下载MindSpore Lite
欢迎使用MindSpore Lite,我们提供了支持多种操作系统和硬件平台的模型转换、模型推理、图像处理等功能,你可以下载适用于本地环境的版本包直接使用。
其中Linux-x86_64、Linux-aarch64的包已在Linux发行版Euleros2.0、Centos7.8、Ubuntu18.04版本上经过测试验证。
1.9.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理和训练runtime库、推理和训练aar包、Micro库、benchmark工具 |
CPU |
Android-aarch32 |
1662181a50852bd4759a821ca29872897efb835c4c318c738609ac4218b80905 |
|
推理和训练runtime库、推理和训练aar包、Micro库、benchmark工具 |
CPU/GPU |
Android-aarch64 |
72a4d51dad9e17b46ae90986ff52cf33393cb549186bf9615dcb32a9d5d084a6 |
|
推理和训练runtime库、推理和训练jar包、Micro库、benchmark工具、converter工具、cropper工具 |
CPU |
Linux-x86_64 |
6175127898766e8a71b4968c8d7588ebe5b192e2c51925814d2b5b3e69c64a7f |
|
推理runtime库、Micro库、benchmark工具、converter工具 |
CPU |
Linux-aarch64 |
00e87b88f095b24467cdb3ca752f4e7300b5ee58e28b33a584d9ff81f0647d7b |
|
推理runtime库、Micro库、benchmark工具、converter工具 |
CPU |
Windows-x86_64 |
e9a5ac1146860bd576790bb46242a2cb5f8711eb25267cf726a7bf7c510216a2 |
|
iOS 推理runtime库 |
CPU |
iOS-aarch32 |
655d9e71e8812e8afd6ff19ab523b3a942f526687bb9ff5165506ba32a1c510e |
|
iOS 推理runtime库 |
CPU |
iOS-aarch64 |
bfb56c4295d006d96b20b52636361519dde8ad2141abb54f5fa6fc3b61311401 |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
5fccaa2c6f267f6edf52f8cb699697adc8253fbc3855248029a67d5b33cf8766 |
|
NNIE 推理runtime库、benchmark工具 |
Hi3516D |
Linux-aarch32 |
926c96e9a9f77952689eb0ef757ea79600f5c5fcfcef2c26e6178751efbd7a4d |
|
NNIE 推理runtime库、benchmark工具 |
Hi3519A |
Linux-aarch32 |
5a98ffeaea6a0ec712b60cc8017a6623d0d8fa75ddb230c9a9e855fec4390456 |
|
NNIE 推理runtime库、benchmark工具 |
Hi3559A |
Linux-aarch64 |
be897dd1f481d59a8979d0d028e30dcdb10727e6ae068c779dddeba775ba5a2b |
|
Ascend 推理runtime库、benchmark工具、converter工具 |
Ascend |
Linux-x86_64 |
82ba2ac9218f832719030ed93d0fc0116380679f6739e8bc0fc9d978ad644261 |
|
Ascend 推理runtime库、benchmark工具、converter工具 |
Ascend |
Linux-aarch64 |
017a340d58b2b1a282fbeba2d2039dfe0932909a36f3e66180c57abc0bc49f58 |
|
轻鸿蒙推理runtime库、Micro库 |
Hi3516D |
OpenHarmony-aarch32 |
71261e4ee7e364fd4be04a9e919e84adb52e1e49d0eefff9725c5aa0cdb62ce9 |
|
Micro库 |
Cortex-M7 |
None |
c101088a199bd4348d75e374a4ab2926db83a3b48bfc68dbcad91a3316c045af |
Python 接口 Wheel安装包
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
converter及推理runtime |
CPU |
Linux-x86_64 |
b8e2385f385103d5ce1c841ba9d83e9a972bdace715b549aad0b9a8eb2ed15ac |
|
converter及推理runtime |
CPU |
Linux-aarch64 |
afb027398fc1474b60113480983da4df04caef7ef6ce3a93bd9ce1424468fbf4 |
|
Ascend converter及推理runtime |
Ascend |
Linux-x86_64 |
d0a9d979e769d5e2eed64a7603c843b98c2b62adaa5b181ae5a5d5d52db62a73 |
|
Ascend converter及推理runtime |
Ascend |
Linux-aarch64 |
05b9620b70ea5a964c6ab68da659ac838cc155e07bd39d4f54957071fe46551d |
1.8.1
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理和训练runtime库、推理和训练aar包、Micro库、benchmark工具 |
CPU |
Android-aarch32 |
7cc56ef7c90e41b6df980c3c43cb326a7c3f1bff8b827fbba3bf19ac5c827f6e |
|
推理和训练runtime库、推理和训练aar包、Micro库、benchmark工具 |
CPU/GPU |
Android-aarch64 |
4dbee13424c347549903426ae794bf4520115625f78871d6c18eca9b62533551 |
|
推理和训练runtime库、推理和训练jar包、Micro库、benchmark工具、converter工具、cropper工具 |
CPU |
Linux-x86_64 |
8a60dbbfee8a9f25853ebbf297c7c727cab365171347c701c0aa51cd0b1290ec |
|
推理runtime库、Micro库、benchmark工具、converter工具 |
CPU |
Linux-aarch64 |
598841c2018c977fbdec5fa2d1b3c5f82fc3ffca41c7726faee8dcb850e6ba03 |
|
