File size: 15,481 Bytes
3209708 3134de2 3209708 abc2516 3209708 abc2516 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 3134de2 3209708 abc2516 3209708 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 | {
"name": "Swin-Transformer",
"id": "ko3ugg54j400",
"description": "Swin-Transformer是针对于图片处理设计的基于Transformer架构的神经网络。该网络针对原始Transformer迁移到图片端后计算量过大，复用困难的问题，提出了新的swin-block以代替原有的attention架构。模型以窗口的attention方式极大地减少了图像不同区域间的互相响应，同时也避免了部分冗余信息的产生。最终，模型在减少了大量计算量的同时，在不同的视觉传统任务上也有了效果的提升。",
"category": "计算机视觉",
"framework": [
"PyTorch"
],
"supportOs": [
"OpenHarmony",
"Linux"
],
"computingPower": [
"Hi3403V100 SVP_NNN",
"Hi3403V100 NNN"
],
"tags": [
"分类"
],
"repositoryUrl": "https://gitee.com/HiSpark/modelzoo/tree/master/samples/built-in/classification/Swin-Transformer",
"licenseUrl": "https://github.com/microsoft/Swin-Transformer/blob/main/LICENSE",
"downloads": [
{
"fileName": "swin.onnx",
"variants": [
"源模型 / 源模型元数据"
]
},
{
"fileName": "swin_tiny_patch4_window7_224.pth",
"variants": [
"源模型 / 源模型元数据"
]
},
{
"fileName": "swin_sim.onnx",
"variants": [
"源模型 / 源模型元数据"
]
},
{
"fileName": "swin.om",
"variants": [
"编译模型 / OM 元数据 / A8W8"
]
},
{
"fileName": "swin_dlite.om",
"variants": [
"编译模型 / OM 元数据 / FP16"
]
},
{
"fileName": "SVP_NNN_PC_V1.0.6.0.tgz",
"variants": [
"附加资源 / 附加资源"
]
},
{
"fileName": "swin_om-A8W8.om",
"variants": [
"编译模型 / 镜像补充 / A8W8"
]
}
],
"apiDetail": {
"createdBy": 139036723332644,
"creationDate": "2026-08-26 16:38:28",
"creationUserCN": "y-0089",
"lastUpdatedBy": null,
"lastUpdateDate": "2026-09-01 16:13:06",
"lastUpdateUserCN": "y-0089",
"rowIdx": -1,
"id": "ko3ugg54j400",
"name": "Swin-Transformer",
"isBeta": 0,
"betaVersionDesc": "",
"description": "Swin-Transformer是针对于图片处理设计的基于Transformer架构的神经网络。该网络针对原始Transformer迁移到图片端后计算量过大，复用困难的问题，提出了新的swin-block以代替原有的attention架构。模型以窗口的attention方式极大地减少了图像不同区域间的互相响应，同时也避免了部分冗余信息的产生。最终，模型在减少了大量计算量的同时，在不同的视觉传统任务上也有了效果的提升。",
"parentId": "h6s4m1ogto00",
"coverImageId": 1701485580845059,
"coverImageUrl": "https://openxinhuo-board-image.obs.cn-east-3.myhuaweicloud.com/1701485580845059%2Fswint.png",
"modelEffectId": null,
"modelEffectUrl": "",
"computerVersion": [
"分类"
],
"naturalLanguageProcess": [],
"multimodal": [],
"video": [],
"framework": [
"PyTorch"
],
"modelRepository": "https://gitee.com/HiSpark/modelzoo/tree/master/samples/built-in/classification/Swin-Transformer",
"originModel": [
{
"id": "1706935124623362",
"name": "swin.onnx",
"url": null,
"size": "115360652"
},
{
"id": "1706935036477443",
"name": "swin_tiny_patch4_window7_224.pth",
"url": null,
"size": "114342173"
},
{
"id": "1706935070031874",
"name": "swin_sim.onnx",
"url": null,
"size": "115268484"
}
],
"originModelLink": "",
"dataSet": "https://image-net.org/",
"modelLicense": "https://github.com/microsoft/Swin-Transformer/blob/main/LICENSE",
"detailParams": [
{
"name": "输入",
"value": "224x224"
},
{
"name": "参数量",
"value": "28.798M"
},
{
"name": "计算量",
"value": "9.567GFLOPs"
}
