<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Windows on myByways</title><link>https://myByways.com/tags/windows/</link><description>Recent content in Windows on myByways</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sun, 31 Aug 2025 00:00:00 +0000</lastBuildDate><atom:link href="https://myByways.com/tags/windows/index.xml" rel="self" type="application/rss+xml"/><item><title>Kokoro TTS and Abogen on Windows</title><link>https://myByways.com/post/kokoro-tts-and-abogen-on-windows/</link><pubDate>Sun, 31 Aug 2025 00:00:00 +0000</pubDate><guid>https://myByways.com/post/kokoro-tts-and-abogen-on-windows/</guid><description>&lt;p&gt;In my last post, I described trying out &lt;a href="https://huggingface.co/hexgrad/Kokoro-82M" rel="external"&gt;Kokoro text-to-speech (TTS) model&lt;/a&gt; via &lt;a href="https://myByways.com/post/running-kokoro-tts-via-macos-containerisation-framework"&gt;Kokoro-FastAPI web UI in a macOS (native) container&lt;/a&gt;. Here, I install &lt;a href="https://github.com/nazdridoy/kokoro-tts" rel="external"&gt;Kokoro-TTS&lt;/a&gt; and &lt;a href="https://github.com/denizsafak/abogen" rel="external"&gt;Abogen&lt;/a&gt; on Windows, to take advantage of my Nvidia GPU.&lt;/p&gt;</description></item><item><title>Faster Flux.1</title><link>https://myByways.com/post/faster-flux1/</link><pubDate>Wed, 14 Aug 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/faster-flux1/</guid><description>&lt;p&gt;More &lt;a href="https://huggingface.co/black-forest-labs/FLUX.1-dev" rel="external"&gt;Flux.1&lt;/a&gt;-based models! Go faster with FP8 or NF4! New LoRAs and ControlNets! There is quite a bit of interest with this model, as evidenced by the speed of community-led enhancements.&lt;/p&gt;</description></item><item><title>Wow! Flux.1 by Black Forrest Labs</title><link>https://myByways.com/post/wow-flux-by-black-forrest-labs/</link><pubDate>Sun, 04 Aug 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/wow-flux-by-black-forrest-labs/</guid><description>&lt;p&gt;While I have many posts about SDXL, I do not use Stable Diffusion 3 at all - license concerns aside, it is simply not good, and may never get better. But just a few days ago, a new, freely available, offline model that is better than SDXL was released by the team that presented Latent Diffusion and created Stable Diffusion, &lt;a href="https://blackforestlabs.ai/" rel="external"&gt;Flux.1 by Black Forrest Labs&lt;/a&gt;!&lt;/p&gt;</description></item><item><title>Fun with Llama 3 (8B) in Ollama</title><link>https://myByways.com/post/fun-with-llama-3-8b-in-ollama/</link><pubDate>Sun, 21 Apr 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/fun-with-llama-3-8b-in-ollama/</guid><description>&lt;p&gt;I saw a post on Reddit, entitled &lt;a href="https://www.reddit.com/r/LocalLLaMA/comments/1c90gl8/llama_3_rocks_with_taking_on_a_personality/" rel="external"&gt;“Llama 3 rocks with taking on a personality!”&lt;/a&gt;. A fun experiment! I thought to replicate it, blatantly copying the puzzle presented. &lt;a href="https://ai.meta.com/blog/meta-llama-3/" rel="external"&gt;Llama 3, released by Meta&lt;/a&gt; just before the weekend, is impressive.&lt;/p&gt;</description></item><item><title>Testing new PAG and Perp-Neg nodes in ComfyUI</title><link>https://myByways.com/post/testing-new-pag-and-perp-neg-nodes-in-comfyui/</link><pubDate>Wed, 17 Apr 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/testing-new-pag-and-perp-neg-nodes-in-comfyui/</guid><description>&lt;p&gt;I know it’s bad form to start off with a disclaimer: but the truth is, I do not know what I am doing. I am just testing out two new &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; nodes, &lt;kbd&gt;PerturbedAttentionGuidance&lt;/kbd&gt; and &lt;kbd&gt;PerpNegGuider&lt;/kbd&gt;.&lt;/p&gt;</description></item><item><title>Mistral System Message setup to improve Image Generation Prompts</title><link>https://myByways.com/post/mistral-system-message-setup-to-improve-image-generation-prompts/</link><pubDate>Sun, 31 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/mistral-system-message-setup-to-improve-image-generation-prompts/</guid><description>&lt;p&gt;&lt;a href="https://myByways.com/post/llm-prompt-generation-with-ollama-in-comfyui"&gt;In my last post&lt;/a&gt;, I used &lt;a href="https://github.com/if-ai/ComfyUI-IF_AI_tools" rel="external"&gt;ComfyUI-IF_AI_tools&lt;/a&gt; to integrate to the &lt;a href="https://ollama.com/brxce/stable-diffusion-prompt-generator" rel="external"&gt;brxce/stable-diffusion-prompt-generator model&lt;/a&gt; running in &lt;a href="https://ollama.com/" rel="external"&gt;Ollama&lt;/a&gt;. I wonder if I could use the base &lt;a href="https://mistral.ai/news/announcing-mistral-7b/" rel="external"&gt;Mistral 7B model&lt;/a&gt; to help improve my uncreative prompts instead&amp;hellip;&lt;/p&gt;</description></item><item><title>LLM Prompt Generation with Ollama in ComfyUI</title><link>https://myByways.com/post/llm-prompt-generation-with-ollama-in-comfyui/</link><pubDate>Thu, 28 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/llm-prompt-generation-with-ollama-in-comfyui/</guid><description>&lt;p&gt;In my last post, I described running &lt;a