推理runtime库、Micro库、benchmark工具、converter工具 |
CPU |
Windows-x86_64 |
44adced8645f69ea452e19069306d425cd8a78de377db18f801327689e28cf0e |
|
iOS 推理runtime库 |
CPU |
iOS-aarch32 |
22758b8ff36258e95a08dae79e09c70c150389adf7e74f543abb526b734fe64f |
|
iOS 推理runtime库 |
CPU |
iOS-aarch64 |
deb75103438d45fb6d7c42fa6feaef2a570d1ecc2a4bc2e1e4379b97b1413ec9 |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
0ee5427a82c1e876113e305e5613232ee93fc168d6eeb9c127261fb4b0ccefcc |
|
NNIE 推理runtime库、benchmark工具 |
Hi3516D |
Linux-aarch32 |
f3bb6245fdd0ae51d090d428ce8e65fc8ca8934fc1d1e2cd262f046130db0835 |
|
NNIE 推理runtime库、benchmark工具 |
Hi3519A |
Linux-aarch32 |
2552e6fac5c36d6c44c35eedba43a2e92aa28ebc0ec1125eb895c1d2ad62318c |
|
NNIE 推理runtime库、benchmark工具 |
Hi3559A |
Linux-aarch64 |
25a7d30c82c1e90b6e832f2edd23ac45c8f4f75ba1fdcef633b7ee99af6856d0 |
|
Ascend 推理runtime库、benchmark工具、converter工具 |
Ascend |
Linux-x86_64 |
a57de862c2e0809194a9183c09968f0f871f3b3c9be729c482b068ed672e6d32 |
|
Ascend 推理runtime库、benchmark工具、converter工具 |
Ascend |
Linux-aarch64 |
e91c74f75d581c2a88ad42f75dc4007eb254469588dd42cec264477f0eab425b |
|
轻鸿蒙推理runtime库、Micro库 |
Hi3516D |
OpenHarmony-aarch32 |
67bd55222bf8c67e01ff413ac2226b7f9444f5a9875b82aa6faf3fcae753c1bf |
|
Micro库 |
Cortex-M7 |
None |
5b1d1de977aac04f29c9e8e84931b0b7ec2a700e103d0940e5afc3ced5f49bd0 |
Python 接口 Wheel安装包
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
converter及推理runtime |
CPU |
Linux-x86_64 |
60c5a974ad76ee06f63ce784ae536d65c42fde6c41eaea3375ecbef5fc3efef0 |
|
converter及推理runtime |
CPU |
Linux-aarch64 |
1ca4cc3ac0df52ca578389502b7ace581d835683b18b1fecafd8bb10b89b3d84 |
|
Ascend converter及推理runtime |
Ascend |
Linux-x86_64 |
95e8ff0fa4296428781e70d32f90cb44ddd97c55372cc7d2a4f61f7d9adbdd2c |
|
Ascend converter及推理runtime |
Ascend |
Linux-aarch64 |
9d361d71b2c8efc809a7fba939b8e2ba1e06368ba7d8a0ff6e192dbc0d777fc9 |
1.8.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
0ff34972be36b9c9ef896094b1b0e5cd39d2253cbed86a41303fd9c21f1d5417 |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
f58139a33524191b6a42761948429aed4bc450d1e8c2986f58bfae1f5d81e626 |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Linux-x86_64 |
90be22a9f2a34329b46b1f0b08954a4b8cb2bd80bc69f8416f780f5f21f31547 |
|
推理runtime以及benchmark/converter工具 |
CPU |
Linux-aarch64 |
e17bfddc47feb17a430f3aea1d2f9b8895505518ec0a0dd109d9c4ed2f8d3b01 |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x86_64 |
f18b3e1697b27bcaf67bad2793494bf8babddd8d5b235b861cf693a44d9e9f14 |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
2270a87ec1888a5f9c62c37177476ee3aece5da8c98d2f379b83889e2f0db8c0 |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
ba368afa146f6e156d840e73145a81c6b2524a51dd0d138bcd2f9c2db5c59771 |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
9d4fce71a3963a0fa779f7996abb315fd9572fe8aca8b562f65f5793b42812b4 |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
b5739551b5217171e522aea27d93f64ce23d4bcd46e6b87bdaec2b09ec4c395b |
|
Ascend 推理runtime以及benchmark/converter工具 |
Ascend |
Linux-x86_64 |
a141a7886bde607fdc80fc7f35b28c14b11d36f28bbc3f917db6ae065cc32954 |
|
Ascend 推理runtime以及benchmark/converter工具 |
Ascend |
Linux-aarch64 |
0dd382c848bbb6ae2bd45a31098d311ec046bf219787480156cf1e0123e9f602 |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony-aarch32 |
