Mistral MoE - Better than ChatGPT?

Поделиться
HTML-код
  • Опубликовано: 2 дек 2024

Комментарии • 47

  • @engineerprompt
    @engineerprompt  11 месяцев назад

    Need help with Building your LLM powered Applications? Reach out if you need help: calendly.com/engineerprompt/consulting-call

  • @alexxx4434
    @alexxx4434 11 месяцев назад +5

    I just entered local LLM operation field. What a time to be alive!

  • @teleprint-me
    @teleprint-me 11 месяцев назад +7

    We already have working GGUF files for the MoE. 2024 is, indeed, going to be extremely interesting.

  • @ArdeniusYT
    @ArdeniusYT 4 месяца назад +1

    hi, which models are currently best for web AI chat bots that are open source, can be fine tuned locally, and run on 8GB vram vps comfortably

    • @engineerprompt
      @engineerprompt  4 месяца назад

      I will look at Mistral-7B, Llama-8B and the Qwen models.

  • @Cingku
    @Cingku 11 месяцев назад +2

    This is crazy. I've just tested the latest version of TheBloke/Mistral-7B-Instruct-v0.2-GPTQ in localGPT and it gives very meaningful and coherence sources when enabled the --show_sources flag of course. And all the previous models cannot do this and simply giving random junks but not this new model! Not to mention it is blazing fast for such quality response and taking only 6 to 12 seconds using my RTX 3060 12 GB with ~3 GB of vector db size. And the system prompt really affecting the result and the default localGPT system prompt is not good for my use case and I just tweak it a little bit and ask it to double check for its mistake and bam! It seems that it can correct its own mistake! Can't imagine the power of MoE Mixtral-8x7B-Instruct-v0.1 if that small model already really good.

  • @avi7278
    @avi7278 11 месяцев назад +4

    I came up with the killers problem as part of my own benchmark, and I proposed this problem to Matthew Berman to part of his testing benchmark and he began to use it. Now I see it here! Pretty cool to see it out in the wild. I said that when an open source model finally answered this question correctly we will have arrived at the point where open source models are just as good as OpenAI's. This was before GPT 4, even chat gpt could not answer it correctly. GPT4 does now answer it correctly but there was even a period where even it gave the wrong answer.

    • @TheReferrer72
      @TheReferrer72 11 месяцев назад +2

      It has very strong reasoning capabilities from my tests. I am running the instruct version locally using LLM Studio.

    • @Someone-tc4wu
      @Someone-tc4wu 11 месяцев назад

      ​@@TheReferrer72what does instruct model mean ?

    • @alexxx4434
      @alexxx4434 11 месяцев назад +4

      @@Someone-tc4wu It's a fine-tuning of a base model to operate in instruction/response manner. While the base model operates in auto-complete mode basically.

  • @kilianlindberg
    @kilianlindberg 11 месяцев назад +1

    I like Mistral as a challenger, and a lightweight-ness ambition. Perhaps more of open source could be a path, offering beneficial cloud computing integrations.

  • @onoff5604
    @onoff5604 11 месяцев назад +2

    Is it correct that Zephyr (in the past) significantly boosted Mistral with a careful training system? Could Zephyr do this again and make a super-super-moe?

  • @Drone256
    @Drone256 11 месяцев назад +5

    Why not compare to GPT-4? With GPT-3.5 costing just 10% per token it's really a much smaller model and not the leading advancement in LLMs.

    • @jkugler1776
      @jkugler1776 11 месяцев назад

      Because it wouldn't make headlines then

    • @tatsamui
      @tatsamui 11 месяцев назад +1

      Yes sure when gpt-4 is 220m x8 moe and this one just 7x8. By far this is the best open source that can beat gpt-3.5 and not even instructions model yet

    • @Drone256
      @Drone256 11 месяцев назад +2

      @@tatsamui Anything that compares to GPT-3.5 is pretty good. At this point do you think people are figuring out how to game the tests during training? For example, you would not want the prisoner's dilemma on an LLM test anymore because people would train for that specifically.

