- MindSpore-tsc - mailweb.mindspore.cn

[DX -SIG]MindSpore DX-SIG meeting 2021 Apr 27
by lizi (D) 25 Apr '21

25 Apr '21

Dear MindSpore Members, This is a kind reminder for our MindSpore DX - SIG meeting on 2021 Apr 27 19:00 (UTC+08:00). The Developer eXperience Special Interest Group (DX - SIG) is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. The subject of the conference is : ��ǰ��£� 1 - MindSpore Developer Kanban (Routine) -- Zhaoboxuan MindSpore��ָ��ܼ��壨�� 2 �C Dashboard development details -- Zhaoboxuan ��ƽ̨��չ 3 �C Issue participants recommendation prototype �C Xiawenzong ��ڱ�ǩѵ��Issue��Զ��Ƽ�ԭ�ͷ�� 4 - OpenSource Playground 2 �C Mentorship �C Clement Li/Zhaoboxuan һ��濪Դ2 �C ��һ��ʦ 5 - TODO ��߷��Լ��⡣ We use WeLink-Zoom for this meeting. Click https://welink-meeting.zoom.us/j/821568025 then follow the instruction to download Zoom (if you haven��t installed) and join the meeting. You can also join the meeting by meeting ID 738908889 The calendar invitation has been updated as well. Thanks, Li Zi (Clement Li @ Gitee)

1 0

MindSpore DX-SIG meeting 2021 Apr 7
by lizi (D) 02 Apr '21

02 Apr '21

Dear MindSpore Members, This is a kind reminder for our MindSpore DX - SIG meeting on 2021 Apr 7 14:30 (UTC+08:00). The Developer eXperience Special Interest Group (DX - SIG) is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. The subject of the conference is : ��ǰ��£� 1 �C How to create issues in an open source project ��ύissue��ʵ�� 2 �C The design of issue auto-assign robot �Զ��Ƽ�issue��˹��߷�� 3 - TODO ��߷��Լ��⡣ We use WeLink-Zoom for this meeting. Click https://welink-meeting.zoom.us/j/166862733 then follow the instruction to download Zoom (if you haven��t installed) and join the meeting. You can also join the meeting by meeting ID 166862733 The calendar invitation has been updated as well. Thanks, Li Zi (Clement Li @ Gitee)

1 0

MindSpore DX-SIG meeting 2021 Mar 23
by lizi (D) 23 Mar '21

23 Mar '21

Dear MindSpore Members, This is a kind reminder for our MindSpore DX - SIG meeting on 2021 Mar 23 19:00 (UTC+08:00). The Developer eXperience Special Interest Group (DX - SIG) is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. The subject of the conference is : 当前议题如下： 1 - MindSpore comments network analysis and core developers recognition MindSpore评论网络分析及核心开发者识别 2 - PR process improvement and tools requirement PR流程优化与工具需求 3 - Label-reminder - A robot for reminding developer to add issue labels Label-reminder bot插件方案及代码评审如果您有其他的诉求或者分享可以继续增加议题。 We use WeLink-Zoom for this meeting. Click https://welink-meeting.zoom.us/j/738908889 then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 738908889 The calendar invitation has been updated as well. Thanks, Li Zi (Clement Li @ Gitee)

1 0

[MindSpore-sig] MindSpore DX-SIG meeting 2021 Mar 23
by lizi (D) 18 Mar '21

18 Mar '21

Dear MindSpore Members, This is a kind reminder for our first MindSpore DX - SIG meeting on 2021 Mar 23 19:00 (UTC+08:00). The Developer eXperience Special Interest Group (DX - SIG) is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. The subject of the conference is : 当前议题如下： 1 - MindSpore comments network analysis and core developers recognition MindSpore评论网络分析及核心开发者识别 2 - PR process improvement and tools requirement PR流程优化与工具需求 3 - Label-reminder - A robot for reminding developer to add issue labels Label-reminder bot插件方案及代码评审如果您有其他的诉求或者分享可以继续增加议题。 We use WeLink-Zoom for this meeting. Click https://welink-meeting.zoom.us/j/738908889 then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 738908889 The calendar invitation has been updated as well. Thanks, Li Zi (Clement Li @ Gitee)

1 0

[MindSpore-sig] MindSpore DX-SIG meeting 2021 Mar 9
by lizi (D) 05 Mar '21

05 Mar '21

Dear MindSpore Members, This is a kind reminder for our first MindSpore DX - SIG meeting on 2021 Mar 9 19:00 (UTC+08:00). The Developer eXperience Special Interest Group (DX - SIG) is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. The subject of the conference is : ��ǰ��£� 1 �C Issue and PR label system improvement Issue/PR ��ǩ��ϵ�Ż�� 2 �C Issue process improvement and tools requirement Issue��Ż�� 3 �CAn invitation for upcoming IBMC Shanghai meetup ��ڼ��IBMC�Ϻ�meetup�� ߷��Լ��⡣ We use WeLink-Zoom for this meeting. Click https://welink-meeting.zoom.us/j/151035006 then follow the instruction to download Zoom (if you haven��t installed) and join the meeting. You can also join the meeting by meeting ID 151035006 The calendar invitation has been updated as well. Thanks, Li Zi

