From 35862d19ec68924db9bf6dd34146146d869491a3 Mon Sep 17 00:00:00 2001
From: hiyouga <467089858@qq.com>
Date: Sun, 6 Oct 2024 10:33:11 +0800
Subject: [PATCH] fix #5611

Former-commit-id: 76c813d37c1d945a8bb6d3e4168e15fbe97c7a87
---
 src/llamafactory/model/patcher.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/llamafactory/model/patcher.py b/src/llamafactory/model/patcher.py
index 06d41af5..126e9723 100644
--- a/src/llamafactory/model/patcher.py
+++ b/src/llamafactory/model/patcher.py
@@ -110,6 +110,9 @@ def patch_config(
     if getattr(config, "model_type", None) == "qwen2" and is_trainable and model_args.flash_attn == "fa2":
         setattr(config, "use_cache", False)  # qwen2 does not support use_cache when using flash attn
 
+    if "LlavaLlamaForCausalLM" in getattr(config, "architectures", []):
+        raise ValueError("Please download llava models with hf-compatible format: https://huggingface.co/llava-hf")
+
     # deepspeed zero3 is not compatible with low_cpu_mem_usage
     init_kwargs["low_cpu_mem_usage"] = model_args.low_cpu_mem_usage and (not is_deepspeed_zero3_enabled())