href="https://myByways.com/post/a-game-with-mistral-7b-using-ollama"&gt;Mistral, a Large Language Model, locally using Ollama&lt;/a&gt;. To accompany that piece, I created a prompt and manually used AI to generate an image. Today, I’ll wire up a &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; workflow to &lt;a href="https://ollama.com/" rel="external"&gt;Ollama&lt;/a&gt; to do this seamlessly, thanks to &lt;a href="https://github.com/if-ai/ComfyUI-IF_AI_tools" rel="external"&gt;ComfyUI-IF_AI_tools&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>A game with Mistral 7B using Ollama</title><link>https://myByways.com/post/a-game-with-mistral-7b-using-ollama/</link><pubDate>Wed, 20 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/a-game-with-mistral-7b-using-ollama/</guid><description>&lt;p&gt;I keep posting about Stable Diffusion, but I do experiment with &lt;strong&gt;Large Language Models&lt;/strong&gt; too! I do not have much to contribute in this regard, instead, here is the transcript of a game I played with the open source &lt;a href="https://mistral.ai/news/announcing-mistral-7b/" rel="external"&gt;Mistral 7B model&lt;/a&gt; via &lt;a href="https://ollama.com/" rel="external"&gt;Ollama&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Stable Video Diffusion</title><link>https://myByways.com/post/stable-video-diffusion/</link><pubDate>Sun, 10 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/stable-video-diffusion/</guid><description>&lt;p&gt;More and more AI generated images are shared as short video clips. So, here a quick test of &lt;a href="https://stability.ai/news/stable-video-diffusion-open-ai-video-model" rel="external"&gt;Stable Video Diffusion&lt;/a&gt; - which was released back in November last year. Don’t know why I didn’t post this when I posted about &lt;a href="https://myByways.com/post/generating-animations-videos-with-sdxl-lcm"&gt;AnimateDiff and the Hotshot Motion model&lt;/a&gt; around the same time.&lt;/p&gt;</description></item><item><title>TripoSR image-to-3D-object</title><link>https://myByways.com/post/triposr-image-to-3d-object/</link><pubDate>Sat, 09 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/triposr-image-to-3d-object/</guid><description>&lt;p&gt;Do you want to convert a 2D image into a 3D model &lt;em&gt;auto-magically&lt;/em&gt;? On 5 March 2024, &lt;strong&gt;Stability AI&lt;/strong&gt; and &lt;strong&gt;Tripo AI&lt;/strong&gt; released &lt;a href="https://stability.ai/news/triposr-3d-generation" rel="external"&gt;TripoSR: Fast 3D Object Generation from Single Images&lt;/a&gt; that does exactly that!&lt;/p&gt;</description></item><item><title>Differential Diffusion for in-painting</title><link>https://myByways.com/post/differential-diffusion-for-in-painting/</link><pubDate>Tue, 05 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/differential-diffusion-for-in-painting/</guid><description>&lt;p&gt;&lt;a href="https://differential-diffusion.github.io/" rel="external"&gt;Differential Diffusion&lt;/a&gt; is the newest method (framework) of in-painting without an in-painting model. Instead, all that is needed is a &lt;strong&gt;mask&lt;/strong&gt; (map) where the lighter the area, the greater the re-painting applied.&lt;/p&gt;</description></item><item><title>Generate transparent images with Layer Diffusion</title><link>https://myByways.com/post/generate-transparent-images-with-layer-diffusion/</link><pubDate>Sun, 03 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/generate-transparent-images-with-layer-diffusion/</guid><description>&lt;p&gt;Ever wished you could generate &lt;a href="https://stability.ai/stable-image" rel="external"&gt;Stable Diffusion XL&lt;/a&gt; images with transparent backgrounds? Well, your wish has been answered by the smart people behind the &lt;a href="https://arxiv.org/html/2402.17113v1" rel="external"&gt;Transparent Image Layer Diffusion using Latent Transparency paper&lt;/a&gt;. They have made their code and models available, and what do you know, &lt;a href="https://github.com/huchenlei/ComfyUI-layerdiffusion" rel="external"&gt;Chenlei Hu&lt;/a&gt; has ported it to &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt;!&lt;/p&gt;</description></item><item><title>SDXL-based 4-step models compared</title><link>https://myByways.com/post/sdxl-based-4-step-models-compared/</link><pubDate>Sat, 02 Mar 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/sdxl-based-4-step-models-compared/</guid><description>&lt;p&gt;With the advent of techniques like Adversarial Diffusion Distillation and Latent Consistency models, A.I. image synthesis based on &lt;a href="https://stability.ai/stable-image" rel="external"&gt;Stable Diffusion XL&lt;/a&gt; has been getting faster and faster. Here is just quick comparison of a few models at 4-steps, some of which are fine-tuned and trained for realism.