1 0

MindSpore TSC meeting 2021 Feb 18
by Liu Yedong 17 Feb '21

17 Feb '21

Dear TSC Members, This is a kind reminder for our upcoming MindSpore TSC Monthly meeting on 2021 Feb 18 16:00 (UTC+08:00). This TSC meeting takes place on the first day after the Chinese New Year, we review on a minor release of MindSpore version 1.1.1, and a new SIG group. We use WeLink-Zoom for this month's meeting. Click https://welink-meeting.zoom.us/j/448774327 then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 448774327 The calendar invitation has been updated as well. Thanks, Yedong

1 0

MindSpore TSC meeting 2021 Jan 21
by Liu Yedong 20 Jan '21

20 Jan '21

Dear TSC Members, This is a kind reminder for our upcoming MindSpore TSC Monthly meeting on 2021 Jan 21th 16:00 (UTC+08:00). We review the first calendar year of MindSpore, discuss new SIGs/WGs and the 1.1.0 version. We use WeLink-Zoom for this month's meeting. Click https://welink-meeting.zoom.us/j/543118603 <https://www.google.com/url?q=https://zoom.us/j/96044179444?pwd%3DRmg5cWRFc0…> then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 543118603 The calendar invitation has been updated as well. Thanks, Yedong

1 0

MindSpore TSC meeting Dec 17
by Liu Yedong 16 Dec '20

16 Dec '20

Dear TSC Members, This is a kind reminder for our upcoming MindSpore TSC Monthly meeting on Dec 17th 16:00 (UTC+08:00). We review the release plan of MindSpore v 1.1.0 , discuss a new SIG and MindCon competition. We use WeLink-Zoom for this month's meeting. Click https://welink-meeting.zoom.us/j/209041120 <https://www.google.com/url?q=https://zoom.us/j/96044179444?pwd%3DRmg5cWRFc0…> then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 209 041 120 The calendar invitation has been updated as well. Thanks, Yedong

1 0

MindSpore TSC Monthly Meeting
by Liu Yedong 18 Nov '20

18 Nov '20

Dear TSC Members, This is a kind reminder for our upcoming MindSpore TSC Monthly meeting on Nov 19th 16:00 (UTC+08:00). Zhipeng cannot attend this meeting ,so Yedong Liu will be the host this time. In this month, we have a minor release of v1.0.1 version , a new SIG and complete 21-days online tutorial. We use WeLink-Zoom meeting for this month's meeting. Click https://welink-meeting.zoom.us/j/850801418 <https://www.google.com/url?q=https://welink-meeting.zoom.us/j/850801418&sa=…> then follow the instruction to download Zoom (if you haven’t installed) and join the meeting. You can also join the meeting by meeting ID 850801418 The calendar invitation has been updated as well. Thanks, Yedong

1 0

邀请：MindSpore Monthly TSC Meeting (EU-Asia Friendly) @ 2020年11月19日 (周四) 16:00 - 18:00 (CST) (mindspore-tsc@mindspore.cn)
by liuyedong1992＠gmail.com 16 Nov '20

