HackerTrans
TopNewTrendsCommentsPastAskShowJobs

rawsh

34 karmajoined vor 6 Jahren

Submissions

Flash-MSA: Accelerating Million-Token Training with Sparse Attention Kernels

nanduruganesh.github.io
11 points·by rawsh·vor 3 Stunden·0 comments

Zyphra ZAYA1-base: First large-scale model trained on AMD

zyphra.com
6 points·by rawsh·vor 8 Monaten·1 comments

Debugging divergence between engine and transformers logprobs for RL

gist.github.com
2 points·by rawsh·vor 10 Monaten·0 comments