Jump to content

All Activity

This stream auto-updates

  1. Past hour
  2. @roytam1 is there any chance you might be able to fix or support the pixai thing i been trying to mention the past few weeks ?? or is it not possible ?? im sorry if i been annoying about it
  3. About UE4's shader compilation stutter issue in P3R, obviously R550 drivers have better performance than R535 drivers. Now I am using 552.61 and trying to find an ideal solution.
  4. Today
  5. Well I have a mini-PC AceMagician T8 Plus with 256 GB of hard drive space and 8 GB RAM so I have enough space for that machine to do anything. I was going to install Windows 7 or Tiny7 and then the VM environment and then OSX ontop of that with Think or Swim TOS as the final layer. But since Windows 7 some say has security breaches, could someone hack into my system through Windows 7 then into the VM environment and then into OSX and then into TOS? Like hacking though each layer? The really small 32 GB hard Drive space with 2 GB of RAM is on these tiny travel mini-PC sticks that I use only when I travel. They are easy to transport. So I was going to copy the layered system idea as I stated above into those Asus VivoSticks TS-10 as well and just run these systems infront of every day not remotely tapping in or using another computer to offload the op system hard drive space. Can I do that with this set up? And I prefer OSX over Windows. And I can't shutdown the Windows updates.
  6. Windows Update pulled down the Cumulative Update today on both Server 2025 and IOT Ent LTSC to 26100.863 - no SAB issues seen. Nice. Don't you love it when software just works?
  7. Tomorrow I do an intensive test of those 3 of my Neurons, who learn the XOR function without direct calculating XOR. Some thoughts about my idea: No noise at all, no vanishing gradient, always a clear signal, most easy calculation, just learning during the adaptation of each threshold Dietmar class MyNeuron { private double threshold1; private double threshold2; public MyNeuron(double threshold1, double threshold2) { this.threshold1 = threshold1; this.threshold2 = threshold2; } public int activate(double input) { if (input < threshold1) { return 0; } else if (input < threshold2) { return 1; } else { return 0; } } } class XORNeuralNetwork { private MyNeuron[] hiddenLayer; private MyNeuron outputNeuron; private double[][] hiddenWeights; private double[] outputWeights; private double learningRate; public XORNeuralNetwork() { hiddenLayer = new MyNeuron[] { new MyNeuron(0.1, 0.2), new MyNeuron(0.1, 0.2) }; outputNeuron = new MyNeuron(0.1, 0.2); hiddenWeights = new double[][] { {1.0, 1.0}, {1.0, 1.0} }; outputWeights = new double[] {1.0, 1.0}; learningRate = 0.1; // Lernrate einstellen } public int predict(int[] inputs) { int[] hiddenLayerOutputs = new int[hiddenLayer.length]; for (int i = 0; i < hiddenLayer.length; i++) { double weightedInput = 0; for (int j = 0; j < inputs.length; j++) { weightedInput += inputs[j] * hiddenWeights[i][j]; } hiddenLayerOutputs[i] = hiddenLayer[i].activate(weightedInput); } double outputWeightedInput = 0; for (int i = 0; i < hiddenLayerOutputs.length; i++) { outputWeightedInput += hiddenLayerOutputs[i] * outputWeights[i]; } return outputNeuron.activate(outputWeightedInput); } public void train(int[][] inputs, int[] expectedOutputs, int epochs) { for (int epoch = 0; epoch < epochs; epoch++) { for (int i = 0; i < inputs.length; i++) { int predicted = predict(inputs[i]); int error = expectedOutputs[i] - predicted; for (int j = 0; j < outputWeights.length; j++) { outputWeights[j] += learningRate * error * hiddenLayer[j].activate(inputs[i][j]); } for (int j = 0; j < hiddenLayer.length; j++) { for (int k = 0; k < hiddenWeights[j].length; k++) { hiddenWeights[j][k] += learningRate * error * inputs[i][k]; } } } // Ausgabe der Vorhersagen nach jeder Epoche System.out.println("Epoche " + epoch + ":"); for (int[] input : inputs) { int output = predict(input); System.out.println("Input: " + input[0] + ", " + input[1] + " -> Output: " + output); } System.out.println(); } } public static void main(String[] args) { XORNeuralNetwork xorNN = new XORNeuralNetwork(); int[][] inputs = { {0, 0}, {0, 1}, {1, 0}, {1, 1} }; int[] expectedOutputs = {0, 1, 1, 0}; xorNN.train(inputs, expectedOutputs, 100); // Anzahl der Epochen auf 100 reduziert für Übersichtlichkeit } }
  8. As others say around these here parts, "I don't see admin/moderator next to your name". Your post is just as "spammy" and as much a "pompous public announcement" as mine. Feel free to go tell the teacher. "Mom, D.Draker hit me back!" "What do you mean by hit you back?" "Well, I may have hit him first, but he didn't tell on me, but I'm going to tell on him." "Go to your room until daddy gets home."
  9. OK, it worked for that guy somehow, so should have worded a bit differently. I think I just got a black window back then. Still, better try this: https://storage.googleapis.com/chromium-browser-continuous/index.html?prefix=Win/291943/ Just remember to disable GPU blocklist on chrome://flags page. Though TBH I'm not sure if canvas accel works specifically, hard to test in reality such an outdated browser, but period correct WebGL stuff seems to work. None of that slow 100% CPU sucking SwiftShader nonsense. The hint comes from https://superuser.com/questions/958238/how-to-enable-webgl-on-chrome-on-xp. I like the last comment from the guy who asked: Interesting, but I have updated to Win10, so I can't test. I think Chrome no longer supports XP anyways.
  10. BTW, they suggest to fully disable sandbox, which is a terrible idea, too!
  11. Yeah, but it's not my idea, I actually wanted to further elaborate on what you wrote about Chrome 49! With a simple calculation, if they already knew it was blocked on XP almost 12 years ago, it brings us back to 2012 - early 2013. What was the Chrome version, then? 21? 20? 19? 18? So, it was blocked centuries ago! Overall, I agree with you, not a good idea at all.