16 Nov '20

您受邀参加以下活动。标题： MindSpore Monthly TSC Meeting (EU-Asia Friendly) Please use ZOOM for Nov 19's meeting:Join Zoom Meeting:https://welink-meeting.zoom.us/j/850801418Meeting ID: 850 801 418Password: Agenda:Sent out to TSC mainling list (mindspore-tsc) before meeting 时间： 2020年11月19日 (周四) 16:00 – 18:00 中国标准时间 - 上海日历： mindspore-tsc(a)mindspore.cn 参与者： * liuyedong1992(a)gmail.com- 组织者 * zhipengh512(a)gmail.com- 创建者 * joel.falcou(a)lri.fr * mggong(a)mail.xidian.edu.cn * a.storkey(a)ed.ac.uk * han.xiao(a)conic.ai * zhoubin812(a)huawei.com * jinxuefeng(a)huawei.com * chenhkust(a)gmail.com * yongle.li(a)gmail.com * u1mail2me(a)gmail.com * iliya.slavutin(a)gmail.com * leichen(a)cse.ust.hk * dcszj(a)mail.tsinghua.edu.cn * jin.li(a)apulis.com * xiangyangli(a)ustc.edu.cn * gorlatch(a)uni-muenster.de * yhtian(a)pku.edu.cn * sgshan(a)ict.ac.cn * chenjj2(a)sdc.icbc.com.cn * prp(a)imperial.ac.uk * thsu(a)hit.edu.cn * fan.yu(a)huawei.com * huangzhipeng(a)huawei.com * cheng.ding(a)huawei.com * wanhanyang(a)huawei.com * zhangqiang27(a)huawei.com * ch.l(a)huawei.com * dengfuqiang(a)huawei.com * slavutin.iliya(a)huawei.com * wanghui71leon(a)gmail.com * mindspore-tsc(a)mindspore.cn 活动详细信息：https://calendar.google.com/calendar/event?action=VIEW&eid=N21lZjdjMTJkZT… 来自 Google 日历的邀请：https://calendar.google.com/calendar/ 由于您是该活动的参加者，所以您会通过 mindspore-tsc(a)mindspore.cn 帐号收到此邮件。要停止接收有关此活动的后续更新，请拒绝此活动。另外，您可以在 https://calendar.google.com/calendar/注册一个Google帐号并控制所有针对日历的通知设置。转发此邀请后，任何收件人都可以向组织者发送回复，并且被添加到邀请对象列表；无论其邀请状态为何，他们都可以邀请其他人；或者，他们也可以修改您的回复。要了解详情，请前往 https://support.google.com/calendar/answer/37135#forwarding

1 0

Adaptive Distributed Training SIG
by sunbeilei 28 Oct '20

28 Oct '20

Dear All, I propose to set up the Adaptive Distributed Training SIG. Improving the resource utilization of a deep learning cluster is of paramount concern for many AI practioners. A promising approach is to use elastic deep learning systems. These systems allow users to dynamically change the number of training resources allocated to training jobs. Hence, practitioners can pack a large number of training jobs into a cluster, significantly improving cluster utilization. Though promising, elastic deep learning systems are difficult to be deployed in practice. State of the art data-parallel elastic ddp learning systems couple the number of training resources with a critical learning hyper-parameter: the batch size of the SGD. Any scaling decisions made by the cluster scheduler therefore must alter the SGD batch size, which affects training results and can even make the training fail to converge. In this SIG, we will enable the cluster scheduler to dynamically scale a training job without affecting its SGD batch size. To achieve this, we want to explore a novel method to decouple the SGD batch size and the number of training resources, so that the change of training resources does not affect the convergence. https://gitee.com/mindspore/community/blob/master/design/meps/mep-adaptive/… https://gitee.com/mindspore/community/tree/master/sigs/adaptivetraining Best regards Beilei Sun

1 0

MindSpore TSC Monthly Meeting Oct 15th
by Zhipeng Huang 15 Oct '20

15 Oct '20

Dear TSC Members, This is a kind reminder for our upcoming MindSpore TSC Monthly meeting on Oct 15th. As I introduced in the last email, we have released the v1.0 version at the end of September, and the community is gaining momentum. In order to further adjust our community structure to adapt to a growing participation, there will be several changes proposed at the Oct 15th meeting: 1. Consolidate into one single monthly meeting for EU-Asia friendly timeslot. 2. Adding two new non-voting organizations : Expert Committee with its subsidiary groups and User Committee with its subsidiary groups. The two proposals will be discussed and then submitted as a resolution PR to the mindspore/community repo for final landing. Look forward to meeting you after a long break :) -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 1

Cancellation of this month's TSC Meeting
by Zhipeng Huang 16 Sep '20

16 Sep '20

Dear TSC Members, In light of several China based TSC members attending Huawei Connect 2020 conference and the development phase of 1.0 release into a critical stage, let's cancel tomorrow's meeting. We will still post 1.0 release plan, the branch maintenance policy and milestone policy after 1.0 via the mailing list for your review, please do not hesitate to provide your key insight. One interesting development is that we will host our first *MindSpore Study Group Russia online meetup *on Thursday UTC0700, please check out my twitter post here <https://twitter.com/nopainkiller/status/1306209352399777794> for meeting information. I will as always try to record the event and post to YouTube if possible. Look forward to more MSGs in other parts of the world :) Stay tuned and stay healthy ! -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 0

MindSpore TSC Monthly Meeting Aug 20, 2020
by Zhipeng Huang 19 Aug '20

19 Aug '20

Dear TSC Members, This is a kind reminder for our monthly meeting tomorrow. Besides the usual items on community progress, we will introduce 2 new SIGs and 2 new WGs :) Respectively: *SIG-MSLITE* for mobile and IoT support *SIG-DPP* for deep probabilistic programming support *WG-MM* for community collaboration on deep learning support in Molecular Modeling *WG-Research* for community collaboration on several research topics For the US-Asia timezone friendly meeting, please join https://zoom.us/j/98865564258 <https://www.google.com/url?q=https://zoom.us/j/98865564258&sa=D&source=cale…> and for EU-Asia timezone friendly meeting, please join https://zoom.us/j/99378539769 <https://www.google.com/url?q=https://zoom.us/j/99378539769&sa=D&source=cale…>. The calendar invitation has been updated as well. -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 0

