From 306c3cb6924c6231c102ff7d75aa3f68e3618ca2 Mon Sep 17 00:00:00 2001 From: Mitja Felicijan Date: Fri, 20 Feb 2026 13:54:21 +0100 Subject: Update to multi model for embeddings and prompting --- context.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'context.c') diff --git a/context.c b/context.c index e7aa0cf..66b8cc2 100644 --- a/context.c +++ b/context.c @@ -115,7 +115,10 @@ int main(int argc, char **argv) { cfg = &models[0]; } - struct llama_model *model = llama_model_load_from_file(cfg->filepath, llama_model_default_params()); + struct llama_model_params model_params = llama_model_default_params(); + model_params.n_gpu_layers = cfg->n_gpu_layers; + model_params.use_mmap = cfg->use_mmap; + struct llama_model *model = llama_model_load_from_file(cfg->filepath, model_params); if (model == NULL) { log_message(stderr, LOG_ERROR, "Unable to load embedding model"); llama_backend_free(); @@ -123,6 +126,8 @@ int main(int argc, char **argv) { } struct llama_context_params cparams = llama_context_default_params(); + cparams.n_ctx = cfg->n_ctx; + cparams.n_batch = cfg->n_batch; cparams.embeddings = true; struct llama_context *embed_ctx = llama_init_from_model(model, cparams); -- cgit v1.2.3