],
"quickStart": {
"url": "https://gitee.com/HiSpark/modelzoo/tree/master/samples/built-in/classification/Swin-Transformer",
"markDownUrl": "",
"developLanguage": [
{
"language": "C++",
"context": "{\"ops\":[{\"attributes\":{\"size\":\"16px\",\"color\":\"#333333\",\"line-height\":\"inherit\"},\"insert\":\"Swin-Transformer可以通过以下代码完成快速推理, 该代码仅展示主要流程，完整实现参考src/main.cpp文件和readme\"},{\"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\\\"; // Swin-Transformer 模型文件路径 \"},{\"attributes\":{\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"insert\":\" std::string imagePath = \\\"/path/to/image.jpg\\\"; // 输入图片路径\"},{\"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::SwinT) != 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(imagePath, FileType::SingelImageFile);\"},{\"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\":{\"text-indent\":\"0px\",\"code-block\":\"plain\"},\"insert\":\"\\n\"},{\"attributes\":{\"size\":\"16px\",\"color\":\"#333333\",\"line-height\":\"inherit\"},\"insert\":\"备注：头文件和动态库位于\"},{\"attributes\":{\"size\":\"16px\",\"color\":\"#4183c4\",\"line-height\":\"inherit\",\"link\":\"https://gitee.com/HiSpark/modelzoo/tree/master/samples/common\"},\"insert\":\"/samples/common\"},{\"attributes\":{\"size\":\"16px\",\"color\":\"#333333\",\"line-height\":\"inherit\"},\"insert\":\"目录下，编译配置参考\"},{\"attributes\":{\"size\":\"16px\",\"color\":\"#4183c4\",\"line-height\":\"inherit\",\"link\":\"https://gitee.com/HiSpark/modelzoo/blob/master/samples/built-in/classification/Swin-Transformer/src/CMakeLists.txt\"},\"insert\":\"文件\"},{\"attributes\":{\"text-indent\":\"0px\"},\"insert\":\"\\n\"}]}"
}
]
},
"status": "released",
"currentHandler": "",
"currentHandlerName": "",
"jsonPath": "https://gitee.com/HiSpark/modelzoo-dev/blob/master/samples/built-in/classification/Swin-Transformer/Swin-Transformer.json",
"modelAdaptor": [
{
"createdBy": null,
"creationDate": null,
"creationUserCN": null,
"lastUpdatedBy": null,
"lastUpdateDate": null,
"lastUpdateUserCN": null,
"rowIdx": -1,
"id": "i8ttm5k1tc00",
"name": "Hi3403V100 SVP_NNN",
"modelId": "ko3ugg54j400",
"modelName": "Swin-Transformer",
"supportNames": [
"A8W8"
],
"toolkit": [
{
"name": "CANN工具",
"url": "https://hispark-obs.obs.cn-east-3.myhuaweicloud.com/SVP_NNN_PC_V1.0.6.0.tgz",
"desc": "Al异构计算架构；提升计算效率的关键平台",
"imgId": "cann"
},
{
"name": "编译工具链",
"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#241%E5%AE%89%E8%A3%85clang%E4%BA%A4%E5%8F%89%E7%BC%96%E8%AF%91%E5%99%A8",
"desc": "高效编译，精准适配；AI性能优化，应用流畅运行",
"imgId": "tool"
},
{
"name": "SDK",
"url": "https://gitee.com/HiSpark/ss928v100_clang/tree/Beta-v0.9.1/",
"desc": "稳定、易用的设计；支撑客户快速产品量产",
"imgId": "sdk"
}
],
"supportOs": [
"OpenHarmony",
"Linux"
],
"supportQuantify": [
{
"createdBy": 139036723332644,
"creationDate": "2026-08-26 16:38:28",
"creationUserCN": "y-0089",
"lastUpdatedBy": 139036723332644,
"lastUpdateDate": "2026-08-26 16:38:28",
"lastUpdateUserCN": "y-0089",
"rowIdx": -1,