[Release]Plan For 0.7.0-beta
by ljllujiale 12 Aug '20

12 Aug '20

Please find below the release plan for our upcoming release 0.7.0 beta targeting the end of August. We've also documented the release management related info at https://gitee.com/mindspore/mindspore/wikis/August%202020?sort_id=2684223 Please also note that this is not the final release note, which will contain additional information on bug fixes and so forth. Feel free to ask any questions or provide any feedback through the list! Planned Features and Improvements For 0.7.0-beta release MindSpore Major Features and Improvements Ascend 910 Training and Inference Framework New models Mask-RCNN: a simple and flexible deep neural network extended from Faster-RCNN for object instance segmentation on COCO 2017 dataset. TinyBert: a smaller and faster version of BERT using transformer distillation for natural language understanding on GLUE benchmark. SENet: a convoluation neural network using Squeeze-and-Excitation blocks to improve channel interdependencies for image classification on ImageNet 2017 dataset. Inception V3: the third version of Inception convolutional architectures for image classification on ImageNet 2012 dataset. Hub: third-party model hosting, downloading and usage; interface definition specifications; benchmarking process and test of preset model on the device side. Frontend and user interface Embedding operator high-level packaging to support segmented by field for Wide&Deep. Load multi-node checkpoint into single-process to support host-device hybrid inference. Support Concat/Tile/Strideslice distributed operators. Support cumulative gradient and batch training split. Support variable parameter input for Cell object. Parameter mixed calculation optimization for pynative mode. Deep Probabilistic Programming Support statistical distributions classes used to generate stochastic tensors. Support probabilistic inference algorithms. Support BNN layers used to construct BNN. Support interfaces for the transformation between BNN and DNN. Support uncertainty estimation to estimate epistemic uncertainty and aleatoric uncertainty. Executor and performance optimization Minspore graph compilation process performance improved by 20%. Decoupling C++ and Python modules to achieve separate compilation of core modules. Serving module supports restful interface. Data processing, augmentation, and save format Support automatic data augmentation Support GNN distributed cache in single node Batch operator optimization Other Hardware Support GPU platform New model supported: VGG16, ResNet101, InceptionV3, NASNet, Transformer, TinyBert, YoloV3-DarkNet, MNASNet, EffecientNet-B0, ShuffleNet. Support some distributed operators in ResNet50 and Wide&Deep. Support automatic parallel for Wide&Deep. Support function funcsi (such as switch-case). Support distributed training with parameter server. Performance optimization of the distributed training with allreduce. Performance optimization of the mixed precision training. Performance optimization of the pynative mode. Performance optimization of the convolution operator, batch normalization operator. Serving module supports GPU backend. CPU platform Support MobileNetV2 Re-Training: Re-train the network with different class number. MindSpore Lite Major Features and Improvements Converter Support third party model, including TFLite/Caffe/ONNX. Add 96 TFLite op. Add 18 Caffe op. Add support for windows. Add 11 optimized passes, include fusion/const fold. Support aware-training and Post-training quantization. API Add JAVA APIs. Add pre/post data processing APIs. CPU Add 100+ops，support fp32, int8/uint8, FP16 ops Support fast convolution algorithms: Sliding Window, Img2col + Gemm, Strassen, Winograd Support assembly/neon instruction. Support CPU fp16 and sdot on ARM v8.2+. GPU Add 50+ ops for OpenCL. Support image2D/buffer format. Optimize online initialization time. add optimized convolution1X1/3X3/depthwise/convolution_transposed for OpenCL. NPU Support Kirin NPU. Tool & example Add benchmark and TimeProfile tools. Add image classification and object detection Android Demo. GraphEngine Major Features and Improvements Conditional operator memory supports separate allocation of 4G memory space; In the zero-copy scenario, atomic_clean supports cleaning the memory of each part of the output when the network is multi-output; Support profiling of multiple levels of data in inference scenarios; In the online compilation scenarios, GE compilation time optimization. MindArmour Major Features and Improvements Privacy leakage evaluation. Using Membership inference to evaluate the privacy-preserving of AI model. Fuzzing based Adversarial Robustness testing. Coverage-guided test set generation. MindInsight Major Features and Improvements Support GPU operator profiling Web UI supports English Computation graph show full_name in anf_ir.proto as node name Provides a code wizard tool to quickly generate classic network scripts in a friendly and easy way. Thanks Lujiale