&lt;/p&gt;</description></item><item><title>Consistent portraits revisisted: InstantID</title><link>https://myByways.com/post/consistent-portraits-revisisted-instantid/</link><pubDate>Sun, 25 Feb 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/consistent-portraits-revisisted-instantid/</guid><description>&lt;p&gt;Not long ago, in a attempt to obtain &lt;a href="https://myByways.com/post/consistent-portraits-using-ip-adapters-for-sdxl"&gt;Consistent portraits using IP-Adapters for SDXL&lt;/a&gt;, I &lt;a href="https://myByways.com/post/comparing-face-ip-adapters-for-sdxl"&gt;shared a comparison&lt;/a&gt; between &lt;a href="https://huggingface.co/h94/IP-Adapter" rel="external"&gt;IP-Adapter-Plus-Face&lt;/a&gt; and &lt;a href="https://huggingface.co/h94/IP-Adapter-FaceID" rel="external"&gt;IP-Adapter-FaceID&lt;/a&gt;. Today I’ll look at &lt;a href="https://huggingface.co/InstantX/InstantID" rel="external"&gt;InstantID&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>New Stable Cascase Checkpoints for ComfyUI</title><link>https://myByways.com/post/new-stable-cascase-checkpoints-for-comfyui/</link><pubDate>Sat, 24 Feb 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/new-stable-cascase-checkpoints-for-comfyui/</guid><description>&lt;p&gt;An update to my &lt;a href="https://myByways.com/post/stable-cascade-with-comfyui"&gt;previous post on Stable Cascade with ComfyUI&lt;/a&gt; - instead of requiring four separate model files, we now only need &lt;a href="https://huggingface.co/stabilityai/stable-cascade/tree/main/comfyui_checkpoints" rel="external"&gt;two checkpoints&lt;/a&gt;, and the &lt;a href="https://comfyanonymous.github.io/ComfyUI_examples/stable_cascade/" rel="external"&gt;ComfyUI workflow is now very straightfoward&lt;/a&gt;!&lt;/p&gt;</description></item><item><title>Stable Cascade with ComfyUI</title><link>https://myByways.com/post/stable-cascade-with-comfyui/</link><pubDate>Sun, 18 Feb 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/stable-cascade-with-comfyui/</guid><description>&lt;p&gt;On 12 Feb 2024, &lt;a href="https://stability.ai/" rel="external"&gt;Stability.ai&lt;/a&gt; released &lt;a href="https://stability.ai/news/introducing-stable-cascade" rel="external"&gt;Stable Cascade&lt;/a&gt; “research preview” (non-commercial license), and over the weekend, &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; was updated to support this new model! Time to give it a go!&lt;/p&gt;</description></item><item><title>Comparing face IP-Adapters for SDXL</title><link>https://myByways.com/post/comparing-face-ip-adapters-for-sdxl/</link><pubDate>Sun, 14 Jan 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/comparing-face-ip-adapters-for-sdxl/</guid><description>&lt;p&gt;As a follow up to my last post regarding &lt;a href="https://myByways.com/post/consistent-portraits-using-ip-adapters-for-sdxl"&gt;Consistent portraits using IP-Adapters for SDXL&lt;/a&gt;, this is a short comparison of the two face IP-Adapters for SDXL by &lt;a href="https://huggingface.co/h94" rel="external"&gt;h94 / xiaohu&lt;/a&gt;: namely, &lt;a href="https://huggingface.co/h94/IP-Adapter" rel="external"&gt;ip-adapter-plus-face_sdxl_vit-h.bin&lt;/a&gt; and &lt;a href="https://huggingface.co/h94/IP-Adapter-FaceID" rel="external"&gt;ip-adapter-faceid_sdxl.bin&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Consistent portraits using IP-Adapters for SDXL</title><link>https://myByways.com/post/consistent-portraits-using-ip-adapters-for-sdxl/</link><pubDate>Sun, 07 Jan 2024 00:00:00 +0000</pubDate><guid>https://myByways.com/post/consistent-portraits-using-ip-adapters-for-sdxl/</guid><description>&lt;p&gt;Getting consistent character portraits generated by SDXL has been a challenge&amp;hellip; until now! &lt;a href="https://github.com/cubiq/ComfyUI_IPAdapter_plus" rel="external"&gt;ComfyUI IPAdapter Plus&lt;/a&gt; (dated 30 Dec 2023) now supports both &lt;a href="https://huggingface.co/h94/IP-Adapter" rel="external"&gt;IP-Adapter&lt;/a&gt; and &lt;a href="https://huggingface.co/h94/IP-Adapter-FaceID" rel="external"&gt;IP-Adapter-FaceID&lt;/a&gt; (released 4 Jan 2024)!&lt;/p&gt;</description></item><item><title>QR Code Monster for SDXL</title><link>https://myByways.com/post/qr-code-monster-for-sdxl/</link><pubDate>Tue, 14 Nov 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/qr-code-monster-for-sdxl/</guid><description>&lt;p&gt;Time to try another &lt;strong&gt;ControlNet&lt;/strong&gt; for Stable Diffusion XL - &lt;a href="https://huggingface.co/monster-labs/control_v1p_sdxl_qrcode_monster" rel="external"&gt;QR Code Monster v1&lt;/a&gt; in &lt;strong&gt;ComfyUI&lt;/strong&gt;. This ControlNet can influence SDXL such that the generated image “hides” a scan-able QR code, which at first glance, looks like a photo!