {
116
+ "name": "CANN配置",
117
+ "url": "",
118
+ "desc": "6.10.t01spc030b660;(请联系FAE获取)",
119
+ "imgId": "cann"
120
+ },
121
+ {
122
+ "name": "编译工具库",
123
+ "url": "https://gitee.com/HiSpark/pegasus/blob/Beta-v0.9.1/docs/Hi3403V100%E7%8E%AF%E5%A2%83%E6%90%AD%E5%BB%BA%E6%8C%87%E5%8D%97/Hi3403V100%E7%8E%AF%E5%A2%83%E6%90%AD%E5%BB%BA%E6%8C%87%E5%8D%97.md",
124
+ "desc": "高效编译,精准适配;AI性能优化,应用流畅运行",
125
+ "imgId": "tool"
126
+ },
127
+ {
128
+ "name": "SDK",
129
+ "url": "https://gitee.com/HiSpark/ss928v100_clang/tree/Beta-v0.9.1/",
130
+ "desc": "稳定、易用的设计;支撑客户快速产品量产",
131
+ "imgId": "sdk"
132
+ }
133
+ ],
134
+ "supportOs": [
135
+ "OpenHarmony",
136
+ "Linux"
137
+ ],
138
+ "supportQuantify": [
139
+ {
140
+ "createdBy": 132241120926760,
141
+ "creationDate": "2026-03-26 19:50:46",
142
+ "creationUserCN": "liaoshibin",
143
+ "lastUpdatedBy": 132241120926760,
144
+ "lastUpdateDate": "2026-03-26 19:50:46",
145
+ "lastUpdateUserCN": "liaoshibin",
146
+ "rowIdx": -1,
147
+ "id": "j6tf932ci000",
148
+ "name": "a16w8",
149
+ "computingId": "i8ttm5k1tc00",
150
+ "computingName": "Hi3403V100 SVP_NNN",
151
+ "omOfflineModelUrl": null,
152
+ "omOfflineModelId": 1722129460559875,
153
+ "omOfflineModelSize": "40041177",
154
+ "omOfflineModelName": "fastspeech_hifigan_en.om",
155
+ "omOfflineModel": [
156
+ {
157
+ "id": "1722129460559875",
158
+ "name": "fastspeech_hifigan_en.om",
159
+ "url": null,
160
+ "size": "40041177"
161
+ }
162
+ ],
163
+ "omOfflineModelLink": null,
164
+ "releaseTime": "2026-01-11",
165
+ "boardOs": null,
166
+ "modelLicense": "https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/audio/FastSpeech2/LICENSE",
167
+ "modelPerformance": [
168
+ {
169
+ "performanceValue": "52.49",
170
+ "unit": "耗时(ms)",
171
+ "desc": null
172
+ },
173
+ {
174
+ "performanceValue": "19.05",
175
+ "unit": "性能(fps)",
176
+ "desc": ""
177
+ },
178
+ {
179
+ "performanceValue": "488.799",
180
+ "unit": "单帧内存带宽(MB)",
181
+ "desc": ""
182
+ },
183
+ {
184
+ "performanceValue": "1.418",
185
+ "unit": "内存(MB)",
186
+ "desc": ""
187
+ }
188
+ ],
189
+ "deleted": 0
190
+ }
191
+ ],
192
+ "deleted": 0
193
+ }
194
+ ],
195
+ "saveType": null,
196
+ "deleteType": null,
197
+ "latest": "Y",
198
+ "deleted": 0,
199
+ "modelPhase": "released",
200
+ "remark": null,
201
+ "fileInfo": null,
202
+ "reviewType": null,
203
+ "owner": "sloanqin",
204
+ "ownerBy": 139318985286440,
205
+ "optional": null,
206
+ "optionalList": [
207
+ "whalenowings",
208
+ "0e900d99dee8461b8"
209
+ ],
210
+ "optionalBy": null,
211
+ "downloadNum": 23,
212
+ "collectNum": null,
213
+ "isCollect": null
214
+ }
215
+ }