1 0

MindSpore TSC Monthly Meeting Jul 16th
by Zhipeng Huang 16 Jul '20

16 Jul '20

Dear TSC Members, This is a kind reminder for today's TSC meeting, please access the meeting via ZOOM link provided in the calendar invitation. Given the current limitation of ZOOM subscription restriction, we have 40mins of discussion time. As usual, US-Asia friendly meeting time is about in 1 hour (Beijing time 12:00), and EU-Asia friendly meeting time will start on UTC0800 (Beijing time 16:00) The initial agenda is as follow: Roll Call Approval of previous minute Community Progress Update (with 0.5.0 release) Release Plan Peview (0.6.0 release preview) Operational Matters (MindSpore Study Group) -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

2 1

[Release]Plan For 0.6.0-beta
by ljllujiale 08 Jul '20

08 Jul '20

Please find below the release plan for our upcoming release 0.6.0 beta targeting the end of July. We've also documented the release management related info at https://gitee.com/mindspore/mindspore/wikis/July%202020?sort_id=2378784 Please also note that this is not the final release note, which will contain additional information on bug fixes and so forth. Feel free to ask any questions or provide any feedback through the list! Planned Features and Improvements For 0.6.0-beta release MindSpore Ascend 910 Training and Inference Framework New models MaskRCNN: a simple and flexible deep neural network for object instance segmentation on COCO 2014 dataset. TinyBERT: a smaller version of the base BERT model for natural language understanding using transformer distillation and two-stage learning framework. Frontend and user interface Supports user side operator compilation and graph execution error rendering. Uniform definition dynamic learning rate behavior in optimizers. Support ps and allreduce mixing in optimizers. Support IndexSlice in sparse expression. Support mixed precision in pynative mode. Support the process of forward execution, dynamically constructing reverse graph. Support use parent construct method during construct. Support asynchronous execution save checkpoint file. Support implicit type conversion in pynative mode. Executor and performance optimization Decouple C++ and python, so make the architecture more extensible. Parameter Server for distributed deep learning supported, and is verified in Wide&Deep model. Quantitative training of YoloV3 on Ascend-910 is supported, and quantitative inference on Ascend-310 is also supported. Serving：a flexible service deployment framework for deep learning models. Memory reuse is enhanced, and the batch size of Bert large model is increased from 96 to 160 on a single server. Data processing, augmentation, and save format Support single cache after data processing Support MindRecord save operator after date processing Support automatic fusion operator, such as decode/resize/crop Support CSV dataset loading Other Hardware Support GPU platform New model supported: ShuffleNet, NASNet, RetinaFace + MobileNetV2. Support hyperparametric search and data enhanced automl on GPU. Support Resnet50 automatic parallel in GPU backend. GraphEngine GE supports function control operators such as If/Case/While/For. In a single operator call scenario, GE supports recording the correspondence between operators and tasks for performance commissioning. GE supports new operator overflow positioning solution. MindArmour Differential privacy model training Optimizers with differential privacy Differential privacy model training now supports some new policies. Adaptive Norm policy is supported. Adaptive Noise policy with expontional decrease is supported. Differential Privacy Training Monitor A new monitor is supported using zCDP as its asmpotetic budget estimator. MindInsight Provide monitoring capabilities for each of Ascend AI processor and other hardware resources,including CPU, memory and disk. Visualization of weight, gradient and other tensor data in model training Provide tabular form presentation of tensor data. Provide histogram to show the distribution of tensor data and its change over time. Thanks Lujiale

1 0

[Meeting]Jitsi Technical Problem
by Zhipeng Huang 18 Jun '20

18 Jun '20

Dear TSC Members, It seems there are issues with Jitsi today, so let's cancel the US-Asia meeting slot for today, and I will try to setup a ZOOM call link for the EU-Asia meeting later today. All the meeting materials will be shared later for offline review -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 2

