LobeChat for Ollama.... any good?

Поделиться
HTML-код
  • Опубликовано: 4 июл 2024
  • Get ready to dive into the world of Ollama integrations with me! In today's video, we're taking a look at LobeChat. With its busy webpage and unclear documentation, it can be tough to figure out how to get started with this tool. But don't worry, I'm here to walk you through the process step by step. From setting up your settings to exploring the features of LobeChat, we'll cover everything you need to know. And let me tell you, it's not all sunshine and rainbows - there are some frustrating parts to this tool that might leave you feeling like you're stuck in a never-ending loop! So if you're curious about how LobeChat stacks up against other Ollama integrations, or if you just want to see what makes this tool tick, then keep watching!
    I showed this page on my site: bit.ly/ollamaintegrations which points to this ugly address: www.technovangelist.com/notes...
    You can find the code for every video I make at github.com/technovangelist/vi.... Then find the folder name that starts with the date this video was published and a title that makes sense for what the video covers.
    Be sure to sign up to my monthly newsletter at technovangelist.substack.com/...
    I have a Patreon at / technovangelist
    You can find the Technovangelist discord at: / discord
    The Ollama discord is at / discord
    (they have a pretty url because they are paying at least $100 per month for Discord. You help get more viewers to this channel and I can afford that too.)
  • НаукаНаука

Комментарии • 48

  • @technovangelist
    @technovangelist  25 дней назад +6

    I go back and forth about whether I should make stuff that has a mostly negative feel or not. If I can’t say anything nice, should I say anything at all? That’s what’s held me back from doing a video on LMStudio and others. But then folks seem to want them.

    • @1vEverybody
      @1vEverybody 24 дня назад +1

      @@technovangelist I am all for this simply because it’s basically a “what not to do” guide, plus the AI space is full of people hyping things without any criticality.

    • @HerroEverynyan
      @HerroEverynyan 22 дня назад

      I'm so happy that you made this video. I've been confused repeatedly by LobeChat. It seemed so popular based on stars, while also having frequent updates, and a sleek UI. But when trying it it was crowded and didn't bring anything new. I installed it on 3 separate occasions I think, because I just felt like maybe I missed something. I love that you put in the time to do thorough research. Finally, I'm at peace. Thank you!

    • @CyberSyntek
      @CyberSyntek 20 дней назад

      Given your knowledge, there is nothing wrong with telling it how it is. It will help others save themselves headaches and potentially give the devs constructive actual user feedback (if they watch the video) for them to potentially improve upon. Both of which can produce positive outcomes from a negative approach. :)

    • @corruptedMegabit
      @corruptedMegabit 7 дней назад

      I appreciate the good info, whether good or bad!

  • @y.m.o6171
    @y.m.o6171 25 дней назад +15

    the license alone is what would set me away from lobechat .. its a disgusting way of releasing something as ASL2 when it is absolutely not. what an insult to people intelligence !

  • @iamAliAhad
    @iamAliAhad 25 дней назад +5

    This is a great example of why modern web development and design can be problematic. Avoid overengineering or overdesigning your website just to make it look good. Keep it simple instead.

    • @technovangelist
      @technovangelist  25 дней назад +1

      totally agree.... i don't know if this is an example of bad design, but rather an example of copywriting to the extreme.... a little too much time reading adweeks guide to copywriting trying to convince us it's great.

  • @1vEverybody
    @1vEverybody 25 дней назад +12

    “Hey this thing sucks”
    Could’ve been a 4 second video

    • @technovangelist
      @technovangelist  25 дней назад +8

      there would be no reason to believe it.

    • @juanjesusligero391
      @juanjesusligero391 24 дня назад

      @@technovangelist But if you'd have just said that at the beginning, you'd have saved time to some people (not to the majority of us here, who would surely wait until the end of the video for some water drinking silence XD).

    • @technovangelist
      @technovangelist  24 дня назад +5

      I will now double down on the effort to never reveal anything like that early and keep you waiting till after the drinking silence.

    • @themax2go
      @themax2go 22 дня назад

      @@technovangelist and at ~10min of vid, it's not too long to watch... and if someone doesn't want to spend those ~10 mins to watch, then they can go through the trouble and have it summarized via a yt vid transcript gpt query (ideally via ollama) 😎

  • @trsd8640
    @trsd8640 24 дня назад +1

    Thank you for this important review! I thought I was doing something wrong, but now I know it wasn’t my fault!

