From d6f74975a40beba94ed6f55c964c35aae66f5314 Mon Sep 17 00:00:00 2001
From: Radek Pilar <github@mrkva.eu>
Date: Tue, 12 Dec 2023 17:05:37 +0100
Subject: [PATCH] convert : use sentencepiece tokenizer for Mixtral-instruct

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

diff --git a/convert-hf-to-gguf.py b/convert-hf-to-gguf.py
index 770e10eb9..40358409f 100755
--- a/convert-hf-to-gguf.py
+++ b/convert-hf-to-gguf.py
@@ -180,6 +180,8 @@ class Model:
             return StableLMModel
         if model_architecture == "QWenLMHeadModel":
             return QwenModel
+        if model_architecture == "MixtralForCausalLM":
+            return MixtralModel
         return Model
 
     def _is_model_safetensors(self) -> bool:
@@ -848,6 +850,11 @@ class StableLMModel(Model):
         self.gguf_writer.add_parallel_residual(hparams["use_parallel_residual"] if "use_parallel_residual" in hparams else True)
         self.gguf_writer.add_layer_norm_eps(1e-5)
 
+class MixtralModel(Model):
+    def set_vocab(self):
+        self._set_vocab_sentencepiece()
+
+
 
 class QwenModel(Model):
     @staticmethod