Commit Graph

451 Commits

Author SHA1 Message Date
xiaying c6f25cafc6 MNN:Sync: Sync Internal 2.9.3 2024-07-22 20:51:06 +08:00
xiaying 358251d11f MNN:Sync: Sync Internal 2.9.2 2024-07-04 17:13:44 +08:00
xiaying 65ec0ea406 MNN:Sync: Fix bug for llama2/llama3 attention fuse, refract llm usage 2024-06-15 15:39:59 +08:00
xiaying 5b19cbbbf1 [MNN:Sync] Sync Internal 2.9.1 2024-06-03 20:09:34 +08:00
xiaying 7cad2ee83f MNN:Sync Sync Internal 2.9.0 2024-05-11 19:17:02 +08:00
xiaying 5895243607 [MNN:Sync] Sync Internal 2.8.4 2024-04-19 11:58:21 +08:00
zhaode.wzd 2972fe71dc [MNN:Sync] Sync Internal 2.8.3 2024-03-13 14:55:54 +08:00
zhaode.wzd 970b63f3b4 [MNN:Sync] Sync Internal 2.8.2 2024-02-29 16:21:40 +08:00
zhaode.wzd 3b978d9d16 [MNN:Sync] Sync Internal 2.8.1 2023-12-27 17:26:44 +08:00
zhaode.wzd 72fa060815 [MNN::Bugfix] Some bugfix sync. 2023-12-12 10:49:15 +08:00
zhaode.wzd f6e60cc950 [MNN:Bugfix] Fix Windows compile error. 2023-12-04 16:40:47 +08:00
zhaode.wzd 387775be2a [MNN:Sync] Sync Internal 2.8.0 2023-12-04 11:19:10 +08:00
xiaying 3ff49cbf4a [MNN:Sync] Sync Internal 2.7.2 2023-10-18 10:31:02 +08:00
王召德 1bbb3b19a6
Merge pull request #2571 from xiezheng-XD/master
Update Dockerfile and testResource.py
2023-10-11 14:22:29 +08:00
zhaode.wzd bdf15442f4 [MNN:Sync] Sync Internal 2.7.1 2023-09-21 09:01:35 +08:00
xiaying db7851491f Fix bug for reconvert fp16 error 2023-09-05 17:42:20 +08:00
xiaying ea4f13d3cf [MNN:Sync] Sync Internal 2.7.0 2023-09-04 10:42:11 +08:00
xiezheng-XD 431a980310
Update testResource.py 2023-08-30 19:29:06 +08:00
xiezheng-XD 95a27fca78
Update testResource.py 2023-08-30 19:21:51 +08:00
xiezheng-XD cfb5633060
Update testResource.py
做了以下几处改动:
1.更新为Python 3 语法,并且优化了代码风格
2.使用os.path.join() 构建路径,革除了手动拼接路径的弊端
3.捕获错误输出流,构建用户友好的错误退出机制
2023-08-30 19:19:47 +08:00
xiezheng-XD c3070188fc
Update Dockerfile
修复了pip包名错误导致的下载问题
2023-08-30 19:13:40 +08:00
xiaying 98ba00c2f3 [MNN:Sync] Sync Internal 2.6.3 2023-08-21 14:53:27 +08:00
xiaying 84d6bd75ae [MNN:Sync] Sync Internal 2.6.2 2023-07-31 14:24:48 +08:00
xiaying ac5b331845 [MNN:Sync] Sync Internal Gitlab 2023-07-18 09:36:26 +08:00
zhaode.wzd 67eceb8abb [MNN:Sync] Sync Internal code, support low_memory for conv. 2023-06-27 10:33:16 +08:00
xiaying 930a9345c1 [MNN:Sync] Sync Internal 2.5.3 2023-06-16 09:42:45 +08:00
xiaying c70ecef660 [MNN:Sync] Sync Internal Gitlab: 2.5.1 2023-05-18 19:11:50 +08:00