  • @Leocesar3D-mx6rm
    @Leocesar3D-mx6rm 25 дней назад +5

    If tools don't support local AI, like this frontend and Microsoft 's Graph RAG, it is a no go for me.

    • @technovangelist
      @technovangelist  25 дней назад +3

      well, this thing does support local ai, just not very completely.

    • @poisonza
      @poisonza 25 дней назад

      i think you can run ms graph rag with local llm.

    • @themax2go
      @themax2go 22 дня назад

      graphRag does support local (but a bit tricky to implement AKA gotta read its docs and also ollama's to understand how to make it work... it's rather easy then)

  • @sammcj2000
    @sammcj2000 25 дней назад +2

    Pretty much mirrors my experience. It’s like a frontend dev with some really nice UI designs (and the idea for plugins) whipped it up then stepped back and asked - architecture? What’s architecture?

  • @KhalilKhamlichi-yn5wn
    @KhalilKhamlichi-yn5wn 25 дней назад +3

    I just Love your videos

  • @theMonkeyMonkey
    @theMonkeyMonkey 24 дня назад

    i notice the notes page doesn't include Flowise - that does have a working set of ollama integrations

    • @technovangelist
      @technovangelist  24 дня назад

      Then someone should add it to the GitHub ollama repo

  • @tohur
    @tohur 21 день назад

    Anything that runs locally typically can't use your microphone due to https or the lack there of on localhosts.. in chrome based browsers you can allow chrome to treat a list of hosts/IPs as if they were running on https thus allowing you to use the voice to text

    • @technovangelist
      @technovangelist  21 день назад

      Perhaps you should tell that to the developers of other products that have it working. Or suggest to lobechat that they remove a feature they can’t get to work.

  • @elwoodfanwwod
    @elwoodfanwwod 25 дней назад

    Had the same issues and fully agree.

  • @nocodecreative
    @nocodecreative 25 дней назад

    Id like to have see it working with the online models

  • @jimlynch9390
    @jimlynch9390 25 дней назад

    I still don't know what it is supposed to do.

    • @technovangelist
      @technovangelist  25 дней назад

      Barely does anything unless you are using the online models

  • @martin22336
    @martin22336 24 дня назад

    I use page assist in a browser and its pretty good for what it does. I get it does not have the craziest updates but its pretty good fast and I cant complain

    • @technovangelist
      @technovangelist  24 дня назад

      i was really impressed with page assist. it still stands out as one of the better simple tools out there

  • @hibanajah-z6m
    @hibanajah-z6m 16 дней назад

    its not lobechat fault that plugins dont work , most of them can work nicely with gpt however using free option limits you burry you does your chosen model support function call?

    • @technovangelist
      @technovangelist  16 дней назад

      It’s totally lobechats fault. Pretty much all models support function calling with ollama.

    • @hibanajah-z6m
      @hibanajah-z6m 15 дней назад

      @@technovangelist just try the last version now everything work better with gpt-x

    • @technovangelist
      @technovangelist  15 дней назад

      Gpt-x???

  • @ringpolitiet
    @ringpolitiet 25 дней назад

    It's okay to not publish a video if it doesn't show anything of value.

    • @technovangelist
      @technovangelist  25 дней назад +4

      I want to look at all the tools available. Some are going to be great and some aren’t.

    • @technovangelist
      @technovangelist  25 дней назад +3

      I first read that as it being ok not to publish if it’s not positive. But just now realized you may have meant simply if there was no value. I would disagree with that. I think there is a lot of value pointing out things that don’t work so tools can get better.

  • @everaldogomes7343
    @everaldogomes7343 24 дня назад

    I like your channel very much. But I just wasted 11m48s watching a video about a tool that doesn't do anything.

    • @technovangelist
      @technovangelist  24 дня назад +3

      Sounds like a good deal. You could have spent 45 minutes trying it out and been even more upset.

    • @everaldogomes7343
      @everaldogomes7343 24 дня назад +1

      @@technovangelist fair enough. Actually, I dont use any tools yet. I just watch the vídeos. And thanks to you I've learned a lot about Ollama. In the future, I Will use Ollama for my projects, thanks to you. I never comment on RUclips, but I like very much your work, that's why I gave my 2 cents. Best regards from Brazil!

    • @technovangelist
      @technovangelist  24 дня назад +2

      Thanks so much