To Do DeadLine: 2020-08-20

Memories I still remember clearly that I was particularly interested in mathematics and theoretical physics when I graduated from junior high school, and studied advanced mathematics and basic physics on my own. After that, I was full of pleasure in exploring physics and mathematics, and every day after entering my first year of high school, I would go to the library after lunch and flip through science and technology magazines, then I participated in the National High School Students Applied Physics Knowledge Contest for the first time, and I remember the final question of the paper I used calculus to solve the problem, and when I got out of the examination room, I didn’t even think that I would get it. I was very happy at that time, and I am still very grateful to Mr. Yao, our class teacher and chemistry teacher in my first year of high school, for having seen the ranking list of science scores in his hands when I went to the school to register. He was the class representative in chemistry, and in every exam he would praise me for the clarity of the solution to a particular problem. Every Sunday, on the way to school, I would stand in the Xinhua bookstore for one to two hours to look up books and scientific materials related to relativity. ...

The last two days the company planned a hackathon competition, they simply signed up to participate, thinking that they are currently researching the topic of deep learning, on the Google search blockchain deep learning, and finally found this paper Energy-recycling Blockchain with Proof-of-Deep-Learning, and then developed a PoDL chain for storing deep models along the lines of the paper, which is presented below, followed by Mathematica and Python implementations. Program Code. ...

Wolfram 2020 08-20 China Student Ambassador Presentation The following topics are of interest to me, and I am currently starting my own business in Wolfram and Python, with Wolfram being the main language for prototyping and exploring initial ideas. After watching the following presentations, I will focus on biomedical and retinal image analysis as a deep learning research topic. 18:00-19:00 Jingxian Wang, Department of Physics, Peking University. Meta-programming Basis in Mathematica. ...

Today, in the course of work, I need to parse a JSON file to get a keyword, and I’m going to use Mathematica to practice. ``` `` { “apiVersion”: “v1”, “data”: { “voting-keystore.json”: “{ "version": 4, "pubkey": " a94300da6d73a11d8ba72c99dd385b7a9d5043c9ee83ffd4190d89b7677e68d384fbceb556d2a0652e81250ad4b8477d", ""uuid”: "09e5c322-bf6b-401a -b90d-1aef6d57996a", "crypto": {"checksum": {"message": "704e947d244bc8020ef6d6f0844c4aea08af18ecf345bb98ab999df6441377d7\ "function": "sha256", "params": {}}, "kdf": {"message": """, "function": "scrypt", "params": {"p": 1, "n": 32768, "salt": "b891518a06f1d3f67a6d485bba8627f5320497f98c08cb52e4391e50071ac97d", "dklen": 32, "r": 8}}, "cipher": {\" message": "bcd85ce03e82e645e135e13501f717d907c03a23ee1f8dc6bd045c53001719ddc7a5\", "function": "aes-128-ctr", "paragraphs": {"iv\ “: "7f05ffce95bcd0157e520c6b949b90b1"}}}, "path": """}”" }, “kind”: “ConfigMap”, “metadata”: { “creationTimestamp”: “2020-07-25T11:35:24Z”, “managedFields”: [ { “apiVersion”: “v1”, “fieldsType”: “FieldsV1”, “ffieldsV1”: { “f:data”: { “.” : {}, “f:voting-keystore.json”: {} } }, “manager”: “Swagger-Codegen”, “operation”: “Update”, “time”: “2020-07-25T11:35:24Z” } ], “name”: “lighthouse-validator-a94300da6d73”, “namespace”: “eth”, “resourceVersion”: “1200502”, “selfLink”: “/api/v1/namespaces/eth/configmaps/lighthouse-validator-a94300da6d73”, “uid”: “05b37640-be81-4944-9c9f-868431b6107f” } } ...

The animation program code is as follows: lung-3D.nb lung3d = AnatomyPlot3D[lung anatomical structure, PlotTheme->"XRay"] gif = {}; Do[ image = ImageResize[ Show[lung3d, ViewPoint -> {3 Cos[x], 3 Sin[x], 0}, ViewAngle -> 20 Degree], {256, 256}]; gif = Append[gif, image], {x, 0, 2 Pi, 0.1}] Export["lung-3D.gif", gif, "AnimationRepetitions" -> Infinity ] image = ImageResize[ Show[lung3d, ViewPoint -> {3 Cos[Pi/2], 3 Sin[Pi/2], 0}, ViewAngle -> 20 Degree], {512, 512}] Training datasets covid-chestxray-dataset ...

Horovod I have written an article about how to build a distributed training cluster on Raspberry Pie 4 using the distributed training system that comes with TF2.0. However, there is a drawback: we need to start the training program at each node, and the distributed training will only work after all the nodes are started. MPI is mainly used in the field of supercomputing. Building MPI cluster on Raspberry, firstly, it can be used to learn distributed computing on supercomputing, and secondly, it can observe the performance of TF> distributed training on ARMv8 in practice. ...

Video address of the online seminar The screenshots in the article are from the presentation of the seminar. Since I am not a biology major, and I am interested in viral protein research, I learned some basic knowledge about protein folding, and then I watched the seminar, the first 40 minutes could be roughly understood, but in the second half was confused, perhaps because of the lack of molecular biology knowledge, I share some of my own knowledge from the video (multiple images). ...

TFlite Composition tensorflow lite interpreter Derivation on deployed hardware, which can include cell phones, microcontrollers, embedded devices. Tensorflow Lite Converter Convert models to make them smaller and reason faster. TFlite Converter Tensorflow Converter can convert models to FlatBuffers format. FlatBuffers is a cross-platform serialization tool where structured data is stored in binary form and can be used in the following ways The performance on the microcontroller is memory saving. Python Keras Converter: ...