Log inSign up
Alex Strick van Linschoten
3,289 posts
user avatar
Alex Strick van Linschoten
@strickvl
ML/AI Engineer (@zenml_io / @kitaru_ai), researcher (& author of a few books + PhD 👨‍🎓). feedback welcome admonymous.co/strickvl 🙏
Delft, The Netherlands
alexstrick.com
Joined January 2011
595
Following
3,178
Followers
RepliesRepliesMediaMedia

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
  • Pinned
    user avatar
    Alex Strick van Linschoten
    @strickvl
    Dec 15, 2025
    🎉 The @zenml_io LLMOps Database just crossed 1,000+ case studies! 17 months of curating real-world LLM production stories. Here's how the collection grew over time 📈 What started as a side project in July 2024 is now the largest open collection of production LLM case
    Image
    3.1K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Dec 2, 2024
    🤔 Do you ever wonder how companies are putting LLMs and GenAI apps into production? What stacks do they use? What architecture did they go with? I put together a database of known public technical writeups with summaries of the key technical features.
    Image
    GIF
    76K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    May 9, 2022
    I built an MVP of my redacted document detector as part of @jeremyphoward's course. I used: - @fastdotai to classify & extract redacted pages extracted from PDFs - IceVision (@ai_fast_track) to detect redacted areas - @huggingface Spaces (with @Gradio and @streamlit ) to deploy
    Image
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Apr 7, 2025
    🚀 I built a friendly version of the Gemini SDK docs. Google's @GeminiApp models are real workhorses and while working on some projects I found myself wanting a simpler version of their docs for common use cases and features. @OfficialLoganK @_philschmid @osanseviero
    Front page screenshot of geminibyexample.com
    Screenshot of the image generation docs page of geminibyexample.com
    41K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Nov 10, 2022
    Replying to @vamchale and @elonmusk
    mathstodon.xyz has that!
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Oct 24, 2022
    This week in the @fastdotai part 2 course we started the exciting work of implementing our own deep learning library from scratch (mostly). This involved @jeremyphoward getting the data from the source files, parsing them out, defining a pseudo-tensor (or Matrix) class and so on.
    A tabular overview of the topics covered in the class.
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Apr 1, 2025
    Replying to @Good___Listener @launchschool and 2 others
    Do @fastdotai part one. Do it properly. Do the things @jeremyphoward says to do in the course lectures. (Then maybe do part 2, depending on your needs). You'll already be ahead of 80% of people just by doing that. And once you're done then you can follow your interests with
    24K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Jun 15, 2024
    Had a blast finetuning my first LLMs as part of the @HamelHusain @dan_s_becker course. Used @winglian's axolotl and was most struck by how accessible the process is: as a user, ignore most of the details and get decent results!
    Image
    Finetuning my first LLM(s) for structured data extraction with axolotl – Alex Strick van Linschoten
    From alexstrick.com
    30K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Jul 1, 2024
    I managed to beat GPT-4's performance for structured data extraction using finetuned models. (This was part of my continued work for the @HamelHusain / @dan_s_becker course.)
    Image
    My finetuned models beat OpenAI’s GPT-4 – Alex Strick van Linschoten
    From alexstrick.com
    15K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Jun 25, 2024
    I'm in the process of coding up the manual evaluation data set for my LLM finetuning work around structured data. I wrote this blog to capture the high-level categories for how I'm thinking about these kinds of 'unit-test'-like evaluations of LLM output.
    Image
    How to think about creating a dataset for LLM finetuning evaluation – Alex Strick van Linschoten
    From alexstrick.com
    7.1K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    May 4, 2025
    I think about this quite regularly. If I had to distill down 'the @fastdotai way', I'd capture the following: - work with your hands on the controls (prototype quickly -> polish (then maybe -> production)) - smaller slices makes for fast cycles (train or work on tiny subsets
    user avatar
    Radek Osmulski
    @radekosmulski
    May 3, 2025
    A big misconception I unconsciously held about the @fastdotai way is that it only applied to “sciency” fields. But that is not the case. @fastdotai courses worked so well for me precisely because they took a humanising approach, because they incorporated at a deep level an
    28K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    Oct 17, 2022
    Really enjoyed the start of part two of the @fastdotai course (taught by @jeremyphoward, @johnowhitaker, @iScienceLuvr and @wasimlorgat) last week. I wrote up some of my notes and reflections on the first lecture (AKA 'lesson 9'). mlops.systems/computervision…
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    May 30, 2024
    Replying to @Seamus_Malek
    "You guys should know that. You've been to college; I haven't"
    21K
  • user avatar
    Alex Strick van Linschoten
    @strickvl
    May 20, 2025
    A ton to think through after just the first session of @HamelHusain and @sh_reya's AI Evals course. Wrote up some initial notes and reflections (link in the thread) but ultimately I think this process (adapted / cribbed from one presented in the course) is the real gold.
    The analyse, measure, improve loop for improving LLM applications
    10K
Advertisement
Advertisement