zhaode.wzd 9fe780308c [MNN:Sync] Sync Internal 2.5.0 2023-04-27 15:11:05 +08:00
xiaying 39e2e7e45f [MNN:Sync] Sync Internal Gitlab 2.4.3 2023-04-18 18:54:46 +08:00
xiaying f26fc2efd8 [MNN:Sync] Sync 2.4.2 2023-04-11 11:12:00 +08:00
xiaying 7629ba674e [MNN:Sync] Sync Internal 2.4.1 2023-03-20 11:32:29 +08:00
xiaying 4e2ad365e8 [MNN:Sync] Sync Internal Gitlab 2023-02-28 10:41:24 +08:00
xiaying 4a609006eb [MNN:Sync] Sync Internal 2.3.1 2023-02-15 10:30:27 +08:00
zhaode.wzd b142695010 [Sync] Sync Internal changes. 2023-01-11 15:08:58 +08:00
xiaying b87429461d Fix compile bug for windows of GridSampleGrad 2022-12-30 16:08:26 +08:00
xiaying ee92164b1d Fix bug for compile in windows 2022-12-30 15:57:14 +08:00
xiaying d46b6b998d [MNN:Sync] Sync Internal 2.3.0 2022-12-30 15:18:58 +08:00
xiaying b1f5664ced [MNN:Internal] Sync to 2.2.3 2022-12-24 09:42:39 +08:00
xiaying ad5d243c9f [MNN:Sync] A few bugfixes
1. 支持 Onnx If 空子图的情况(这种情况是条件判断一定为真或假)
    2. 修正 Where 算子在 zeroshape 下维度计算出错的问题
    3. 修正 Reduce 计算 zeroshape 的非 prod 情况
    4. 修正 arch64-linux 上编译错误
    5. 修正 头文件 NNAPI 的注释错误
    6, 部分训练相关问题修正
2022-12-04 15:17:36 +08:00
xiaying 8908e80d44 [Sync] Sync Internal 2.2.2 2022-11-18 22:35:31 +08:00
zhaode.wzd c683c5c6c2 [Sync] Sync Internal Gitlab 2.2.1 2022-11-08 17:05:14 +08:00
xiaying acb3bb6c62 [Sync] Sync Internal Gitlab 2.2.0 2022-10-30 08:44:24 +08:00
xiaying 6027a441d5 Fix bug for Conv1dSqueezeMove for squeeze with multi inputs 2022-10-13 11:13:19 +08:00
xiaying db53f951e6 [Sync] Sync Internal 2.1.2 2022-09-30 10:02:52 +08:00
wtiandong 2768a397bd Merge remote-tracking branch 'origin/master' into interp3D 2022-09-27 17:13:37 +08:00
wtiandong 9e284352b5 merge interp3d_op_param into interp_op_param
merge interp3d_op_param into interp_op_param
2022-09-27 17:07:23 +08:00
NiuChenyu 5c67cfa8d6 Add ConvTranspose3D 2022-09-23 17:59:27 +08:00
jxt1234 4ba6232657
Merge pull request #2066 from DaydreamCoding/patch-16
MNN_BUILD_SHARED_LIBS=OFF build OnnxClip
2022-09-19 10:57:11 +08:00
jxt1234 52a275abb3
Merge pull request #2060 from DaydreamCoding/patch-13
Fix MNN_BUILD_PROTOBUFFER=OFF
2022-09-19 10:52:44 +08:00
DaydreamCoding d03dcfab7d
MNN_BUILD_SHARED_LIBS=OFF build OnnxClip 2022-09-17 17:03:01 +08:00
DaydreamCoding af57e11c75
converter caffe adapt Eltwise 2022-09-15 15:28:36 +08:00
DaydreamCoding 7756350155
Fix MNN_BUILD_PROTOBUFFER=OFF 2022-09-15 11:31:51 +08:00
zhaode.wzd 4753255227 [MNN:Sync] Sync Internal 2.1.1 contain below changes.