[Release]Plan For 0.5.0-beta
by 刘存伟 18 Jun '20

18 Jun '20

Dear TSC members, Please find below the release plan for our upcoming release 0.5.0-beta targeting the end of June. We’ve also documented the release management related info at https://gitee.com/mindspore/mindspore/wikis/June%202020?sort_id=2339889 Please also note that this is not the final release note, which will contain additional information on bug fixes and so forth. Feel free to ask any questions or provide any feedback through the list ! Planned Features and Improvements For 0.5.0-beta release MindSpore Ascend 910 Training and Inference Framework New models MASS: a pre-training method for sequence to sequence based language generation tasks on Text Summarization and Conversational Response Generation using News Crawls 2007-2017 dataset, Gigaword corpus and Cornell movie dialog corpus. Transformer: a neural network architecture for language understanding on WMT 2014 English-German dataset. DeepFM: a factorization-machine based neural network for CTR prediction on Criteo dataset. Wide&Deep: jointly trained wide linear models and deep neural networks for recommender systems on Criteo dataset. DeepLabV3: significantly improves over our previous DeepLab versions without DenseCRF post-processing and attains comparable performance with other state-of-art models on the PASCAL VOC 2007 semantic image segmentation benchmark. InceptionV3: the third edition of Google's Inception Convolutional Neural Network on ImageNet 2012 dataset. Faster-RCNN: towards real-time object detection with region proposal networks on COCO 2017 dataset. SSD: a single shot multibox detector for detecting objects in images using a single deep nueral network on COCO 2017 dataset. GoogLeNet: a deep convolutional neural network architecture codenamed Inception V1 for classification and detection on CIFAR-10 dataset. Frontend and User Interface Complete numpy advanced indexing method. Supports value and assignment through tensor index. Some optimizers support separating parameter groups. Different parameter groups can set different learning_rate and weight_decay. Support setting submodule's logging level independently, e.g. you can set logging level of module A to warning and set logging level of module B to info. Support weights to be compiled according to shape to solve the problem of large memory overhead. Add more operators implement and grammar support in pynative mode. To be consistent with graph mode. Executor and Performance Optimization Support doing evaluation while in training process, so that the accuracy of training can be easily obtained. Enable second-order optimization for resnet50, which can achieve 75.9% accuracy in 45 epochs (Resnet50 @ImageNet). Optimize pynative implementation and improve it's execution performance. Optimize summary record implementation and improve its performance. Data processing, augmentation, and save format Support simple text processing, such as tokenizer/buildvocab/lookup. Support simple gnn dataset processing. Other Hardware Support GPU platform New models supported: MobileNetV2, MobileNetV3. Support quantization aware training. Support mixed precision training. Support device memory swap in/out during training. CPU platform New model supported: LSTM. GraphEngine Optimize Allreduce trailing parallelism, rebuild the calculation graph dependencies, adjust the calculation order, and maximize the efficiency of calculation and gradient aggregation communication in parallel, especially in large data volume gradient aggregation and low bandwidth/large cluster scenarios You can get a bigger income. Advance constant folding, variable fusion, conversion operator related optimization pass to the end of the graph preparation. Modify memory allocation algorithm, optimize GE memory allocation, and reduce memory usage in training multi-PCS scenarios. Support IR composition, model compilation, inference execution in the same process. MindArmour Optimizers with differential privacy Differential privacy model training now supports both Pynative mode and graph mode. Graph mode is recommended for its performance. MindInsight MindSpore Profiler Provide performance analyse tool for the input data pipeline. Provide timeline analyse tool, which can show the details of the streams/tasks. Provide a tool to visualize the step trace information, which can be used analyse the general performance of the neural network in each phase. Provide profiling guides for the users to find the performance bottlenecks quickly. CPU summary operations support for CPU summary data Over threshold warn support in scalar training dashboard Provide more user-friendly callback function for visualization Provide unified callback SummaryCollector to log most commonly visualization event. Discard the original visualization callback SummaryStep, TrainLineage and EvalLineage. SummaryRecord provide new API add_value to collect data into cache for summary persistence. SummaryRecord provide new API set_mode to distinguish summary persistence mode at different stages. Mindconverter supports conversion of more operators and networks, and improves its ease of use. Thanks Liucunwei This e-mail and its attachments contain confidential information from HUAWEI, which is intended only for the person or entity whose address is listed above. Any use of the information contained herein in any way (including, but not limited to, total or partial disclosure, reproduction, or dissemination) by persons other than the intended recipient(s) is prohibited. If you receive this e-mail in error, please notify the sender by phone or email immediately and delete it!

1 1

[Meeting]MindSpore TSC Monthly Meeting Jun 18th Agenda
by Zhipeng Huang 15 Jun '20

15 Jun '20

Dear TSC Members, Please find the initial agenda for TSC meeting this week on Thursday as follows: 1.Roll Call and approval for previous minutes. 2.Community Progress Update 3.0.3.0 release note review 4.0.5.0 release plan preview 5.Operational Matters (setup learning groups) Please feel free to suggest any topics you would like to add. Again we will have US-Asia friendly meeting starting at noon Beijing time and then EU-Asia friendly meeting starting at 16:00pm Beijing time. -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 0

[Meeting]MindSpore TSC Monthly Meeting May 21th Agenda
by Zhipeng Huang 18 May '20

