mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-01-31 03:27:39 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Incorporate some fixes and updates for GGML

Browse source
This commit is contained in:
Justine Tunney 2023-04-28 20:24:55 -07:00
parent b31ba86ace
commit d9e27203d4
No known key found for this signature in database
GPG key ID: BE714B4575D6E328
2 changed files with 24 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
third_party/ggml/ggml.c vendored
View file

@ -3641,6 +3641,24 @@ static void ggml_vec_dot_q8_0_q8_0(const int n, float * restrict s, const void *
}
*s = vaddvq_f32(sumv0) + vaddvq_f32(sumv1);
#elif defined(__AVX2__)
// Initialize accumulator with zeros
__m256 acc = _mm256_setzero_ps();
// Main loop
for (int i = 0; i < nb; ++i) {
// Compute combined scale for the block
const __m256 d = _mm256_mul_ps( _mm256_broadcast_ss( &x[i].d ), _mm256_broadcast_ss( &y[i].d ) );
__m256i bx = _mm256_loadu_si256((const __m256i *)x[i].qs);
__m256i by = _mm256_loadu_si256((const __m256i *)y[i].qs);
const __m256 q = mul_sum_i8_pairs_float(bx, by);
// Multiply q with scale and accumulate
acc = _mm256_fmadd_ps( d, q, acc );
}
*s = hsum_float_8(acc);
#else
// scalar
float sumf = 0.0;

6
third_party/ggml/main.cc vendored
View file

@ -52,6 +52,7 @@ asm(".include \"libc/disclaimer.inc\"");
static console_state con_st;
static llama_context ** g_ctx;
static int g_verbose;
static bool is_interacting = false;
#define EPHEMERAL(fmt) "\r\e[K\033[1;35m" fmt " \033[0m"
@ -64,6 +65,9 @@ void sigint_handler(int signo) {
if (!is_interacting) {
is_interacting=true;
} else {
if (g_verbose) {
llama_print_timings(*g_ctx);
}
_exit(128 + signo);
}
}
@ -108,6 +112,8 @@ int main(int argc, char ** argv) {
win32_console_init(params.use_color);
#endif
g_verbose = params.verbose;
if (params.perplexity) {
printf("\n************\n");
printf("%s: please use the 'perplexity' tool for perplexity calculations\n", __func__);