  12. You would! You're just very lucky to be allowed to join the EU, you don't live under Broz Tito anymore. Otherwise, 100% of all interesting Western websites would be blocked, then. Now think of those that live in such territorial formations as China, Iran, Russia, Cuba, etc. All those entities are publicly known to zealously block huge segments of the internet! So, Microsoft is a kind and caring company that thinks of its customers and their safety!
  13. Good! In this case we (at least) know who will read through our communications! That VPN is supposedly aimed to use SHA 2, https://en.wikipedia.org/wiki/SHA-2 SHA-2 (Secure Hash Algorithm 2) is a set of cryptographic hash functions designed by the United States National Security Agency (NSA), NSA are the good guys, I don't mind them reading my emails, I don't plot against America.
  14. I just have a crazy cool idea of a new KI. Take a look at this Neuron. In some kind it is more intelligent than human Neurons in brain. It uses the "All or Nothing" idea for to make a Neuron to fire (first threshold1). BUT: It has a second threshold2, so that it can solve with only such 3 Neurons the XOR problem. And everything with only positiv(!) weights Dietmar PS: My version of Netbeans 16 allows only 1,3 as input for testing for a double number instead of 1.3 package myneuron; import java.util.Scanner; public class Myneuron { private double threshold1; private double threshold2; public Myneuron() { threshold1 = 1.0; threshold2 = 5.0; } public int activate(double input) { if (input < threshold1) { return 0; } else if (input < threshold2) { return 1; } else { return 0; } } public static void main(String[] args) { Scanner scanner = new Scanner(System.in); Myneuron neuron = new Myneuron(); System.out.println("Test meines speziellen Neurons:"); System.out.print("Gib eine Kommazahl als Input für das Neuron ein: "); double input = scanner.nextDouble(); int output = neuron.activate(input); System.out.println("Antwort des Neurons: " + output); scanner.close(); } }
  15. Implying WineD3D actually works, much less works well (tried the suggested hack years ago, didn't get anywhere). Bogging down Pentium 4 with translating D3D to OpenGL, it's a great idea!
  16. Technically, no need to spam up the thread if you aren't gonna request the name change, and block whomever you want, without such pompous public announcements, keep it to yourself. Thanks.
  17. "Chrome developers disabled Canvas HW Accel for XP (the artcile is about 12 years old!), it calls Direct3DCreate9Ex from d3d9.dll which is DirectX 9L available only on Vista+, but it is possible to wrap D3D9Ex on Windows XP via WineD3D Direct3D to OpenGL wrapper. (special thanks to WineHQ/Crossover) Then we need to disable blacklist and disable sandbox in launch options." https://stackoverflow.com/questions/15073593/does-chrome-support-gpu-hardware-acceleration-for-the-2d-canvas-under-windows-xp
  18. Chrome 49 requires D3D9Ex for HW accel, which is Vista+.
  19. Alright. I've found out that such values were set by setupdd.sys and values like Compute Server nor Storage Server not mentioned at all. I've decided to do some hex-editing and it did the trick. The problem is length of both strings had to be the same. If the string is shorter, replacing the rest of the string with 00 does the trick. However if the string is longer, it is impossible without decompilation.
  20. I have a local branch as staging to "merge" from NM27 repo, sans most of browser frontend changes, and lots of NSS/PSM changes. of course it builds on VC2010. way to NM27.9.6 is still far, and I think I need to skip interim NSS updates to reduce workload on porting them to VC2010 compatible syntax.
  21. @ClassicNick Posts cleaned. I see you have already asked about changing your username in the appropriate thread. Unfortunately, I cannot do this for you, only the supervisor @Tripredacus can now do it. If you don't hear from him in the name changing thread in the next few days, try sending him a PM. He should pick up the request though, so be patient!
  22. @Dave-H I think you can just report your own post to have things like that taken care of.
  23. Nothing will change. MV2 extensions that will not transition by choice of their developers to MV3,such as uBlock Origin will be abandoned after June 2025. Indeed, it is likely that in this transition year the development of uBlock Origin for Chromium-based browsers will slow down compared to the same version for Firefox-based browsers. On the other hand more and more users will install in their Chromium-based browsers uBlock Origin Lite,AdGuard extension MV3........etc.....and thus this state of abandonment will be accelerated by several parties Although I have the company policy already installed,personally if there is a slowdown in the development of uBlock Origin for Chromium based browsers,such that the functionality of the extension is reduced compared to the same version for Firefox I will switch to another MV3 adblocker without delay. Probably adGuard extension whose development I have been following for at least 6 months. I am just waiting for the problems that I have also opened to be solved. And of course that the current stable version be merged with the experimental MV3 version. If the future events hypothesized above come true.
  24. Technically, I think this is you flattering yourself. As I've pointed out before, I read ALL NEW posts at MSFN (with two exceptions - I don't use nor care for "Start Is Back" and "Vista Extended Kernel" and my custom style sheet prevents them from even appearing). I suppose I could customize my style sheet to prevent "various members" posts from appearing also, but why bother, lol.
  1. Load more activity
×
×
  • Create New...