18 May '20

Dear TSC Members, Please find the initial agenda for TSC meeting this week on Thursday as follows: 1.Roll Call and approval for previous minutes. 2.Community Progress Update 3.0.2.0 release review 4.0.3.0 release preview 5.Operational Matters Please feel free to suggest any topics you would like to add. Again we will have US-Asia friendly meeting starting at noon Beijing time and then EU-Asia friendly meeting starting at 16:00pm Beijing time. -- Zhipeng (Howard) Huang Principle Engineer OpenStack, Kubernetes, CNCF, LF Edge, ONNX, Kubeflow, OpenSDS, Open Service Broker API, OCP, Hyperledger, ETSI, SNIA, DMTF, W3C

1 0

[Meeting]MindSpore TSC Monthly Meeting May 21th
by huangzhipeng 16 May '20

16 May '20

Dear TSC Members, As discussed at the meeting last month, we will have our monthly meeting on Thursday the third week of every month. A calendar invitation has been sent to you for the monthly meeting reminder. For May meeting, we will have the meeting on May 21 with US-Asia friendly meeting starting 12:00 (UTC+8) and then EU-Asia friendly meeting starting 16:00 (UTC+8) respectively. Sorry for any confusion that cause unnecessary problems. Please feel free to contact me if you have not received the calendar invitation. I will send out the agenda on Monday to the mailing list for your review. Thank you very much !

1 0

[Release]Plan For 0.3.0-alpha
by huangzhipeng 12 May '20

12 May '20

Dear TSC members, Please find below the release plan for our upcoming release 0.3.0-alpha targeting the end of May. We’ve also documented the release management related info at https://gitee.com/mindspore/mindspore/wikis/May%202020?sort_id=2196349 Please also note that this is not the final release note, which will contain additional information on bug fixes and so forth. Feel free to ask any questions or provide any feedback through the list ! Planned Features and Improvements For 0.3.0-alpha release MindSpore Ascend 910 Training and Inference Framework * New models * MASS: a pre-training method for sequence to sequence based language generation tasks on Text Summarization and Conversational Response Generation using News Crawls 2007-2017 dataset, Gigaword corpus and Cornell movie dialog corpus. * Transformer: a neural network architecture for language understanding on WMT 2014 English-German dataset. * DeepFM: a factorization-machine based neural network for CTR prediction on Criteo dataset. * Wide&Deep: jointly trained wide linear models and deep neural networks for recommender systems on Criteo dataset. * DeepLabV3: significantly improves over our previous DeepLab versions without DenseCRF post-processing and attains comparable performance with other state-of-art models on the PASCAL VOC 2007 semantic image segmentation benchmark. * InceptionV3: the third edition of Google's Inception Convolutional Neural Network on ImageNet 2012 dataset. * Faster-RCNN: towards real-time object detection with region proposal networks on COCO 2017 dataset. * SSD: a single shot multibox detector for detecting objects in images using a single deep nueral network on COCO 2017 dataset. * GoogLeNet: a deep convolutional neural network architecture codenamed Inception V1 for classification and detection on CIFAR-10 dataset. * Frontend and User Interface * Complete numpy advanced indexing method. Supports value and assignment through tensor index. * Some optimizers support separating parameter groups. Different parameter groups can set different learning_rate and weight_decay. * Support setting submodule's logging level independently, e.g. you can set logging level of module A to warning and set logging level of module B to info. * Support weights to be compiled according to shape to solve the problem of large memory overhead. * Add more operators implement and grammar support in pynative mode. To be consistent with graph mode. * Executor and Performance Optimization * Support doing evaluation while in training process, so that the accuracy of training can be easily obtained. * Enable second-order optimization for resnet50, which can achieve 75.9% accuracy in 45 epochs (Resnet50 @ImageNet). * Optimize pynative implementation and improve it's execution performance. * Optimize summary record implementation and improve its performance. * Data processing, augmentation, and save format * Support simple text processing, such as tokenizer/buildvocab/lookup. * Support simple gnn dataset processing. Other Hardware Support * GPU platform * New models supported: MobileNetV2, MobileNetV3. * Support quantization aware training. * Support mixed precision training. * Support device memory swap in/out during training. * CPU platform * New model supported: LSTM. GraphEngine * It supports dynamic batches and shapes with certain fixed levels. * Scope fusion interfaces are opened allowing user defined scope fusion rules. * Enhance the maintenance and measurement capability. * A package of compiled libraries is generated after compilation to facilitate code deployment.(#I1D1QF<https://gitee.com/mindspore/dashboard/issues?id=I1D1QF>) MindArmour Differential Privacy is coming! By using Differential-Privacy-Optimziers, one can still train a model as usual, while the trained model preserved the privacy of training dataset, satisfying the definition of differential privacy with proper budget. * Optimizers with Differential Privacy * Some common optimizers now have a differential privacy version (SGD/Adam-SGD). We are adding more. * Automatically and adaptively add Gaussian Noise during training to achieve Differential Privacy. * Automatically stop training when Differential Privacy Budget exceeds. * Differential Privacy Accoutant * Calculate overall budget consumed during training, indicating the altimate protect effect. MindInsight * Profiling * Provide easy to use apis for profiling start/stop and profiling data analyse (on Ascend only). * Provide AICore and AICPU operators performance display and analysis on MindInsight UI. * Large scale network computation graph visualization. * Optimize summary record implementation and improve its performance. * Improve lineage usability * Optimize lineage display and enrich tabular operation. * Decouple lineage callback from SummaryRecord. * Support scalar compare of multiple runs. * Scripts conversion from other frameworks. Support for converting PyTorch scripts within TorchVision to MindSpore scripts automatically. 黄之鹏 Zhipeng (Howard) Huang 主任工程师 - 智能计算&IT开源生态部 Principle Engineer - Intelligent Computing & IT Open Source Ecosystem Department 华为技术有限公司 Huawei Technologies Co., Ltd. Tel : +86 755 28780808 / 18576658966 Email : huangzhipeng(a)huawei.com [cid:image001.png@01D62890.24914470] This e-mail and its attachments contain confidential information from HUAWEI, which is intended only for the person or entity whose address is listed above. Any use of the information contained herein in any way (including, but not limited to, total or partial disclosure, reproduction, or dissemination) by persons other than the intended recipient(s) is prohibited. If you receive this e-mail in error, please notify the sender by phone or email immediately and delete it!

