e_ameisen
Joined 9 years ago
Karma
544
Submissions
67
Member since
Jun 23
Recent submissions
Comment · 1 year ago
Comment · 6 years ago
http://mlpowered.com/
I share practical tips about ML. I focus specifically on things that will make it easier to ship ML applications. As much as possible, I include code.
Popular posts:
https://mlpowered.com/posts/how-to-solve-90-nlp/
https://mlpowered.com/posts/image-search/
I've paused writing for a bit while I wrote my book (https://mlpowered.com/book/) but now am planning to pick back up.
I have a newsletter at the bottom of each page and an RSS feed you can subscribe to to receive new posts.
Comment · 6 years ago
Disclaimer: This is a book I wrote.
Building Machine Learning Powered Applications walks you through building an ML application end-to-end, from product idea to a (simple) deployed version.
The free first chapter is available here https://mlpowered.com/book/
The github is at https://github.com/hundredblocks/ml-powered-applications
Comment · 6 years ago
I'm glad you have found a workaround. Again, this is for an example script outside of the main narrative of the book, so the impact should be minimal.
If pip struggles to find a specific version, I'll usually remove the specific version requirement, and give whichever version pip finds a try by running `pip install tensorflow`
Comment · 6 years ago
Interesting. I wasn't aware of that issue. It seems like after running activate, `pip` does point to the right location (`which pip` points to the virtualenv pip), but I'll look into it.
Comment · 6 years ago
Sorry to hear that! Has tensorflow been your main issue? It is only used in one of the example notebooks, so you can skip that requirement without too much of an impact.
If I can help with troubleshooting, send me an email at [email protected]
In the meantime I'll bump up the version in requirements.txt to solve the conflict.
Comment · 6 years ago
That’s a great question, and I’m not sure. Googling around led me to an HN thread where a commenter claims that it is TheSans (lucasfonts.com/fonts/the-sans) but I do not have a good enough eye to confirm!
Comment · 6 years ago
O'Reilly did organize the tech review process, where they found multiple reviewers to give detailed feedback on every part of the book. The O'Reilly staff themselves helped immensely with the editing process, as well as decisions about which topics would be the most worthwhile to include.
Comment · 6 years ago
I answered the next comment down the chain, but I agree with you. As framed, there is no satisfactory solution to making a question "good".
Improving question quality could however be a product goal for StackOverflow, or for a company focusing on writing tools (Grammarly, Textio,...). The book describes a process for turning that vague product goal into a more tangible set of metrics, which lead to choosing an ML approach, and iterating on it.
Eventually there is a finished prototype, for which you can find a GitHub link in the PDF preview. It has definitely not solved how to evaluate if a question is "good", but aims to provide a narrower set of recommendations (the first chapter actually dives into an approach for this).
Comment · 6 years ago
I'm so glad to hear you have been enjoying it!
This is a new tool which relies on existing introspection libraries like TransformerLens (which is similar in spirit to Garcon) to build an attribution graph. This graph displays intermediate computational steps the model took to sample a token.
For more details on the method, see this paper: https://transformer-circuits.pub/2025/attribution-graphs/met....
For examples of using it to study Gemma 2, check out the linked notebooks: https://github.com/safety-research/circuit-tracer/blob/main/...)
We also document some findings on Claude 3.5 Haiku here: https://transformer-circuits.pub/2025/attribution-graphs/bio...)