hx507

Currently clang-tidy hook is running in serial, which can be slow if multiple large files are modified. Clang-tidy is trivially parallelize-able per file, so why don't we take advantage of...

Hi! I appreciate your good work! Since the supernet is not release, is there anyway we can replicate your results on correlations between architecture density and their robustness? It would...

Scale buf_size linearly with n_ctx

17

This appear to solve https://github.com/ggerganov/llama.cpp/issues/153 where error of `ggml_new_tensor_impl: not enough space in the context's memory pool` is thrown in interactive mode, if using a larger context size. At least...