    • @jopansmark
      @jopansmark 11 месяцев назад

      Why compare free AI model to paid service? Are you stupid?

    • @shaftymaze
      @shaftymaze 11 месяцев назад +1

      It's actually open. Free on a service. And now costs nothing. Uncensored version. The largest is going to be an open source version of gpt4. If not tomorrow. Within months. Open source is going to blow past closed by end of 2024.

  • @malikrumi1206
    @malikrumi1206 11 месяцев назад +1

    How big a difference on the benchmarks is actually statistically significant?

  • @XShollaj
    @XShollaj 11 месяцев назад +2

    Love the Mistral team - great job guys!

    • @BobbyMasteria
      @BobbyMasteria 11 месяцев назад

      great job training the Woke lobby

    • @engineerprompt
      @engineerprompt  11 месяцев назад

      I agree, they are really able to ship. Their speed is amazing.

  • @vitorbortolin6810
    @vitorbortolin6810 11 месяцев назад +4

    This restriction was removed

    • @DihelsonMendonca
      @DihelsonMendonca 11 месяцев назад +1

      What restriction, please ?

    • @mimotron
      @mimotron 11 месяцев назад

      @@DihelsonMendonca the restrictions on using the output to train another model. You're just not allowed to reverse engineering the model now.

  • @kilianlindberg
    @kilianlindberg 11 месяцев назад +1

    Noticed Phi-2 (2.7B) today, anyone tested it or an opinion in comparison?

  • @nazihfattal974
    @nazihfattal974 11 месяцев назад +1

    I hope they make an AI assistant similar to OpenAI's soon with the same tools

  • @onoff5604
    @onoff5604 11 месяцев назад +1

    excellent report! Many thanks. More api options are great. (more fully open-er source would have been great-er-er)

  • @jopansmark
    @jopansmark 11 месяцев назад +1

    It's over for OpenAI.

  • @tech-genius
    @tech-genius 11 месяцев назад

    Where does this model stand against the MPT models from MosaicML

    • @engineerprompt
      @engineerprompt  11 месяцев назад +1

      This is far better than MPT models

  • @thomashalpin2251
    @thomashalpin2251 11 месяцев назад

    This is awesome! Could you make a tutorial on getting to run on Apple Silicon?

  • @PelletsEaseFrance
    @PelletsEaseFrance 11 месяцев назад

    I've try this morning, but it's complicate to run on windows.

  • @BobbyMasteria
    @BobbyMasteria 11 месяцев назад +1

    I tested it and indeed it's pretty good compared to GPT 3.5 ! unfortunately is has the same stupid censorship which is against human rights

  • @Bot-68
    @Bot-68 11 месяцев назад

    Please recommend me a open source model to ADD EYE BLINK TO THE AVATARS ( REALLY NEED IT :( )

  • @mlyw7918
    @mlyw7918 11 месяцев назад +1

    Every new Ai chatbot is labelled as chatgpt killer 😂 … but no one managed to kill it …

    • @jopansmark
      @jopansmark 11 месяцев назад

      It did, beats 3.5 in almost all benchmarks.

  • @vinsmokearifka
    @vinsmokearifka 11 месяцев назад

    Sir, any tips if I using japanese language?

  • @tomaszzielinski4521
    @tomaszzielinski4521 11 месяцев назад +1

    Why do people keep comparing stuff to Llama? It was never a top performer, and it's not lightweight either.

  • @matthew_brown
    @matthew_brown 11 месяцев назад

    Test against GPT-4. 3.5 is good, but it isn't GPT-4

    • @jopansmark
      @jopansmark 11 месяцев назад

      Maybe 9x7 will be better

  • @jeffwads
    @jeffwads 11 месяцев назад

    Baffling that people think it's better than 3.5...haha. If you just test it with reasoning it soon becomes quite clear it isn't.

    • @jopansmark
      @jopansmark 11 месяцев назад

      It is better than gpt-3.5

  • @annwang5530
    @annwang5530 11 месяцев назад +1

    Not uncensored sucks