Відео

i love the sound and narration

Good tutorial, but how tf is Bing superior to Google? It's basically the same thing but worse and also not private

Stupid question: I can't find Prompt template config in Text Generator. How to overcome this?

HOLY COW YOU ARE A LEGEND THIS IS WHAT IVE ALWAYS WANTED FOR OBSIDIAN!!

I cannot understand why such a feature rich tool is not more well known

How can you highlight from the subtitles automatically

Are they using same ai model, are there projects in which aider is not the best

How do I run this just within LMStudio? I load the onsidian vision model, go to chat, select a picture to describe but the response is just "### Comment: ### Response:" scrolling up the screen ?

5 минут пустого разговора видио не очем

Would be great to pull in audio transcripts by adding video url

Thanks for sharing, I would only expect to do it using python before. Using obsidian is more user friendly, and you can also easily use it on different devices and with different models.

Great stuff! Thanks

Ok...I've tested it for a couple of hours now using many different models. With online models there is no problem, even when throwing at them a couple of folders full of files. With locally hosted models, the same results in hallucinations or the model replying it is not capable of accessing local reference files. This happens even if the "/set parameter num_ctx 32768" instruction gets passed to the model via ollama CLI. To the opposite, if the local model is passed only 1 file....then boom! even small models seem to work at first attempt. So my conclusion is that in order to work with multiple files on local models you need to add context window recognition to SystemSculpt so that one can finally understand if the model capability gets exceeded or not.

What about exporting titles only to a few note files as context? Would that work?

Muito obrigada!

Did you find recommended settings for the Git plugin? I over here in analysis paralysis because I want it to sync automatically, but the details like how often, and syncing on startup / shutdown have me frozen as I don't know what the "best" settings for most people are.

Thanks, very helpful & concise

Aider is the best by far. This video is 9 months old, by now Aider got so many updates and is way-way better

cannot thank you enough!

Thanks so much for sharing this! 😊Just wondering, do the link to the original author and the date at the end come from Omnivore or Obsidian? Also, my Omnivore is only able to save the first few lines of a tweet. Is this the case for anyone else?

Would you say this is better than the Mistral 7B model you mentioned in an earlier video? I'm a chemist who's completely out of my field asking this, but I stumbled upon this channel through your plugin, which I love. It inspired me to start learning typescript, just not enough hours in a day to get to all this fun stuff.

Is it not possible to put the folder on one drive? Or are you actively using the versioning?

I little modify it " where file.cday = date(this.file.name) and file.name != this.file.name" so it didn't show your daily note in this list of created notes.

Hello Mike, I love your plugin and have been using it for a few weeks now. It's replaced many other plugins and has significantly sped up my workflow, so thank you! However, since updating the plugin today, it’s not working at all. I’m using the latest version of Obsidian on Windows. How can I resolve this issue? I don't have your email, but I regularly watch your videos, so I'm reaching out here in the hopes that you'll see this comment.

Table of Contents plugin does not do what you show in the video. It’s a plugin that creates a TOC inside of the note instead. The side panel you show is the core Outline plugin.

Sure wish the implementation of Vim in Obsidian was more robust and less buggy. I see there are many Vim related plugins now, so maybe I need to explore some of them. But things like having to hit escape twice to get out of edit mode (sometimes) and not being able to navigate through an entire note by going to top or bottom on note using G and gg (only moves to the top and bottom of the visible portion of the note on the screen) are just a couple things that come to mind.

I just went through a few of your videos. Subscribed! Fantastic stuff. Any chance of Anthropic/Claude support?

Great video, thanks for sharing. Highly informative stuff.

I ended up firing you off an email. Still having issues getting the Claude API to run. I will be trouble shooting more today. Thanks!

Is it not amusing that we use 'kick start' to indicate an effortful sudden climactic start to something. I prefer a power-start myself having ridden motorcycles with both kick-starts (hard work, unreliable) and electric starts (instant, reliable, full power, first time). So I don't 'kick-start' anything. How primitive that would be. I 'power-start', or just plain 'begin'.

Hey, great plugin, can't wait to try. Question. I've installed the plugin from the Community Plugins UI, set an OpenAI key, but does not work. With an old key I can select the model but does not work. With the new Project Keys, it says "valid key" but no models appear to select 😞 any idea?

Not sure what I missed but: I am trying to add Claude as my default and its not giving me the option to (any option actually)....

I've just found this - really interesting! Is it possible to use a local model to do the transcription/whisper? rather than the OpenAI or Grok one? (I want to do it all local and for free!)

It is incredible; thank you so much. Does it support Windows?

ive got a monthly subscription for open AI, i wish I could just use that instead of paying for API on top of it.

This looks amazing. You’re doing some great things. It’s quite overwhelming. There’s a lot of “update videos” but no walk-through of how to get started using it (not installing it). I’m looking forward to the “How-to” video for all of it’s capabilities.

Great video. Thank you. Where is the script link to get it from your website? Does it require a Patreon membership? Already subscribed and liked it. Thanks for the great work,

Love this already

There is definitely a sizable audience for Vim content. Do it!

this looks fantastic. i like that it has an openrouter option which opens up all the non-openai models

Thank you very much. I look forward to a new video in which you explain how to use your own documents (RAG).

What model would you recommend for local? I am using Mistral

You genius.

why the hell are subtitles on this thing? cant watch

Really good Channel if it comes to Ai and Tech without no unnecessary clickbait or something like that, keep it up man 🔥

Hey, can this use ollama as an end-point ?

very nice. can you make it to support mobile Obsidian?

Why didn't you tell the people Aider can work with ONLY one file total? It's useless for all the projects except the tiny ones. Also, you should show the incremental capabilities of the tools. One thing is to generate first version of the app. But iterate over it and make it better over time is much much more important.

Would you still recommend this or Phi3/Phi3 mini 124k? My notes are a medical textbook devided in pieces. Probably not the easiest to dissect.