talk-llama : sync llama.cpp

2026-10-18 13:17:59 +02:00 · 2026-01-30 15:56:40 +02:00
parent b529c0610f
commit 953e503fd9
38 changed files with 2299 additions and 730 deletions
@@ -67,7 +67,7 @@ ggml_tensor * llm_build_nemotron_h::build_attention_layer(ggml_tensor *
                                                          const llama_model &       model,
                                                          const int64_t             n_embd_head,
                                                          const int                 il) {
-    // compute Q and K and (optionally) RoPE them
+    // compute Q and K
    ggml_tensor * Qcur = build_lora_mm(model.layers[il].wq, cur);
    cb(Qcur, "Qcur", il);
    if (model.layers[il].bq) {