NEW TextGrad by Stanford: Better than DSPy

Поділитися
Вставка
  • Опубліковано 8 січ 2025

КОМЕНТАРІ • 28

  • @dennisestenson7820
    @dennisestenson7820 6 місяців тому +13

    This is a concept I'd been considering myself, but I never thought of it as autodifferentiated text. Fantastic that research is being done in this direction. I knew it'd be a good idea.

    • @Caellyan
      @Caellyan 6 місяців тому +3

      I criticized this has to be done manually, but never thought of chaining 2 LLMs to achieve it. Though, it does make getting slightly better answers 3x more expensive.
      I guess it's useful for unsupervised learning though.

  • @kenchang3456
    @kenchang3456 6 місяців тому +2

    Thanks for the video. I missed the boat with DSPy but it's good to know you can just go ahead with TextGrad.

  • @giladmorad4348
    @giladmorad4348 6 місяців тому +5

    Thanks for the video, it’s very insightful!
    I have 1 thought:
    1. Textgrad and DSPy can be combined. As DSPy is mostly based on ICL and this framework focuses more on signature optimization. Additionally, the researchers in Stanford mentioned that the combined prompt on one occasion improved the prompt by 1% and it should be further studied.

    • @matty-oz6yd
      @matty-oz6yd 6 місяців тому

      DSPy is ICL and prompt optimisation combined. I hope they add text grad in somehow though

    • @giladmorad4348
      @giladmorad4348 6 місяців тому

      @@matty-oz6yd yea, good correction. I hope they add Textgrad in as an optimizer.

    • @hussainshaik4390
      @hussainshaik4390 5 місяців тому

      Their mipro v2 optmizes literally doing the same

  • @jmanhype1
    @jmanhype1 6 місяців тому +2

    sounds like we need a middleware complexity assesor that can sit in the middle and auto reject if it doesnt meet that balance

  • @hoomansedghamiz2288
    @hoomansedghamiz2288 5 місяців тому +2

    Here's an unpopular opinion: could this be considered a misuse of the notation for auto-differentiation and backpropagation? For any graph to be differentiable, it must be acyclic-like a Directed Acyclic Graph (DAG), which is typical for neural networks. However, in the LLM sphere, we see pipelines incorporating cycles, such as the RAG where blocks are repeatedly cycled through, forming what might be described as Directed Cyclic Graphs (DCGs). While using PyTorch's clean and modular syntax is appealing, applying auto-differentiation in this context could be seen as a stretch (personal opinion).

  • @mydetlef
    @mydetlef 2 місяці тому

    OK, I'm a n00b. But why should I use two models when the smarter one can give me the optimal answer straight away? In which scenarios do I need all these expensive iterations? Will I then have predefined prompts for recurring queries of the same type that can be answered directly on my smartphone by a small model?

  • @fingerstyleguitarjustingao729
    @fingerstyleguitarjustingao729 5 місяців тому

    great video, hope for you more advanced explain and experience on TextGrad!

  • @matterhart
    @matterhart 6 місяців тому +17

    Thanks stanford, though I would have called it backpromptigation. ;)

  • @asadad5162
    @asadad5162 4 місяці тому

    Great video, very informative. Textual Gradient is such a pretentious concept for me, but I do look forward to try TextGrad out. At least it is a systematic method to perform prompt optimization.....

  • @brandonheaton6197
    @brandonheaton6197 6 місяців тому +1

    Solid. I knew if the guy behind DSPy could build that, there was a better version imminent

  • @MindEmbedding
    @MindEmbedding 5 місяців тому

    Thanks for another great video! I like your presentation style. What kind of software do you use for your slides?

  • @pensiveintrovert4318
    @pensiveintrovert4318 2 місяці тому

    It is 3 months later, has either of the two approaches proven to be practically useful and is being used today?

  • @DannyGerst
    @DannyGerst 6 місяців тому +1

    You said that you used in on your tasks. Can you release part of that code in the wild? It would be really great to see a live example. That was the thing I found very challenging with DSPy. Only with the storm project I started understanding how it should work ;-)

    • @code4AI
      @code4AI  6 місяців тому +1

      Start with the four Jupyter Notebooks that I provided and you will see that you have immediately multiple new ideas for your specific tasks. I plan a new video on my insights, given my testing and maybe I have an idea how to optimize the TextGrad method further ....

  • @mlcat
    @mlcat 6 місяців тому

    26:51 what does 0 demonstrations mean? No examples of good output, only original prompt?

    • @mydetlef
      @mydetlef 2 місяці тому

      Answer from Copilot: Yes

  • @Anonymous-lw1zy
    @Anonymous-lw1zy 5 місяців тому

    Superb explanation! Thank you!

  • @pensiveintrovert4318
    @pensiveintrovert4318 5 місяців тому

    How is this different from prompt tuning (not engineering)?

    • @code4AI
      @code4AI  5 місяців тому

      Explained in the video.

  • @whig01
    @whig01 6 місяців тому

    Seems like one can prompt optimize for the same level system and never lack coherence.

  • @stephanembatchou5300
    @stephanembatchou5300 6 місяців тому

    Very informative.
    Thanks

  • @artur50
    @artur50 6 місяців тому

    Thanks for the links to colabs…

  • @GeoffLadwig
    @GeoffLadwig 6 місяців тому

    Great! Thanks

  • @spkgyk
    @spkgyk 6 місяців тому +1

    Amazing video!
    But pseudo as in pseudo-code is pronounced like sudo (syuudo)
    Not smart enough to correct anything else in this video lmao, keep up the good work! Love the channel