8c93859a1a7925069906abc59c9d5f5fc53b989b852e80a2ffe3b85e7ff944ba |
Python 接口 Wheel安装包
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
converter及推理runtime |
CPU |
Linux-x86_64 |
80c08d67e2885f68e34c3b99356c1d95c72b1ad4cfdf29d08085ff7195753a08 |
|
converter及推理runtime |
CPU |
Linux-aarch64 |
a6365c96bc6ca04c7482cc2f2d3c1367d56ad7af13ed0e0d56b9349e773bcc30 |
|
Ascend converter及推理runtime |
Ascend |
Linux-x86_64 |
6539df5f1d59c679fd5a4e0b372b410ced226c2a3f4e1aec1a004b6bf2b6f1ba |
|
Ascend converter及推理runtime |
Ascend |
Linux-aarch64 |
8fed6d1ee3ab6673f733bb9071b19e1fc4a34a20813f9d5d4f6468fd6b5a36b6 |
1.7.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
26e31e9dc4a87698f0af18cd51ed219559b9a996b4123502608dd0a130270805 |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
e97a19f6d5a1e8cd6ff1b17d8445801eff496a2f02f1a3de8305a1459fe6eaef |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Linux-x86_64 |
f162a4b2ccfb5ec11f2d510e4df1d76bd1d9f9d6e321b1e068ac9e37db9dfd0e |
|
推理runtime以及benchmark/converter工具 |
CPU |
Linux-aarch64 |
320e2f076a068151c2796bb4a7ceb2190dc86ec78de0c39b52e6a496511b5e85 |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x86_64 |
41229b8be3997dd6ec648c7884b483ca4ca5926f3a7d9f78314af1cc2d2824ad |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
769fe933ebc77ebeeac5486ca418f0f17b491a01253cce8d7368742832bc6a99 |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
1ab0717c8d73ae64570f34ad203c6155f541a4420a24c24f26f0ffd919770832 |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
3965c150ec0428693217dd88dceca4bd01de08f40f8299410af94d9ba3a170bc |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
16b51013b0b2e70e52edeb909535ad9818f7771e6c108f7ab7e207ffcd8673ba |
|
Ascend 推理runtime以及benchmark/converter工具 |
Ascend310 |
Linux-x86_64 |
1fe2fdcabb9900c2d7e4ee9e239df157351574ed92b5f568eac23620f91f2ce8 |
|
Ascend 推理runtime以及benchmark/converter工具 |
Ascend310 |
Linux-aarch64 |
41027f2e070896d8440c89b8b75f66e9c92bbefa72b8b68e959770c2ffea5780 |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony-aarch32 |
269f0dc976523302d372a0500a3c93eb78ea048c51b9c5ae339174f224e1cbee |
1.6.1
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
afcfab6c13d46d25a19d4caedb66c2a16ccd63d5b5993ed1e6268633d03a7820 |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
7bba9b242c1b5bbb21e39313db0d1118aa568813466a82f3ddd689d570910b92 |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Linux-x86_64 |
667f8fae4762281858adfb71e62f11a52f641078496901d3f317b0df8c415452 |
|
推理runtime以及benchmark/converter工具 |
CPU |
Linux-aarch64 |
238589a227534f0514d2371fe9a92eb5ac2f181a22917ac2b5ef6969d51c518e |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x86_64 |
da0781020f8680a7ab5c00d9930d06fe0f9c295aa9e13b62b43c2b30e0bcad91 |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
6a8adfb3ea960af9213ed9474a9b2d9deef57ebe8cb9f9c822a314c7ff8165e7 |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
f7fe5477744bd1d7ee19d3292f150ac9665f60f613a3a8025a870b147885c76a |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
bde6c05aa90193d6fc0431ac3ef7a2c7e36e77db072955db68d16d6a90cfbe64 |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
fc4ab99ae0fb4cad4a63dad591da8d20be624e29b563aba39bba8239c0d64bfc |
|
Ascend 推理runtime以及benchmark/converter工具 |
Ascend310 |
Linux-x86_64 |
