Update model.py

model.py

@@ -4,7 +4,7 @@ from typing import Iterator
 import torch
 from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
 
-model_id = 'codellama/CodeLlama-13b-Instruct-hf'
+model_id = 'TheBloke/CodeLlama-7B-GGML'
 
 if torch.cuda.is_available():
     config = AutoConfig.from_pretrained(model_id)
@@ -22,7 +22,7 @@ else:
     torch.set_default_tensor_type(torch.FloatTensor)
     model = AutoModelForCausalLM.from_pretrained(
         model_id, 
-        torch_dtype=torch.float32,
+        torch_dtype=torch.float16,
         device_map='cpu') 
     
 tokenizer = AutoTokenizer.from_pretrained(model_id)