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

fix loading rope.scaling.original_context_length from GGUF (#3)

Browse source 
Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
This commit is contained in:
Jeffrey Quesnelle 2023-10-30 08:35:51 -07:00 • committed by GitHub
parent 4d5ed8349d
commit 9fc823826e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

5
llama.cpp
View file

@ -8968,10 +8968,13 @@ struct llama_context * llama_new_context_with_model(
cparams.mul_mat_q = params.mul_mat_q; cparams.mul_mat_q = params.mul_mat_q;
cparams.n_ctx = params.n_ctx == 0 ? hparams.n_ctx_train : params.n_ctx; cparams.n_ctx = params.n_ctx == 0 ? hparams.n_ctx_train : params.n_ctx;
cparams.n_yarn_orig_ctx = params.yarn_orig_ctx == 0 ? hparams.n_ctx_train : params.yarn_orig_ctx;
cparams.rope_freq_base = params.rope_freq_base == 0.0f ? hparams.rope_freq_base_train : params.rope_freq_base; cparams.rope_freq_base = params.rope_freq_base == 0.0f ? hparams.rope_freq_base_train : params.rope_freq_base;
cparams.rope_freq_scale = params.rope_freq_scale == 0.0f ? hparams.rope_freq_scale_train : params.rope_freq_scale; cparams.rope_freq_scale = params.rope_freq_scale == 0.0f ? hparams.rope_freq_scale_train : params.rope_freq_scale;
cparams.n_yarn_orig_ctx = params.yarn_orig_ctx != 0 ? params.yarn_orig_ctx :
hparams.n_yarn_orig_ctx != 0 ? hparams.n_yarn_orig_ctx :
hparams.n_ctx_train;
auto rope_scaling_type = params.rope_scaling_type; auto rope_scaling_type = params.rope_scaling_type;
if (rope_scaling_type == LLAMA_ROPE_SCALING_UNSPECIFIED) { if (rope_scaling_type == LLAMA_ROPE_SCALING_UNSPECIFIED) {
rope_scaling_type = hparams.rope_scaling_type_train; rope_scaling_type = hparams.rope_scaling_type_train;