7b46fe61974d0295c1f816707e0330c26085ca5532c57f36b149fa851bcdbf7b |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony-aarch32 |
e70b2d30feafe0bbce3999c0bb551ed70edaf497c31fe84d70e4d5944093b253 |
1.6.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
d043803cffc8a0b75409aab3e4039f1e86756cf618af1538a76865e9fa4fd481 |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
25188266621f4cfedb24970a9a98ef6190fe02c9d034b7285f360da425ffe9d6 |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Linux-x86_64 |
90472996359f64509f38036ed8100605c76dcdc42453c2fc7156048eb981708c |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x86_64 |
4460b8f1bf321eca005074dccffb54d6d3164ba3f78ce34530ec20db4dbc9980 |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
72fe007660abe9c51d0a1852b094fb52d8bbd1610c989e79c9858937102aa59f |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
51bd5f7c21477d7856bea33d31e059f578b6b964a7c43e440e97c44b186db4a4 |
|
NNIE converter工具 |
CPU |
Linux-x86_64 |
81c2a5dadf51978d1c80f75c63fde4edefa2897792ac571fd33ffd35e338736b |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
8133c2326e2defa3614f86592d5691fdb410a4296898e254a33cd33a7e519b16 |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony-aarch32 |
d5daafac4bdcd0d03158e2a7cd3f881869b49cfb77d9654a24ddd967edbe5e91 |
1.5.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
83101ffc38de6c33c94d09bddd0efed31c23f468f76694ecbec5623db5f04afd |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
2f09c9f018c1141f2c415d48ff8ec01c17a2ad5e0b7d3233b8aa0612a2330a9e |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Ubuntu-x64 |
2359084653c1ddb55da738d5daf65e2cb0b0426032232ae455a72f8960961823 |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x64 |
2ed5d767be638787755c3855e28312d655e15219ad8bf500b43d6cea3a8d2dc6 |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
b3cd43f694f051e996cb8b39ed30137f92a4c324adaaceac1d75aa995bd7deb1 |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
9e919cdaf92fbb408ab169113182f7977d51409575e3a864a5a985a3004623d4 |
|
NNIE converter工具 |
CPU |
Ubuntu-x64 |
ab424c967b9ead17cebf1ef353b8b8129b7725e57f045be94a235741d12326c8 |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
976ce83dc89f3ebeab9b706ce0b1a8f49cab7a2c19cd860446c3464164054ceb |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony |
49cf161c90c259415718b0c6b4538bbc06d4629bfa604c4cbcbc04cf5ed1e3e8 |
1.5.0-rc1
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
7317c9359bd6da97c389780a9a6f91987c512a794d3a367e42e411eaacb181e0 |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
d0453a1e5f5f653cc77f160b3ea9252ee1ab96d0bce3e9128d4a54f47279ac46 |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Ubuntu-x64 |
81451541762699c259d13dd1d5a985fde653aeaf91052471bb918eb4f5045f49 |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x64 |
3ee80e7906f6173b805d985b51551c489bd23614b8747fb6e795a7303729913b |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
56a508d49605d46b65b84d4160155ebcbb21d1f9295baac1c96148f0a2b06e79 |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
69af90551a97f6d36379431a63fc31ebff102809fc5ec7e7a48e566f3a578ad8 |
|
NNIE converter工具 |
CPU |
Ubuntu-x64 |
ee0ec7fdc203a5106eccc02ca15c7543c4e6c957e605f8a9aaead4d966f10c4b |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
