Update app.py

app.py

@@ -1760,9 +1760,9 @@ def launch():
 
         if QUANTIZATION == 'awq':
             print(F'Load model in int4 quantization')
-            llm = LLM(model=model_path, dtype="float16", tensor_parallel_size=tensor_parallel, gpu_memory_utilization=gpu_memory_utilization, quantization="awq")
+            llm = LLM(model=model_path, dtype="float16", tensor_parallel_size=tensor_parallel, gpu_memory_utilization=gpu_memory_utilization, quantization="awq", max_model_len=8192)
         else:
-            llm = LLM(model=model_path, dtype=dtype, tensor_parallel_size=tensor_parallel, gpu_memory_utilization=gpu_memory_utilization)
+            llm = LLM(model=model_path, dtype=dtype, tensor_parallel_size=tensor_parallel, gpu_memory_utilization=gpu_memory_utilization, max_model_len=8192)
 
         try:
             print(llm.llm_engine.workers[0].model)