From abe0829984d0a4473628b90cf4418b752470de49 Mon Sep 17 00:00:00 2001
From: Matt Pulver <matt.pulver@heavy.ai>
Date: Fri, 25 Aug 2023 09:18:24 -0400
Subject: [PATCH] Add '// Beam search' heading to llama.{h,cpp} after
 llama_grammar_accept_token().

---
 llama.cpp | 4 ++++
 llama.h   | 5 +++++
 2 files changed, 9 insertions(+)

diff --git a/llama.cpp b/llama.cpp
index af950d3d5..1e4cf4055 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -4326,6 +4326,10 @@ void llama_grammar_accept_token(struct llama_context * ctx, struct llama_grammar
     ctx->t_sample_us += ggml_time_us() - t_start_sample_us;
 }
 
+//
+// Beam search
+//
+
 struct llama_beam {
     std::vector<llama_token> tokens;
     float p;  // Cumulative beam probability (renormalized relative to all beams)
diff --git a/llama.h b/llama.h
index e88a45078..81a27d438 100644
--- a/llama.h
+++ b/llama.h
@@ -465,6 +465,10 @@ extern "C" {
     /// @details Accepts the sampled token into the grammar
     LLAMA_API void llama_grammar_accept_token(struct llama_context * ctx, struct llama_grammar * grammar, llama_token token);
 
+    //
+    // Beam search
+    //
+
     struct llama_beam_view {
         llama_token const* tokens;
         size_t n_tokens;
@@ -482,6 +486,7 @@ extern "C" {
         size_t common_prefix_length;  // Current max length of prefix tokens shared by all beams.
         bool last_call;               // True iff this is the last callback invocation.
     };
+
     // Type of pointer to the beam_search_callback function.
     // void* callback_data is any custom data passed to llama_beam_search, that is subsequently
     // passed back to beam_search_callback. This avoids having to use global variables in the callback.