[Pymnn:Bugfix] Fix usage and small bug in pymnn.
    [Docs:Update] Update docs/cpp markdown
    [Docs:Update] Add docs check.
    [MNN:Update] Update VecHalf.hpp
    [MNN:Bugfix] Fix compile errors caused by "#define MNN_THREAD_LOCK_CPU"
    [Geometry:Bugfix] Fix bug for resize of broadcastto: https://github.com/alibaba/MNN/issues/2040
    [Docs:Update] Update inference api usage.
    [Pymnn:Bugfix] Close hiai load to fix resource leak.
    [MNN:Update] Down gradle version for demo compile
2022-09-09 17:24:37 +08:00
wtiandong e7e3d139b1 Fix Conv3D Relu6
In ONNX, when neural network is Conv3D -> Clip(1, -1), the converter will convert these two OPs in to Conv3DRelu6, regardless the min max values are not 0 and 6. This patch fix this bug.
2022-09-09 09:36:26 +08:00
wtiandong 71aae927ff Add Interp3D Support
1. add PyTorch interpolation 3D to Onnx to MNN converter
2. add interpolation3D nearest CPU/OpenCL implementation

all added OPs are verified
update opencl_program.cc

update opencl_program.cc
2022-09-07 15:56:08 +08:00
xiaying fafafef5c5 [MNN:Sync] Sync Internal 2.1.0 2022-08-31 20:11:16 +08:00
zhaode.wzd 76b8ace520 Sync Internal 2.0.5 2022-08-23 21:21:29 +08:00
xiaying 68708c5d66 Sync Internal 2.0.4 2022-08-12 10:30:48 +08:00
xiaying b6e4bf1050 [Converter:Bugfix] Fix bug for revertbuffer 2022-08-04 16:11:40 +08:00
xiaying 8330da263a [Sync] Sync internal 2.0.3 2022-07-22 09:59:30 +08:00
xiaying eb51926f84 [MNN:Sync] Sync internal Gitlab to 2.0.2 2022-07-19 13:52:07 +08:00
Brian Li c61a2f52f3 Use model base path for external data 2022-07-12 11:30:11 +08:00
Brian Li 6a23f85ab4 Add external data support for onnx converter 2022-07-12 11:30:11 +08:00
Brian Li 2801621ea5 Add missing headers 2022-07-12 01:28:27 +08:00
xiaying 2ec9495719 [MNN:Sync] Sync 2.0. 2022-07-11 10:56:37 +08:00
xiaying 2d13d6a495 2022-06-27 10:52:11 +08:00
xiaying d3ffdf4229 [MNN:Sync] Sync internal gitlab 2022-06-24 18:30:05 +08:00
xiaying aeaac3fde3 [MNN:Sync] Sync internal gitlab 2022-06-10 10:39:50 +08:00
xiaying 951a57b07c Support zero bias for LSTM 2022-06-07 19:57:43 +08:00
jxt1234 007a6e867f
Merge pull request #1943 from jokerz0624/bug_fix_converter
fix(converter): fix one bug of GridSampleOnnxClassic
2022-06-01 10:20:15 +08:00
xiaying c685d6ab85 Fix bug for python3 can't run convertOnnxTest 2022-05-31 20:55:38 +08:00
xiaying 38dd86f8db For Reshape for 0 / -1 , remove duplicateReshape will cause error 2022-05-31 20:55:38 +08:00
xiaying 692a544053 Fix onnx deconvolution bugfix 2022-05-31 19:52:09 +08:00
Joker 3196e4a93f fix(converter): fix one bug of GridSampleOnnxClassic 2022-05-31 19:05:37 +08:00
xiaying c95df2a932 Fix compile bug for arm82-armv7a 2022-05-30 17:24:20 +08:00
Yulv-git 77cc100153 Fix some typos. 2022-05-27 23:48:09 +08:00