&lt;/p&gt;</description></item><item><title>Generating animations / videos with SDXL LCM</title><link>https://myByways.com/post/generating-animations-videos-with-sdxl-lcm/</link><pubDate>Mon, 13 Nov 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/generating-animations-videos-with-sdxl-lcm/</guid><description>&lt;p&gt;I never tried generating video clips or animations with SDXL before, simply because my GPU is not powerful enough. But after testing out the &lt;a href="https://myByways.com/post/crazy-fast-image-generation-with-lcm-lora-for-sdxl"&gt;LCM LoRA for SDXL yesterday&lt;/a&gt;, I thought I’d try the &lt;a href="https://huggingface.co/blog/lcm_lora" rel="external"&gt;SDXL LCM LoRA&lt;/a&gt; with &lt;a href="https://github.com/hotshotco/Hotshot-XL" rel="external"&gt;Hotshot-XL&lt;/a&gt;, which is something akin to &lt;a href="https://animatediff.github.io/" rel="external"&gt;AnimateDiff&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Crazy fast image generation with LCM LoRA for SDXL</title><link>https://myByways.com/post/crazy-fast-image-generation-with-lcm-lora-for-sdxl/</link><pubDate>Sun, 12 Nov 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/crazy-fast-image-generation-with-lcm-lora-for-sdxl/</guid><description>&lt;p&gt;Stable Diffusion keeps improving at an astounding pace! This time, it’s the idea of &lt;em&gt;distilling&lt;/em&gt; a model into a &lt;strong&gt;Latent Consistency Model (LCM)&lt;/strong&gt; for very, very fast image generation with a quality trade-off. On 24 Oct 2023, the distilled &lt;a href="https://huggingface.co/segmind/SSD-1B" rel="external"&gt;Segmind Stable Diffusion 1B (SSD-1B) model&lt;/a&gt; was released, followed by a better implementation in the form of &lt;a href="https://huggingface.co/blog/lcm_lora" rel="external"&gt;Latent Consistency LoRAs for SDXL and SDD-1B&lt;/a&gt; released on 9 Nov 2023.&lt;/p&gt;</description></item><item><title>Using the SDXL-Inpainting 0.1 Model</title><link>https://myByways.com/post/using-the-sdxl-inpainting-01-model/</link><pubDate>Sun, 03 Sep 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/using-the-sdxl-inpainting-01-model/</guid><description>&lt;p&gt;Stability AI just released an new &lt;a href="https://huggingface.co/diffusers/stable-diffusion-xl-1.0-inpainting-0.1" rel="external"&gt;SD-XL Inpainting 0.1 model&lt;/a&gt;. Here is how to use it with &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>SDXL Revision workflow in ComfyUI</title><link>https://myByways.com/post/sdxl-revision-workflow-in-comfyui/</link><pubDate>Sun, 20 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/sdxl-revision-workflow-in-comfyui/</guid><description>&lt;p&gt;Yesterday, I &lt;a href="https://myByways.com/post/stability-ai-control-loras"&gt;tried out Stability AI’s four Control-LoRAs&lt;/a&gt; but mentioned that I did not understand the output of the &lt;strong&gt;Revision&lt;/strong&gt; “image-mixing” workflow. I’ve since done a bit more experimentation&amp;hellip;&lt;/p&gt;</description></item><item><title>Stability AI Control LoRAs</title><link>https://myByways.com/post/stability-ai-control-loras/</link><pubDate>Sat, 19 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/stability-ai-control-loras/</guid><description>&lt;p&gt;Less than a week after my post testing &lt;a href="https://myByways.com/post/sdxl-1-0-with-sdxl-controlnet-canny"&gt;diffusers/controlnet-canny-sdxl-1.0&lt;/a&gt;, along comes Stability AI’s own &lt;strong&gt;ControlNets&lt;/strong&gt;, which they call &lt;a href="https://huggingface.co/stabilityai/control-lora" rel="external"&gt;Control-LoRAs&lt;/a&gt;! Not one but 4 of them - &lt;strong&gt;Canny&lt;/strong&gt;, &lt;strong&gt;Depth&lt;/strong&gt;, &lt;strong&gt;Recolor&lt;/strong&gt; and &lt;strong&gt;Sketch&lt;/strong&gt; models!&lt;/p&gt;</description></item><item><title>SDXL 1.0 with SDXL-ControlNet: OpenPose (v2)</title><link>https://myByways.com/post/sdxl-1-0-with-sdxl-controlnet-openpose-v2/</link><pubDate>Fri, 18 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/sdxl-1-0-with-sdxl-controlnet-openpose-v2/</guid><description>&lt;p&gt;As &lt;a href="https://myByways.com/post/fooocus-ksampler-custom-node-for-comfyui-sdxl"&gt;promised in my last post&lt;/a&gt;, today I am testing out &lt;strong&gt;Thibaud Zamora&lt;/strong&gt;’s &lt;a href="https://huggingface.co/thibaud/controlnet-openpose-sdxl-1.0" rel="external"&gt;SDXL-controlnet: OpenPose (v2) model&lt;/a&gt; using &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt;. I keep saying I’ll keep my posts short but never do&amp;hellip;&lt;/p&gt;</description></item><item><title>Fooocus KSampler Custom Node for ComfyUI SDXL</title><link>https://myByways.com/post/fooocus-ksampler-custom-node-for-comfyui-sdxl/</link><pubDate>Thu, 17 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/fooocus-ksampler-custom-node-for-comfyui-sdxl/</guid><description>&lt;p&gt;Recently I tried &lt;a href="https://github.com/lllyasviel/Fooocus" rel="external"&gt;Fooocus by Lyumin Zhang (Illyasviel)&lt;/a&gt; which fulfills its promise to allow one to “Focus on prompting and generating” - it is certainly easy to use! But shortly after its release, someone has “ported” the code to &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; as a &lt;strong&gt;Custom Node&lt;/strong&gt;! So of course it’s time to test it out&amp;hellip;&lt;/p&gt;</description></item><item><title>SDXL 1.0 with SDXL-ControlNet: Canny</title><link>https://myByways.com/post/sdxl-1-0-with-sdxl-controlnet-canny/</link><pubDate>Sun, 13 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/sdxl-1-0-with-sdxl-controlnet-canny/</guid><description>&lt;p&gt;The Stability AI documentation now has a pipeline supporting &lt;a href="https://huggingface.co/docs/diffusers/main/en/api/pipelines/controlnet_sdxl" rel="external"&gt;ControlNets with Stable Diffusion XL&lt;/a&gt;! Time to try it out with &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; for Windows.