gotty VS transformers

gotty + lnav

My bootstrap script ends in launching AutoGPT with GoTTY https://github.com/sorenisanerd/gotty

Setup GoTTY and allow it to run a single script only that will restart VPS. Assign password to GoTTY and give it to only trusted people to use. No need for them to install anything since GoTTY is basically "terminal in browser"

First, I suggest you to use forked GoTTY version since original one in your post wasn't updated for more than 6 years.

GoTTY. You dont need to know Golang, just download binary from release section for your platform

Honestly graylog/loki is only worth it if you want to have automatic processing/stats generation/graphing and complex log management rules. If you just want to read logs in a web interface I suggest either frontail (very basic, a bit too much for my taste) or lnav (I use this 99% of the time, over SSH) + gotty to access a terminal/lnav from a web browser - be careful to secure it properly as it basically gives shell access to your server.

Write a script(s) (docker container plex restart) that you will allow to run via web and point to it gotty, so users will run it as they are in terminal. (Just don't forget to set GOTTY_CREDENTIAL on an exposed link).

I created a custom image from https://github.com/sorenisanerd/gotty with lnav in order to view application logs in a browser. The Dockerfile and docker compose I used are below. I created symbolic links to the apps I am interested in into a single directory and mounted it to /docker/logs in the container. Thought I would share if someone else is looking for a lightweight solution to consolidate and view logs in a browser.

Having trouble with ssh? No problem, browser is all what people need, just run GoTTY and get full (or read only) access to any terminal in multiple tabs in web browser, as much as you need.

sorenisanerd/gotty: Share your terminal as a web application

Fascinating work, very promising.

Can you summarise how the model in your paper differs from this one ?

https://github.com/huggingface/transformers/issues/27011

transformers is a package by Hugging Face, that helps you interact with models on HF Hub (GitHub)

Is t5x an encoder/decoder architecture?

Some more general options.

The Flax ecosystem

https://github.com/google/flax?tab=readme-ov-file

or dm-haiku

https://github.com/google-deepmind/dm-haiku

were some of the best developed communities in the Jax AI field

Perhaps the “trax” repo? https://github.com/google/trax

Some HF examples https://github.com/huggingface/transformers/tree/main/exampl...

Sadly it seems much of the work is proprietary these days, but one example could be Grok-1, if you customize the details. https://github.com/xai-org/grok-1/blob/main/run.py

The HuggingFace transformers library already has support for a similar method called prompt lookup decoding that uses the existing context to generate an ngram model: https://github.com/huggingface/transformers/issues/27722

I don't think it would be that hard to switch it out for a pretrained ngram model.

Most of this tutorial is based on Hugging Face course about Transformers and on Niels Rogge's Transformers tutorials: make sure to check their work and give them a star on GitHub, if you please ❤️

* Superconvergence + LR range finder + Fast AI's Ranger21 optimizer was the goto optimizer for CNNs, and worked fabulously well, but on transformers, the learning rate range finder sadi 1e-3 was the best, whilst 1e-5 was better. However, the 1 cycle learning rate stuck. https://github.com/huggingface/transformers/issues/16013

Thanks! :) I'm pushing them into transformers, pytorch-gemma and collabing with the Gemma team to resolve all the issues :)

The RoPE fix should already be in transformers 4.38.2: https://github.com/huggingface/transformers/pull/29285

My main PR for transformers which fixes most of the issues (some still left): https://github.com/huggingface/transformers/pull/29402