CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-02 09:12:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

gguf : start implementing quantization (WIP)

Browse Source
This commit is contained in:
M. Yusuf Sarıgöz
2023-08-12 15:55:58 +03:00
parent b2571af255
commit fa7c39540c
2 changed files with 19 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
gguf-util.h
View File

@@ -109,18 +109,22 @@ struct gguf_file {
size_t write_str(const std::string & val) {
size_t total_written = 0;
const int32_t n = val.size();
total_written += fwrite((const char *) &n, sizeof(n), 1, fp);
total_written += fwrite(val.c_str(), n, 1, fp);
fwrite((const char *) &n, sizeof(n), 1, fp);
total_written += sizeof(n);
fwrite(val.c_str(), n, 1, fp);
total_written += n;
return total_written;
}
size_t write_i32(int32_t val) {
return fwrite((const char *) &val, sizeof(val), 1, fp);
fwrite((const char *) &val, sizeof(val), 1, fp);
return sizeof(val);
}
size_t write_u64(size_t val) {
return fwrite((const char *) &val, sizeof(val), 1, fp);
fwrite((const char *) &val, sizeof(val), 1, fp);
return sizeof(val);
}
void write_raw(const void * data, size_t size) {