diff options
| -rw-r--r-- | gguf-py/gguf/constants.py | 2 | ||||
| -rw-r--r-- | src/llama.cpp | 3 |
2 files changed, 4 insertions, 1 deletions
diff --git a/gguf-py/gguf/constants.py b/gguf-py/gguf/constants.py index ecb8234a..489714c4 100644 --- a/gguf-py/gguf/constants.py +++ b/gguf-py/gguf/constants.py @@ -1601,7 +1601,7 @@ KEY_TOKENIZER_CLS_ID = Keys.Tokenizer.CLS_ID KEY_TOKENIZER_MASK_ID = Keys.Tokenizer.MASK_ID KEY_TOKENIZER_HF_JSON = Keys.Tokenizer.HF_JSON KEY_TOKENIZER_RWKV = Keys.Tokenizer.RWKV -KEY_TOKENIZER_PRIFIX_ID = Keys.Tokenizer.PREFIX_ID +KEY_TOKENIZER_PREFIX_ID = Keys.Tokenizer.PREFIX_ID KEY_TOKENIZER_SUFFIX_ID = Keys.Tokenizer.SUFFIX_ID KEY_TOKENIZER_MIDDLE_ID = Keys.Tokenizer.MIDDLE_ID KEY_TOKENIZER_EOT_ID = Keys.Tokenizer.EOT_ID diff --git a/src/llama.cpp b/src/llama.cpp index b8555677..18c7cd0f 100644 --- a/src/llama.cpp +++ b/src/llama.cpp @@ -18665,6 +18665,9 @@ static ggml_type change_type_if_necessary(ggml_type new_type, int nx, int ny) { case GGML_TYPE_IQ4_KS: case GGML_TYPE_IQ4_KS_R4: case GGML_TYPE_IQ4_XS_R8: + case GGML_TYPE_IQ2_KT: + case GGML_TYPE_IQ3_KT: + case GGML_TYPE_IQ4_KT: case GGML_TYPE_IQ4_XS: new_type = GGML_TYPE_IQ4_NL; break; case GGML_TYPE_IQ4_K: case GGML_TYPE_IQ4_K_R4: |