&lt;/p&gt;</description></item><item><title>Scale and Composite Latents with SDXL</title><link>https://myByways.com/post/scale-and-composite-latents-with-sdxl/</link><pubDate>Sat, 05 Aug 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/scale-and-composite-latents-with-sdxl/</guid><description>&lt;p&gt;In this post, I experiment with &lt;strong&gt;latent scaling&lt;/strong&gt; and &lt;strong&gt;latent compositing&lt;/strong&gt; with &lt;a href="https://stability.ai/blog/stable-diffusion-sdxl-1-announcement" rel="external"&gt;SDXL 1.0&lt;/a&gt; using &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt;. That is to say, increasing / decreasing the size of the image, and combining multiple images into one à la green screen (chroma key) compositing.&lt;/p&gt;</description></item><item><title>CLIPSeg with SDXL in ComfyUI</title><link>https://myByways.com/post/clipseg-with-sdxl-in-comfyui/</link><pubDate>Mon, 31 Jul 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/clipseg-with-sdxl-in-comfyui/</guid><description>&lt;p&gt;Onward with &lt;a href="https://stability.ai/blog/stable-diffusion-sdxl-1-announcement" rel="external"&gt;SDXL&lt;/a&gt; and &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt;! Sometimes I want to tweak generated images by replacing selected parts that don’t look good while retaining the rest of the image that does look good. Rather than manually creating a mask, I’d like to leverage &lt;a href="https://github.com/timojl/clipseg" rel="external"&gt;CLIPSeg&lt;/a&gt; to generate a masks from a text prompt.&lt;/p&gt;</description></item><item><title>SDXL with Offset Example LoRA in ComfyUI for Windows</title><link>https://myByways.com/post/sdxl-with-offset-example-lora-in-comfyui-for-windows/</link><pubDate>Sun, 30 Jul 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/sdxl-with-offset-example-lora-in-comfyui-for-windows/</guid><description>&lt;p&gt;Yesterday I mentioned in passing that my &lt;a href="https://myByways.com/post/stable-diffusion-sdxl-1-0-with-comfyui"&gt;Nvidia RTX 2060 with 12GB could not run both SDXL 1.0 Base and Refiner models&lt;/a&gt; in a single &lt;a href="https://github.com/comfyanonymous/ComfyUI" rel="external"&gt;ComfyUI&lt;/a&gt; workflow. Today, I show you my workaround and also experiment with adding the &lt;a href="https://stability.ai/blog/stable-diffusion-sdxl-1-announcement" rel="external"&gt;SDXL 1.0&lt;/a&gt; &lt;strong&gt;Official Offset Example LoRA&lt;/strong&gt; to the workflow.&lt;/p&gt;</description></item><item><title>Configure a mouse for macOS trackpad-like gestures</title><link>https://myByways.com/post/configure-a-mouse-for-macos-trackpad-like-gestures/</link><pubDate>Sun, 16 Jul 2023 00:00:00 +0000</pubDate><guid>https://myByways.com/post/configure-a-mouse-for-macos-trackpad-like-gestures/</guid><description>&lt;p&gt;A short post about setting up a mouse on macOS to mimic trackpad gestures.&lt;/p&gt;</description></item><item><title>Browser support for HTTP/3 QUIC</title><link>https://myByways.com/post/browser-support-for-http3-quic/</link><pubDate>Sun, 31 Jul 2022 00:00:00 +0000</pubDate><guid>https://myByways.com/post/browser-support-for-http3-quic/</guid><description>&lt;p&gt;It’s been over a year (or two) since &lt;a href="https://en.wikipedia.org/wiki/HTTP/3" rel="external"&gt;three major browsers enabled HTTP/3 using the QUIC protocol over UDP&lt;/a&gt;. Chrome and Edge enabled HTTP/3 by default in April 2020 and Firefox followed “shortly after” in April 2021. I am not sure - does &lt;strong&gt;Safari in macOS Monterey 12.5&lt;/strong&gt;, released July 2022, enable HTTP/3 by default?&lt;/p&gt;</description></item><item><title>Enabling AMD Ryzen 5600X fTPM for Windows 10</title><link>https://myByways.com/post/enabling-amd-ryzen-5600x-ftpm-for-windows-10/</link><pubDate>Fri, 25 Jun 2021 00:00:00 +0000</pubDate><guid>https://myByways.com/post/enabling-amd-ryzen-5600x-ftpm-for-windows-10/</guid><description>&lt;p&gt;You might have read that Microsoft’s upcoming &lt;a href="https://docs.microsoft.com/en-us/windows/compatibility/windows-11/" rel="external"&gt;Windows 11 requires a Trusted Platform Module (TPM)&lt;/a&gt;. More recent AMD CPUs support &lt;strong&gt;Firmware TPM&lt;/strong&gt; (fTPM) which implements TPM in software on the CPU instead of requiring a dedicated (discrete) chip. This is how I enabled fTPM.&lt;/p&gt;</description></item><item><title>Tweaks and Hacks to Configure Windows 10</title><link>https://myByways.com/post/tweaks-and-hacks-to-configure-windows-10/</link><pubDate>Sat, 13 Mar 2021 00:00:00 +0000</pubDate><guid>https://myByways.com/post/tweaks-and-hacks-to-configure-windows-10/</guid><description>&lt;p&gt;I’ve been meaning to post this from the time I built my Windows desktop. But I keep adding to the list of things I need to configure to get &lt;strong&gt;Windows 10&lt;/strong&gt; into a ”usable” state for me. Unlike with macOS - for which the default installation is well configured, secure, and pretty much good-to-go - the installation and setup of Windows is extremely painful and a terrible experience. I have to spend hours trying different methods to remove junk, disable telemetry, and secure Windows (and still