mirror of https://github.com/nomic-ai/gpt4all
backend: fix buffer overrun in repeat penalty code
Caught with AddressSanitizer running a basic prompt test against llmodel standalone. This fix allows ASan builds to complete a simple prompt without illegal accesses but there are still notably several leaks.pull/913/head
parent
4f2b7f7be4
commit
08f3bd2a82
Loading…
Reference in New Issue