Researchers seek small-scale Chinchilla scaling law replications in PyTorch or JAX
turing-math-labs · reddit · 2026-11-04
turing-math-labs posted on Reddit asking PhDs and research engineers about small-scale experiments replicating the Chinchilla scaling laws using PyTorch or JAX, seeking to compare notes.
- Links to DeepMind's blog and the NeurIPS 2022 paper "Training Compute-Optimal Large Language Models" as reference
- An open call for community experience rather than a report of concrete results
More from Research
- 20-40% of alignment data is noisy: a loss-reduction scoring framework to fix it — burkov · 2026-11-04
- Scientists image quantum entangled photons in real time, and the result looks like a taijitu — CurieuxExplorer · 2026-11-04
- kalomaze: crude diffusion autoencoders may lose low-frequency info at the latent stage — kalomaze · 2026-11-04
- Structured captions boost text-to-music alignment, new Music DSG metric beats CLAP — affige_yang · 2026-11-04
- Fruit fly's 166,000-neuron circuit map is being used to drive robot behavior — CurieuxExplorer · 2026-11-04
- Orbital engineers use AI OS CurieOS to draft two-phase cold plate designs — TimothyDuignan · 2026-11-04