Mechanical Dirk

Mechanical Dreams

Science EN ↓ 170 Folgen

An automatically generated podcast about machine learning and natural language processing. The two fictional hosts talk about papers that I want to learn more about on my way to work. It's not good, but it's useful.

Besuch unbedingt die Website des Podcasts und unterstütze die Macher: mechanicaldreams.mechanicaldirk.com

Autor

Mechanical Dirk

Kategorie

Science

Neueste Folge

27. Sep 2026

Wo hören?

Podcasts in der App Replaio Radio Bald verfügbar

Podcasts kommen bald in die App. Installiere sie jetzt und erlebe als Erster einen ganz neuen Blick auf Podcasts

Bei Google Play herunterladen Kostenlos installieren Android fast 10 Mio. Downloads · Bewertung 4,8 iOS bald

Folgen

The Zamba2 Suite 29.11.2024
Small-scale proxies for large-scale Transformer training instabilities 26.11.2024
Hardware Scaling Trends and Diminishing Returns in Large-Scale Distributed Training 25.11.2024

I slightly tweaked the personality of the hosts.

Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations 19.11.2024
Understanding WSD Learning Rates 18.11.2024
Toward Understanding Why Adam Converges Faster Than SGD for Transformers 16.11.2024

New generation algorithm! Should make the episodes longer, more detailed, and more coherent.

Amuro & Char - Analyzing the Relationship between Pre-Training and Fine-Tuning of Large Language Models 08.11.2024
Evaluation data contamination in LLMs: How do we measure it and (when) does it matter? 07.11.2024
How Does Critical Batch Size Scale in Pre-training? 04.11.2024
The Road Less Scheduled 01.11.2024
Learning-Rate-Free Learning by D-Adaptation 31.10.2024
Scaling FP8 Training to Trillion Token LLMs 30.10.2024
Geometric Dynamics of Signal Propagation Predict Trainability of Transformers 29.10.2024
A Survey on Model MoErging 28.10.2024
Liquid Time-constant Networks 27.10.2024
Scaling Laws for Predicting Downstream Performance in LLMs 26.10.2024
A Spectral Condition for Feature Learning 25.10.2024
Don't decay the learning rate 24.10.2024

A classic paper about learning rates.

OLMoE 23.10.2024

professor norris: Welcome back to Mechanical Dreams, the podcast where we delve into the exciting world of machine learning and natural language processing. I'm Professor Norris, and as always, I'm joined by my brilliant student, Linda. linda: It's great to be back, Professor. And I'm particularly excited about today's paper. It tackles a topic that's been buzzing in...

An Empirical Model of Large Batch Training 23.10.2024

First attempt to automatically generate a podcast from a paper. This one is way too short, but it's a start.

Höre den Podcast Mechanical Dreams in Replaio

Radio und Podcasts in einer App - kostenlos und ohne Anmeldung. Installiere sie noch heute und verpasse den Start nicht

Bei Google Play herunterladen

Replaio ist kein Herausgeber von Podcasts; die Namen der Sendungen, Cover und Audioinhalte gehören ihren Autoren und werden über öffentliche RSS-Feeds verbreitet