Add cotributing.md - Githubissues

tomaarsen / attention_sinks

Extend existing LLMs way beyond the original training length with constant memory usage, without retraining

Apache License 2.0

650 stars 41 forks source link

Open rajveer43 opened 10 months ago