fail)&amp;hellip;&lt;/p&gt;</description></item><item><title>Command to turn off monitor in Windows 10</title><link>https://myByways.com/post/command-to-turn-off-monitor-in-windows-10/</link><pubDate>Fri, 12 Mar 2021 00:00:00 +0000</pubDate><guid>https://myByways.com/post/command-to-turn-off-monitor-in-windows-10/</guid><description>&lt;p&gt;I wanted to turn off my monitor in Windows 10, while listening to music, rather than wait for the &lt;strong&gt;Power plan&lt;/strong&gt; “Turn off the display” timer to kick in. Hmmm, seems Windows has never improved, because I recall having this requirement back-in-the-day - haha, that was during my first project ever, when I was involved in some Win32 development&amp;hellip;&lt;/p&gt;</description></item><item><title>Configuring Windows Subsystem for Linux (WSL2)</title><link>https://myByways.com/post/configuring-windows-subsystem-for-linux-wsl2/</link><pubDate>Sat, 13 Feb 2021 00:00:00 +0000</pubDate><guid>https://myByways.com/post/configuring-windows-subsystem-for-linux-wsl2/</guid><description>&lt;p&gt;Here is how I configure &lt;strong&gt;Windows Subsystem for Linux&lt;/strong&gt; (WSL). I do things a bit strangely: by changing the mounts, by running a GUI Desktop, and by running Docker in WSL2 and not on the Host&amp;hellip;&lt;/p&gt;</description></item><item><title>Custom icons in Windows Start Menu Medium Tiles</title><link>https://myByways.com/post/custom-icons-in-windows-start-menu-medium-tiles/</link><pubDate>Thu, 21 Jan 2021 00:00:00 +0000</pubDate><guid>https://myByways.com/post/custom-icons-in-windows-start-menu-medium-tiles/</guid><description>&lt;p&gt;In the Windows 10 &lt;strong&gt;Start Menu&lt;/strong&gt;, I used &lt;strong&gt;Pin To Start&lt;/strong&gt; to a medium sized tile for Firefox - and the icon is nice and large - but is not the case for other shortcuts. I was wondered why&amp;hellip; and how to make all the icons have a more consistent look.&lt;/p&gt;</description></item><item><title>RTSS layout for Cyberpunk 2077</title><link>https://myByways.com/post/rtss-layout-for-cyberpunk-2077/</link><pubDate>Sun, 13 Dec 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/rtss-layout-for-cyberpunk-2077/</guid><description>&lt;p&gt;I recently posted about configuring RTSS to overlay &lt;a href="https://myByways.com/pages/blog/cpu-and-gpu-monitoring-overlay-for-gaming"&gt;CPU and GPU information&lt;/a&gt; on games in Windows 10. Here is my setup for Cyberpunk 2077, so that I can &lt;a href="https://myByways.com/post/cyberpunk-2077-settings-for-rtx-2060-1080p"&gt;keep an eye on my framerates (when testing the plethora of graphics settings)&lt;/a&gt;, CPU / GPU / RAM utilization and temperatures.&lt;/p&gt;</description></item><item><title>Cyberpunk 2077 settings for RTX 2060 @ 1080p</title><link>https://myByways.com/post/cyberpunk-2077-settings-for-rtx-2060-1080p/</link><pubDate>Sat, 12 Dec 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/cyberpunk-2077-settings-for-rtx-2060-1080p/</guid><description>&lt;p&gt;I though I’d quickly post my &lt;a href="https://www.cyberpunk.net/" rel="external"&gt;Cyberpunk 20774&lt;/a&gt; settings running on a Nvidia RTX 2060 at 1080p (920×1080). The default Cyberpunk &lt;strong&gt;Ray Tracing&lt;/strong&gt; Medium preset results in low frame rates and rather blur graphics, although all effects are enabled.&lt;/p&gt;
&lt;p&gt;&lt;em&gt;Originally written with v1.04, but updated for version 1.05+&lt;/em&gt;&lt;/p&gt;</description></item><item><title>1stPlayer DK keyboard and mouse review</title><link>https://myByways.com/post/1stplayer-dk-keyboard-and-mouse-review/</link><pubDate>Sun, 06 Dec 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/1stplayer-dk-keyboard-and-mouse-review/</guid><description>&lt;p&gt;I bought a cheap-ish mechanical keyboard from &lt;a href="https://www.1stplayer.com/" rel="external"&gt;1stPlayer&lt;/a&gt;, the &lt;a href="https://www.1stplayer.com/productdetails/id/38.html" rel="external"&gt;DK 5.0 Lite 87&lt;/a&gt; along with the &lt;a href="https://www.1stplayer.com/productdetails/id/34.html" rel="external"&gt;DK 3.0 mouse&lt;/a&gt;. &lt;a href="https://myByways.com/post/my-first-pc-build-with-an-amd-ryzen-5-5600x"&gt;I called this purchase “a mistake”&lt;/a&gt; as I am not quite satisfied with the physical build and the very poorly designed software. So, I figured I might as well do a short review that may help prospective buyers.&lt;/p&gt;</description></item><item><title>Fix mechanical keyboard keys 1678923450</title><link>https://myByways.com/post/fix-mechanical-keyboard-keys-1678923450/</link><pubDate>Sun, 06 Dec 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/fix-mechanical-keyboard-keys-1678923450/</guid><description>&lt;p&gt;I suddenly experienced a strange issue today - the row of number keys on my &lt;a href="https://www.1stplayer.com/productdetails/id/38.html" rel="external"&gt;1stPlayer DK 5.0 Lite&lt;/a&gt; mechanical keyboard, i.e. the keys from &lt;code&gt;2&lt;/code&gt; to &lt;code&gt;9&lt;/code&gt;, would output the wrong number! Specifically, hitting all numeric keys in order would output &lt;code&gt;1678923450&lt;/code&gt; instead of &lt;code&gt;1234567890&lt;/code&gt;. This happened while gaming and I have no idea what caused it.