Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Commit 77a7340

Browse files
authored
ggml : add new Q4_2 quantization (ARM only) (#1046)
* ggml : Q4_2 ARM * ggml : add ggml_is_quantized() * llama : update llama_type_name() with Q4_2 entry * ggml : speed-up q4_2 - 4 threads: ~100ms -> ~90ms - 8 threads: ~55ms -> ~50ms * ggml : optimize q4_2 using vmlaq_n_f32 + vmulq_n_f32
1 parent 50a8a2a commit 77a7340

File tree

5 files changed

+287
-11
lines changed

5 files changed

+287
-11
lines changed

examples/quantize/quantize.cpp

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -14,6 +14,7 @@ int main(int argc, char ** argv) {
1414
fprintf(stderr, "usage: %s model-f32.bin model-quant.bin type\n", argv[0]);
1515
fprintf(stderr, " type = %d - q4_0\n", LLAMA_FTYPE_MOSTLY_Q4_0);
1616
fprintf(stderr, " type = %d - q4_1\n", LLAMA_FTYPE_MOSTLY_Q4_1);
17+
fprintf(stderr, " type = %d - q4_2\n", LLAMA_FTYPE_MOSTLY_Q4_2);
1718
return 1;
1819
}
1920

0 commit comments

Comments
 (0)