Correction for Listing 6.5 in Build a Reasoning Model From Scratch
Thanks everyone for all the kind words and feedback. Super happy that you are enjoying Build a Reasoning Model (From Scratch)!
Unfortunately, there’s a small typo in Listing 6.5 on page 198 (see the video below). The line torch.manual_seed(0) should be torch.manual_seed(5).
This is a tiny change, but it’s needed to reproduce the generated response in Listing 6.5 and the corresponding log-probability outputs later in Chapter 6. If you use 0, the code is of course still correct but the numeric results may be a bit different, which could be confusing. Anyway, it should be a small 1 character change :).
This will be fixed in the next printing.
I am sorry about the oversight, and I hope this note saves you some debugging time.
Source: lightly edited website version of my Substack note.
Read Next
MiMo-V2.6 Pro Architecture and Training Notes
Notes on MiMo-V2.6 Pro's GQA and sliding-window attention, agent training tasks, reward signals, and large RL batches.
It's Easy to Dismiss Jev as Just a Classifier
A short note on Jev's generalization, possible encoder-style architecture and training, and Choice and Noul API examples.
Pacing != Pacing Development
My take on AI model pacing as a framework for release checks and the competitive pressure around model releases.