xiaying b9bfa2342f Support GridSample Onnx for classic 2022-05-20 13:44:39 +08:00
xiaying 8f951f4143 Fix compile bug for MNN_SEP_BUILD=OFF and MNN_BUILD_SHARED_LIBS=ON 2022-05-13 14:02:37 +08:00
jxt1234 110145a704
Merge pull request #1871 from szh6/bugfix/master
Bugfix/master
2022-05-06 19:59:11 +08:00
xiaying c0aee19d32 [Sync] Sync internal gitlab 2022-05-06 19:51:20 +08:00
shufu 4f3940c852 fix(MNNConverter):fix the bug that iterators is invalidate when set --forTraining option true
Signed-off-by: shufu <shufuu@qq.com>
2022-03-01 17:28:29 +08:00
shufu dba8029eb4 fix(quantize):fix the bug that readClibrationFiles funtion doesn't work
when quantize model on MSVC
Signed-off-by: shufu <shufuu@qq.com>
2022-03-01 17:25:28 +08:00
xiaying 0c718e552b [Sync] Sync internal Gitlab 2022-02-18 11:30:27 +08:00
xiaying 1b626d72c3 [MNN:Sync] Sync internal gitlab 2022-01-04 10:50:40 +08:00
xiaying 82630820c0 Use Mod instead of Select 2021-12-13 14:11:38 +08:00
xiaying bd9ef418af Fix bug for Gather's indice < 0 2021-12-13 14:11:36 +08:00
xiaying b3c5feefdb [Converter:Bugfix] Support Onnx::TopK for dynamic shape 2021-12-10 15:16:28 +08:00
xiaying 69dba73dc7 [MNN:Sync] Sync internal gitlab
Main Feature:
1. Add OpenCV API and Numpy API Support
2. Protobuf move into MNN
3. Add more op for torchscript convert
4. Add recompute to speed up geometry compute
5. Add ModuleBasic Test
2021-11-30 10:10:53 +08:00
jxt1234 8f934c28c6
Merge pull request #1754 from cmdbug/master
[fix] fix typo
2021-11-12 17:18:48 +08:00
teng 6c56682327
[fix] fix typo
MacOS -> macOS
2021-11-12 10:30:06 +08:00
teng b7e6705ad5
[fix] fix typo
MacOS -> macOS
2021-11-12 10:29:44 +08:00
aaron-wu 8e773602bf fix(schema): merge parameters for RoiPooling and RoiAlign into one table as RoiParameters 2021-11-09 11:11:27 +08:00
aaron-wu 9f26b548b3 feat(convert): add ROIAlignOnnx convert(the onnx model file exported from torchvision.ops.roi_align) 2021-11-09 10:00:50 +08:00
jxt1234 d983b77ef2
Merge pull request #1743 from Napoleon-Jm/dev
fix: cli param keep input format.
2021-11-05 11:18:15 +08:00
恺心 a0b14095c2 fix: cli param keep input format. 2021-11-03 14:22:50 +08:00
xiaying 06d3d23c57 Temply turn inputCount = 1's depthmulti to convolution 2021-11-02 11:59:28 +08:00
xiaying d199ee6f49 Fix compute small error for OnnxScatterND 2021-11-02 10:55:15 +08:00
jxt1234 67f8bc94ef
Merge pull request #1735 from no5-aaron-wu/dev_aaron_wu
add onnx convert and uint test for ROIPooling OP
2021-10-29 14:43:22 +08:00
xiaying 8caa162f87 Fix ConvDilateFuse bug for width pad not equal height pad 2021-10-29 11:37:10 +08:00
insta360 cca07fdf98 feat(convert): add ROIPoolingOnnx Convert(the onnx model file export from torchvision.ops.roi_pool) 2021-10-29 10:26:05 +08:00