&lt;/p&gt;</description></item><item><title>Creating Windows and macOS VMs</title><link>https://myByways.com/post/creating-windows-and-macos-vms/</link><pubDate>Sun, 29 Nov 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/creating-windows-and-macos-vms/</guid><description>&lt;p&gt;Just a bunch of un-related findings that may help when creating Windows and macOS VMs using &lt;a href="https://www.virtualbox.org/" rel="external"&gt;VirtualBox&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>CPU and GPU Monitoring Overlay for Gaming</title><link>https://myByways.com/post/cpu-and-gpu-monitoring-overlay-for-gaming/</link><pubDate>Thu, 26 Nov 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/cpu-and-gpu-monitoring-overlay-for-gaming/</guid><description>&lt;p&gt;On my &lt;a href="https://myByways.com/post/my-first-pc-build-with-an-amd-ryzen-5-5600x"&gt;new Windows 10 desktop&lt;/a&gt; I wanted an on-screen display (OSD) of system performance overlay while gaming - similar those benchmarking videos one finds on YouTube. While I don’t overclock my CPU or GPU, I still do want to know the utilization and temperatures, as these give me a hint as to whether the system is thermal throttled or bottlenecked.&lt;/p&gt;</description></item><item><title>VirtualBox very slow on Windows 10</title><link>https://myByways.com/post/virtualbox-very-slow-on-windows-10/</link><pubDate>Mon, 23 Nov 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/virtualbox-very-slow-on-windows-10/</guid><description>&lt;p&gt;I found that &lt;strong&gt;Oracle VM VirtualBox 6.1&lt;/strong&gt; was running very slowly on my &lt;a href="https://myByways.com/post/my-first-pc-build-with-an-amd-ryzen-5-5600x"&gt;brand new Windows 10 PC&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;Initially I thought it was to do with VirtualBox, e.g. insufficient cores or memory, the display controller (VBoxSVGA), insufficient video memory, etc. And then I thought it might be an AMD thing, so had a look at BIOS, e.g. nested AMD-V, etc.&lt;/p&gt;</description></item><item><title>My first PC build with an AMD Ryzen 5 5600X</title><link>https://myByways.com/post/my-first-pc-build-with-an-amd-ryzen-5-5600x/</link><pubDate>Sat, 14 Nov 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/my-first-pc-build-with-an-amd-ryzen-5-5600x/</guid><description>&lt;p&gt;I just got myself a new desktop PC! I am not much of a Windows user, but am looking forward to Cyberpunk 2077! I have never built a PC before, so I did the research and learnt a lot. Here is what I bought and why.&lt;/p&gt;</description></item><item><title>How to download videos from YouTube</title><link>https://myByways.com/post/how-to-download-videos-from-youtube/</link><pubDate>Sat, 30 May 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/how-to-download-videos-from-youtube/</guid><description>&lt;p&gt;Wondering how to download a video from &lt;strong&gt;YouTube&lt;/strong&gt;, with audio and subtitles intact? Here are a couple of methods I’m documenting in case I ever need to use them.&lt;/p&gt;</description></item><item><title>Fixing Windows 10 100% disk by stopping SysMain</title><link>https://myByways.com/post/fixing-windows-10-100-disk-by-stopping-sysmain/</link><pubDate>Tue, 21 Jan 2020 00:00:00 +0000</pubDate><guid>https://myByways.com/post/fixing-windows-10-100-disk-by-stopping-sysmain/</guid><description>&lt;p&gt;I just spent hours trying to figure out and fix a &lt;strong&gt;Windows 10&lt;/strong&gt; machine that was taking 100% disk, 100% RAM and about 50% CPU with absolutely nothing running. Of course, I initially blamed the user&amp;hellip; then, a misbehaving (hidden, background) app&amp;hellip; then, malware / virus&amp;hellip; the, a runaway Windows Defender&amp;hellip; then, a hard disk / hardware failure&amp;hellip; but ultimately discovering the culprit is - &lt;em&gt;drum roll&lt;/em&gt; - Microsoft Windows itself! As usual!&lt;/p&gt;</description></item><item><title>PowerShell Pipeline to Rename Files</title><link>https://myByways.com/post/powershell-pipeline-to-rename-files/</link><pubDate>Mon, 30 Dec 2019 00:00:00 +0000</pubDate><guid>https://myByways.com/post/powershell-pipeline-to-rename-files/</guid><description>&lt;p&gt;For me, Microsoft &lt;strong&gt;PowerShell&lt;/strong&gt; is hard! I can’t wrap my mind around &lt;a href="https://docs.microsoft.com/en-us/powershell/scripting/learn/understanding-the-powershell-pipeline?view=powershell-6" rel="external"&gt;data pipelines&lt;/a&gt; (&lt;code&gt;|&lt;/code&gt;). Instead, I keep reverting to if/then, loops and other traditional programming paradigms, similar to what I’d do in UNIX shell!&lt;/p&gt;</description></item><item><title>About the Windows Sandbox</title><link>https://myByways.com/post/about-the-windows-sandbox/</link><pubDate>Sun, 01 Sep 2019 00:00:00 +0000</pubDate><guid>https://myByways.com/post/about-the-windows-sandbox/</guid><description>&lt;p&gt;Starting with &lt;strong&gt;Windows 10 May 2019 update&lt;/strong&gt;, there is a new &lt;strong&gt;Windows Sandbox&lt;/strong&gt; feature, which, despite its name, is based on &lt;strong&gt;Windows Containers&lt;/strong&gt;.