diff options
| author | Kawrakow <iwankawrakow@gmail.com> | 2025-03-07 08:54:09 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-03-07 08:54:09 +0200 |
| commit | c67a37b251fc22b0f8b8313ea5c76a73ff6ed49f (patch) | |
| tree | 472c093202025b58d6bc4a8de431bd470f48efde /include | |
| parent | 7bdbf99bbdbfe46b01f7783a7c98a30a1558e2c3 (diff) | |
Custom quantization rules with regular expressions (#244)
* Custom quantization rules with regular expressions
* Add the --custom-q option to the help
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Diffstat (limited to 'include')
| -rw-r--r-- | include/llama.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/include/llama.h b/include/llama.h index 38a12744..5e86cb68 100644 --- a/include/llama.h +++ b/include/llama.h @@ -418,6 +418,7 @@ extern "C" { bool ignore_imatrix_rules; // If set to true, the built-in rules for refusing to quantize into certain quants without imatrix are ignored void * imatrix; // pointer to importance matrix data void * kv_overrides; // pointer to vector containing overrides + void * custom_quants; // pointer to vector containing custom quantization rules } llama_model_quantize_params; // grammar types |