llama : rework embeddings logic (#14208)

* llama : rework embeddings logic ggml-ci * cont : fix rerank ggml-ci * cont : engrish [no ci] * cont : fix rerank ggml-ci * server : support both embeddings and completions with single model ggml-ci * cont : avoid embeddings_org ggml-ci
2025-08-06 01:05:03 -04:00 · 2025-06-16 14:14:00 +03:00
parent 3ba0d843c6
commit d3e64b9f49
16 changed files with 159 additions and 114 deletions
--- a/src/llama-batch.h
+++ b/src/llama-batch.h
@@ -88,7 +88,8 @@ public:
    bool init(
            const llama_batch & batch_inp,
            const llama_vocab & vocab,
-            const llama_memory_i * memory);
+            const llama_memory_i * memory,
+            bool embd_all);

    const llama_batch & get_batch() const;