477628889908fb64ae4b7b989c0181593f9c66bebf119f6a07437911879bff8b |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony |
54654927f6bf0f90d316407c689575c47a263e0f3c608e5b430b9764b4c35f5c |
1.3.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU |
Android-aarch32 |
c1a950feec47a58871956cab74f0b6f76ad2f151dde990228782f76c0d8120df |
|
推理/训练runtime、推理/训练aar包、以及benchmark工具 |
CPU/GPU |
Android-aarch64 |
4def68662f5b249db0fd0f372fab1877e530fe32b6b9317869f01bedde892838 |
|
推理/训练runtime、推理/训练jar包、以及benchmark/codegen/converter/cropper工具 |
CPU |
Ubuntu-x64 |
2d0f77bb3c1a9489bc9511f334fb6cea3266d6bd4d600517b5aa7e58efab1310 |
|
推理runtime以及benchmark/codegen/converter工具 |
CPU |
Windows-x64 |
40c1abdeb4c5f5353844e9798c6a0f20565a5a3bc6de7da3cdc4a2df6fa15ef7 |
|
iOS 推理runtime |
CPU |
iOS-aarch32 |
b4bb1435887b04ce95be5429875e81c3e40a57b0c6182a35d58ea34b27d5fa5c |
|
iOS 推理runtime |
CPU |
iOS-aarch64 |
a28111c2bcc542a70ef98edfe007892855dcfcd9d40586fa98be962caefc26d3 |
|
NNIE converter工具 |
CPU |
Ubuntu-x64 |
73f4dffde69d24a8d0574e771bc6131a45a8fc5ebd18b34fc7afd30a0d149cb1 |
|
NNIE 推理runtime及benchmark工具 |
Hi3516D |
Linux-aarch32 |
07a0d0a8a8f257c01d06fa33f57969eb68385b430f2f5a3a4b09dba463c361d9 |
|
轻鸿蒙推理runtime |
Hi3516D |
OpenHarmony |
1bd481e93b6f3b2467a6d8f0bcc2da0221e4d023d74c584174dedb8854704748 |
1.2.0
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
推理runtime(cpp)、训练runtime(cpp)、推理aar包、以及benchmark/benchmark_train工具。 |
CPU |
Android-aarch32 |
7d073573385a69bff53542c395d106393da241682cd6053703ce21f1de23bac6 |
|
推理runtime(cpp)、训练runtime(cpp)、推理aar包、以及benchmark/benchmark_train工具。 |
CPU/GPU |
Android-aarch64 |
7f8400f0b97fa3e7cbf0d266c73b43a2410905244b04d0202fab39d9267346e0 |
|
推理runtime(cpp)、训练runtime(cpp)、推理jar包、以及benchmark/benchmark_train/codegen/converter/cropper工具。 |
CPU |
Ubuntu-x64 |
3b609ed8be9e3ae70987d6e00421ad4720776d797133e72f6952ba6b93059062 |
|
推理runtime(cpp)以及benchmark/codegen/converter工具。 |
CPU |
Windows-x64 |
bf01851d7e2cde416502dce11bd2a86ef63e559f6dabba090405755a87ce14ae |
|
推理runtime(cpp) |
CPU |
OpenHarmony |
a9987b25815cb69e0f630be1388486e8d727a19815a67851089b7d633bd2f3f2 |
1.1.0
推理
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
MindSpore Lite 模型转换工具(Converter) |
CPU |
Ubuntu-x64 |
d449e38a8493c314d1b5b1a127f62269192da785b012ff892eda775dedca3d82 |
|
CPU |
Windows-x64 |
5e50b7701b97ebe784095f2ba954fc6c377eb157fbc9aaeae2497e38cc4ee212 |
||
MindSpore Lite 模型推理框架(Runtime,含图像处理) |
CPU/GPU/NPU |
Android-aarch64/Android-aarch32 |
a19de5706db57e97a5f04ef08e0e383f8ea497c70bb60e60d056b31a603c0243 |
|
CPU |
Ubuntu-x64 |
176256c2fbef775f1a44aaeccae0c4eea6a60f41fc0baece5479dcb378155f36 |
||
CPU |
Windows-x64 |
30b5545245832a73d84732166f360c77cd09a7a4fe1fb922a8f7b80e7df326c1 |
训练
组件 |
硬件平台 |
操作系统 |
链接 |
SHA-256 |
---|---|---|---|---|
MindSpore Lite 模型转换工具(Converter) |
CPU |
Ubuntu-x64 |
f95a9db98c84ec3d97f88383ecc3832582aa9737ed287c33703deb0b419acf25 |
|
MindSpore Lite 模型训练框架(Runtime,含图像处理) |
CPU |
Android-aarch64/Android-aarch32 |
a6d8152f4e2d674c52af2c379f7d07858d30bc0dceef1dbc366e6fa16a5948b5 |
|
CPU |
Ubuntu-x64 |
1290f0adc790adc9edce654b9a629a9a323cfcb8453eb6bc19b779ef726282bf |
Ubuntu-x64版本包是在GCC版本大于等于7.3.0的环境编译得到的,所以该版本包的部署环境要求GLIBC版本大于等于2.27。
Android-aarch32的版本包不支持GPU和NPU。
MindSpore Lite还提供对Runtime的
libmindspore-lite.a
静态库裁剪工具,用于裁剪静态库文件,有效降低库文件大小。MindSpore Lite下载完成后,需要进行SHA-256完整性校验。