vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Remove deprecated rules for token embeddings

Browse source
This commit is contained in:
Nexesenex 2024-08-24 21:57:07 +02:00
parent 844d11b8f3
commit 53b8eaa316
1 changed files with 0 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

3
src/llama.cpp
View file

@ -16514,9 +16514,6 @@ static ggml_type llama_tensor_get_type(quantize_state_internal & qs, ggml_type n
if (qs.model.hparams.n_vocab >= 127999 || qs.model.hparams.n_head() <= 20) new_type = GGML_TYPE_IQ3_S; if (qs.model.hparams.n_vocab >= 127999 || qs.model.hparams.n_head() <= 20) new_type = GGML_TYPE_IQ3_S;
new_type = GGML_TYPE_IQ4_XS; new_type = GGML_TYPE_IQ4_XS;
} }
else if (ftype == LLAMA_FTYPE_MOSTLY_IQ2_S || ftype == LLAMA_FTYPE_MOSTLY_IQ2_M) new_type = GGML_TYPE_IQ3_XXS;
else if (ftype == LLAMA_FTYPE_MOSTLY_IQ2_XL) new_type = GGML_TYPE_IQ3_XXS;
else if (ftype == LLAMA_FTYPE_MOSTLY_IQ3_XXS) new_type = GGML_TYPE_IQ3_S;
else if (new_type == GGML_TYPE_Q4_0_4_4 || new_type == GGML_TYPE_Q4_0_4_8 || new_type == GGML_TYPE_Q4_0_8_8) { else if (new_type == GGML_TYPE_Q4_0_4_4 || new_type == GGML_TYPE_Q4_0_4_8 || new_type == GGML_TYPE_Q4_0_8_8) {
new_type = GGML_TYPE_Q4_0; new_type = GGML_TYPE_Q4_0;
} }