happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Add B/F16 to get_ftype()

This commit is contained in:
Ed Addario 2025-11-29 23:52:51 +00:00
parent 229109f329
commit b97cda6289
No known key found for this signature in database
GPG Key ID: E7875815A3230993
1 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
tools/quantize/quantize.cpp
View File

@ -512,7 +512,12 @@ static const char * get_ftype(const float bpw) {
{4.5000, "Q4_K"}, {4.5000, "Q4_K"},
{5.5000, "Q5_K"}, {5.5000, "Q5_K"},
{6.5625, "Q6_K"}, {6.5625, "Q6_K"},
{8.5000, "Q8_0"} {8.5000, "Q8_0"},
#ifdef GGML_USE_METAL
{16.0000, "F16"}
#else
{16.0000, "BF16"}
#endif
}; };
return quant_bpw.lower_bound(bpw)->second; return quant_bpw.lower_bound(bpw)->second;