"id": "ko3uggecj400",
"name": "A8W8",
"computingId": "i8ttm5k1tc00",
"computingName": "Hi3403V100 SVP_NNN",
"omOfflineModelUrl": null,
"omOfflineModelId": 1712880949002241,
"omOfflineModelSize": "30888496",
"omOfflineModelName": "swin.om",
"omOfflineModel": [
{
"id": "1734292545798147",
"name": "swin.om",
"url": null,
"size": "30870482"
}
],
"omOfflineModelLink": null,
"releaseTime": "2025-11-18",
"boardOs": "OpenHarmony",
"modelLicense": "https://gitee.com/HiSpark/modelzoo/tree/master/samples/built-in/classification/Swin-Transformer/LICENSE",
"modelPerformance": [
{
"performanceValue": "25.675",
"unit": "耗时（ms）",
"desc": ""
},
{
"performanceValue": "38.949",
"unit": "性能（fps）",
"desc": ""
},
{
"performanceValue": "295.711",
"unit": "单帧内存带宽（MB）",
"desc": ""
},
{
"performanceValue": "38.176",
"unit": "内存（MB）",
"desc": ""
}
],
"deleted": 0
}
],
"deleted": 0
},
{
"createdBy": null,
"creationDate": null,
"creationUserCN": null,
"lastUpdatedBy": null,
"lastUpdateDate": null,
"lastUpdateUserCN": null,
"rowIdx": -1,
"id": "i8ttm5k1tc01",
"name": "Hi3403V100 NNN",
"modelId": "ko3ugg54j400",
"modelName": "Swin-Transformer",
"supportNames": [
"A8W8",
"FP16"
],
"toolkit": [
{
"name": "CANN工具",
"url": "",
"desc": "5.30.t11.7.b110； (请联系FAE获取)",
"imgId": "cann"
},
{
"name": "编译工具链",
"url": "",
"desc": "aarch64-mix210-linux-gcc；(请联系FAE获取)",
"imgId": "tool"
},
{
"name": "SDK",
"url": "",
"desc": "SS928 V100R001C02SPC022； (请联系FAE获取)",
"imgId": "sdk"
}
],
"supportOs": [
"Linux"
],
"supportQuantify": [
{
"createdBy": 139036723332644,
"creationDate": "2026-08-26 16:38:28",
"creationUserCN": "y-0089",
"lastUpdatedBy": 139036723332644,
"lastUpdateDate": "2026-08-26 16:38:28",
"lastUpdateUserCN": "y-0089",
"rowIdx": -1,
"id": "ko3uggfsj400",
"name": "FP16",
"computingId": "i8ttm5k1tc01",
"computingName": "Hi3403V100 NNN",
"omOfflineModelUrl": null,
"omOfflineModelId": 1712880995074050,
"omOfflineModelSize": "62380441",
"omOfflineModelName": "swin_dlite.om",
"omOfflineModel": [
{
"id": "1712880995074050",
"name": "swin_dlite.om",
"url": null,
"size": "62380441"
}
],
"omOfflineModelLink": null,
"releaseTime": "2025-11-18",
"boardOs": "Linux",
"modelLicense": "https://gitee.com/HiSpark/modelzoo/tree/master/samples/built-in/classification/Swin-Transformer/LICENSE",
"modelPerformance": [
{
"performanceValue": "101.97",
"unit": "耗时（ms）",
"desc": "性能工具。"
},
{
"performanceValue": "9.81",
"unit": "性能（fps）",
"desc": "性能工具。"
},
{
"performanceValue": "494.766",
"unit": "单帧内存带宽（MB）",
"desc": ""
},
{
"performanceValue": "215.668",
"unit": "内存（MB）",
"desc": ""
}
],
"deleted": 0
}
],
"deleted": 0
}
],
"saveType": null,
"deleteType": null,
"latest": "Y",
"deleted": 0,
"modelPhase": "released",
"remark": null,
"fileInfo": null,
"reviewType": null,
"owner": "y-0089",
"ownerBy": 139036723332644,
"optional": null,
"optionalList": [
"0e900d99dee8461b8"
],
"optionalBy": null,
"downloadNum": 53,
"collectNum": null,
"isCollect": null
}
}
|