llama : have n_batch default to 512 (#1091)

* set default n_batch to 512 when using BLAS

* spacing

* alternate implementation of setting different n_batch for BLAS

* set n_batch to 512 for all cases
This commit is contained in:
eiery 2023-04-22 04:27:05 -04:00 committed by GitHub
parent 7e312f165c
commit 10f19c1121
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -20,7 +20,7 @@ struct gpt_params {
int32_t repeat_last_n = 64; // last n tokens to penalize int32_t repeat_last_n = 64; // last n tokens to penalize
int32_t n_parts = -1; // amount of model parts (-1 = determine from model dimensions) int32_t n_parts = -1; // amount of model parts (-1 = determine from model dimensions)
int32_t n_ctx = 512; // context size int32_t n_ctx = 512; // context size
int32_t n_batch = 8; // batch size for prompt processing int32_t n_batch = 512; // batch size for prompt processing (must be >=32 to use BLAS)
int32_t n_keep = 0; // number of tokens to keep from initial prompt int32_t n_keep = 0; // number of tokens to keep from initial prompt
// sampling parameters // sampling parameters