Skip to content

Instantly share code, notes, and snippets.

@mvidaldp
mvidaldp / sdcard_fix.md
Created April 11, 2022 12:43
SD card formating fix via ADB shell when Android GUI fails (internal, portable/external or mixed). Works on Retroid Pocket 2+

I wrote this short tutorial because extending my internal storage using my new micro SD card on my Retroid Pocket 2+ failed all the time. Only setting it up as portable/external worked. However, this instructions should work in any Android 5.0+ device.

So, in case you have problems setting up your SD card on your Android device via graphical interface (setting up storage as extended internal memory or portable), and you get a corrupted SD card or any other error, follow these steps to fix it via adb shell:

  1. Make sure you have adb access to your Android device: Settings > System > About, touch/click on Build number until Developer options are enabled:
  2. Go to Settings > System > Developer options and enable USB debugging.
  3. Assuming you have adb installed on your remote terminal run the following:

adb shell

Reinforcement Learning for Language Models

Yoav Goldberg, April 2023.

Why RL?

With the release of the ChatGPT model and followup large language models (LLMs), there was a lot of discussion of the importance of "RLHF training", that is, "reinforcement learning from human feedback". I was puzzled for a while as to why RL (Reinforcement Learning) is better than learning from demonstrations (a.k.a supervised learning) for training language models. Shouldn't learning from demonstrations (or, in language model terminology "instruction fine tuning", learning to immitate human written answers) be sufficient? I came up with a theoretical argument that was somewhat convincing. But I came to realize there is an additional argumment which not only supports the case of RL training, but also requires it, in particular for models like ChatGPT. This additional argument is spelled out in (the first half of) a talk by John Schulman from OpenAI. This post pretty much

@veekaybee
veekaybee / normcore-llm.md
Last active December 6, 2025 11:50
Normcore LLM Reads

Anti-hype LLM reading list

Goals: Add links that are reasonable and good explanations of how stuff works. No hype and no vendor content if possible. Practical first-hand accounts of models in prod eagerly sought.

Foundational Concepts

Screenshot 2023-12-18 at 10 40 27 PM

Pre-Transformer Models