&lt;/p&gt;</description></item><item><title>Windows Permission Problems Copying From NTFS HDD</title><link>https://myByways.com/post/windows-permission-problems-copying-from-ntfs-hdd/</link><pubDate>Wed, 26 Dec 2018 00:00:00 +0000</pubDate><guid>https://myByways.com/post/windows-permission-problems-copying-from-ntfs-hdd/</guid><description>&lt;p&gt;Today I tried to copy files and folders off a USB hard disk (NTFS formatted) on my Windows 7 desktop. I encountered multiple files with permission problems. The manual method to change permissions is very tedious, so I created a script (DOS batch file) to do the same.&lt;/p&gt;</description></item><item><title>Batch file to list deleted files</title><link>https://myByways.com/post/batch-file-to-list-deleted-files/</link><pubDate>Thu, 12 Jul 2018 00:00:00 +0000</pubDate><guid>https://myByways.com/post/batch-file-to-list-deleted-files/</guid><description>&lt;p&gt;Well, it turns out that after creating a &lt;a href="https://myByways.com/post/powershell-to-list-deleted-files"&gt;PowerShell script to list deleted files&lt;/a&gt; (part 1) and another &lt;a href="https://myByways.com/post/powershell-to-monitor-for-deleted-renamed-or-moved-files"&gt;PowerShell script to monitor for deleted, renamed or moved files&lt;/a&gt; (part 2), I&amp;rsquo;ve decided to revert to a simple, batch file instead because running a &lt;strong&gt;batch file&lt;/strong&gt; is so much easier than &lt;strong&gt;PowerShell&lt;/strong&gt;!&lt;/p&gt;</description></item><item><title>PowerShell to monitor for deleted, renamed or moved files (part 2)</title><link>https://myByways.com/post/powershell-to-monitor-for-deleted-renamed-or-moved-files/</link><pubDate>Sun, 17 Jun 2018 00:00:00 +0000</pubDate><guid>https://myByways.com/post/powershell-to-monitor-for-deleted-renamed-or-moved-files/</guid><description>&lt;p&gt;This is part 2, of my attempt to &amp;ldquo;sync&amp;rdquo; photos I deleted on my desktop to my SD card (you can &lt;a href="https://myByways.com/post/powershell-to-list-deleted-files"&gt;read part 1 first&lt;/a&gt;). In this post, I try to use &lt;strong&gt;PowerShell&lt;/strong&gt; with .NET framework to (try) monitor for file system changes, and output to a batch file that &amp;ldquo;replicates&amp;rdquo; the &lt;code&gt;ren&lt;/code&gt; (rename) and &lt;code&gt;del&lt;/code&gt; (delete) to files and folders.&lt;/p&gt;</description></item><item><title>PowerShell to list deleted files (part 1)</title><link>https://myByways.com/post/powershell-to-list-deleted-files/</link><pubDate>Sat, 16 Jun 2018 00:00:00 +0000</pubDate><guid>https://myByways.com/post/powershell-to-list-deleted-files/</guid><description>&lt;p&gt;I have an unusual workflow when copying photos from my camera&amp;rsquo;s SD card to my Windows desktop. For any photos I my desktop, I want to also delete on the SD card. In this post (part 1 of 2), I describe my &lt;strong&gt;PowerShell&lt;/strong&gt; script to &amp;ldquo;repeat&amp;rdquo; what I deleted on the desktop on the SD card. In &lt;a href="https://myByways.com/post/powershell-to-monitor-for-deleted-renamed-or-moved-files"&gt;my next post, part 2&lt;/a&gt;, I expand the &lt;strong&gt;PowerShell&lt;/strong&gt; script to (try) monitor for file system changes instead.&lt;/p&gt;</description></item><item><title>Legacy HP computer info</title><link>https://myByways.com/post/legacy-hp-computer-info/</link><pubDate>Sat, 31 Dec 2016 00:00:00 +0000</pubDate><guid>https://myByways.com/post/legacy-hp-computer-info/</guid><description>&lt;p&gt;It&amp;rsquo;s the last day of 2016&amp;hellip; Time for a bit of nostalgia! Over the years, I posted bits and pieces on Blogger, myByways.net (Microsoft), Google Pages and others. For posterity, I&amp;rsquo;ve transferred selected information about a HP notebook and desktop I used to have, but I doubt any of this information is relevant any more.&lt;/p&gt;</description></item><item><title>Streaming from desktop with NVIDIA GameStream to iOS and macOS</title><link>https://myByways.com/post/streaming-from-desktop-with-nvidia-gamestream-to-ios-and-macos/</link><pubDate>Fri, 07 Oct 2016 00:00:00 +0000</pubDate><guid>https://myByways.com/post/streaming-from-desktop-with-nvidia-gamestream-to-ios-and-macos/</guid><description>&lt;p&gt;I recently discovered an alternative to remote control my Windows 7/10 desktop PC, instead of using Microsoft Remote Desktop. This method requires a computer with a modern NVIDIA graphics card with &lt;strong&gt;GameStream&lt;/strong&gt; and works with iOS and macOS using a client called &lt;strong&gt;Moonlight&lt;/strong&gt;!&lt;/p&gt;</description></item><item><title>Clean Install Windows 10 with Product Key and Tokens.dat</title><link>https://myByways.com/post/clean-install-windows-10-with-product-key-and-tokens-dat/</link><pubDate>Sun, 25 Oct 2015 00:00:00 +0000</pubDate><guid>https://myByways.com/post/clean-install-windows-10-with-product-key-and-tokens-dat/</guid><description>&lt;p&gt;To clean install Windows 10, you need more than the product key from an upgraded instance of Windows 7/8. You&amp;rsquo;ll need a file called &lt;code&gt;tokens.dat&lt;/code&gt; created during the upgrade process.&lt;/p&gt;</description></item></channel></rss>