From 3aa184a8c7c553b5dfcc142d919f3db695df297a Mon Sep 17 00:00:00 2001
From: 0xspringtime <110655352+0xspringtime@users.noreply.github.com>
Date: Sat, 22 Jun 2024 09:37:41 -0400
Subject: [PATCH] convert-hf : change assert to exception (#8015)

---
 convert-hf-to-gguf.py | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/convert-hf-to-gguf.py b/convert-hf-to-gguf.py
index a6751cc80..166e5ded2 100755
--- a/convert-hf-to-gguf.py
+++ b/convert-hf-to-gguf.py
@@ -967,7 +967,13 @@ class XverseModel(Model):
         from transformers import AutoTokenizer
         tokenizer = AutoTokenizer.from_pretrained(dir_model)
         vocab_size = hparams.get("vocab_size", len(tokenizer.vocab))
-        assert max(tokenizer.vocab.values()) < vocab_size
+        # Since we are checking the maximum index, we need to ensure it's strictly less than vocab_size,
+        # because vocab_size is the count of items, and indexes start at 0.
+        max_vocab_index = max(tokenizer.get_vocab().values())
+        if max_vocab_index >= vocab_size:
+            raise ValueError("Vocabulary size exceeds expected maximum size.")
+
+
 
         reverse_vocab: dict[int, str] = {id_: encoded_tok for encoded_tok, id_ in tokenizer.vocab.items()}
         added_vocab = tokenizer.get_added_vocab()