1 0

Report From SIG Compiler
by Suteng 29 Apr '20

29 Apr '20

Hi all, Please see the report from SIG MindSpore Compiler - Auto Parallel #MindSpore Compiler Special Interest Group (SIG) This is the working repo for the Compiler special interest group (SIG). This repo contains all the artifacts, materials, meeting notes and proposals regarding ANF IR, auto differentiation, auto parallel, graph optimizer, VM and any other programs for high level graph compilation in MindSpore. Feedbacks and contributions are welcome. Auto Parallel: Auto Parallel is a novel approach to achieve automatic parallel training and inference, including data parallel, model parallel and hybrid parallel. Auto Parallel is part of MindSpore graph compiler, and aims to improve the usability and performance of model training and inference. #Current Progress： Support resnet50 auto parallel training. #Development plan: We plan to support ReID, recommendation, and NLP models, such as resnet, wid&deep and bert, in next few months. At present, we have support resnet which used in ReID. Some key feature plan to do: 1. Support GPU/CPU backend. At present, only support Ascend. 2. Support auto parallel inference, using multi device to speed up inference, and run bigger model. 3. Improve cost model to search a better performance parallel strategy. 4. Develop more parallel operators. #SIG Leads * Ding Jian (Huawei) #Logistics * SIG leads will drive the meeting. * Meeting annoucement will be posted on our gitee channel: https://gitee.com/mindspore/community/tree/master/sigs/compiler * Feedbacks and topic requests are welcome by all. #Discussion * Slack channel https://app.slack.com/client/TUKCY4QDR/C011RSWRN3S?cdn_fallback=2 * Documents and artifacts: https://gitee.com/mindspore/community/tree/master/sigs/compiler Meeting notes

1 0

Report From SIG Data
by 刘存伟 28 Apr '20

28 Apr '20

Dear tic member, Please see the report from SIG MindSpore Data. # MindSpore Data Special Interest Group (SIG) This is the working repo for the Data special interest group (SIG). This repo contains all the artifacts, materials, meeting notes and proposals regarding dataset - data processing and mindrecord - data format in MindSpore. Feedbacks and contributions are welcome. Data Processing: You can understand it as a Dataset, which is mainly responsible for reading the user's data into a Dataset, then performing related data enhancement operations (such as: resize, onehot, rotate, shuffle, batch ...), and finally provide the Dataset to the training process. Data Format: It can conveniently normalize the user's training data to a unified format (MindRecord). The specific operation steps are as follows: The user can easily convert the training data into MindRecrod data by defining the training data schema and calling the Python API interface. The format is then read into a Dataset through MindDataset and provided to the training process. # SIG Leads Liu Cunwei (Huawei) # Logistics SIG leads will drive the meeting. Meeting annoucement will be posted on our gitee channel: https://gitee.com/mindspore/community/tree/master/sigs/data Feedbacks and topic requests are welcome by all. # Discussion Slack channel https://app.slack.com/client/TUKCY4QDR/C010RPN6QNP?cdn_fallback=2 Documents and artifacts: https://gitee.com/mindspore/community/tree/master/sigs/data # Meeting notes Thursday April 2, 2020 # Current Progress * Support multi-process of GeneratorDataset/PyFunc for high performance * Support variable batchsize * Support new Dataset operators, such as filter,skip,take,TextLineDataset

1 1