From 91f6499393d2d999331fbfdba47a7f8b9f913f0d Mon Sep 17 00:00:00 2001
From: Kerfuffle <44031344+KerfuffleV2@users.noreply.github.com>
Date: Thu, 16 Nov 2023 19:14:37 -0700
Subject: Respect tokenizer.ggml.add_bos_token value when tokenizing (#4040)

* gguf-py: gguf-dump: Respect --no-tensor flag in JSON mode.

* Respect add_bos_token GGUF metadata value

* gguf-py: Try to fix SpecialVocab giving up too easily for the Nth time
---
 common/common.cpp | 6 ++++++
 1 file changed, 6 insertions(+)

(limited to 'common/common.cpp')

diff --git a/common/common.cpp b/common/common.cpp
index 6a711420..e119317d 100644
--- a/common/common.cpp
+++ b/common/common.cpp
@@ -1072,6 +1072,12 @@ std::string llama_detokenize_bpe(llama_context * ctx, const std::vector<llama_to
     return result;
 }
 
+bool llama_should_add_bos_token(const llama_model * model) {
+    const int add_bos = llama_add_bos_token(model);
+
+    return add_bos != -1 ? bool(add_bos) : (llama_vocab_type(model) == LLAMA_VOCAB_TYPE_SPM);
+}
+
 //
 // YAML utils
 //
-- 
cgit v1.2.3