13b model

usage.py

@@ -13,7 +13,7 @@ import torch
 
 #import ipdb; ipdb.set_trace()
 cache_dir="/workspace/asr/peft/qlora" 
-# TODO set this cache_dir to the path where you stored (or, want to store) llama2-7bhf model
+# TODO set this cache_dir to the path where you stored (or, want to store) llama2-13b-hf model
 
 lazylora_dir=os.getcwd() # the path that contains 'adapter_config.json' and 'adapter_model.bin'
 
@@ -41,7 +41,7 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 #model.print_trainable_parameters()
 print(sum(p.numel() for p in model.parameters()))
-# 6,671,979,520 -> half-size of 7B due to 4-bit loading
+# 6,671,979,520 -> half-size of 13B due to 4-bit loading
 
 model = PeftModel.from_pretrained(model, lazylora_dir)
 print('after adding lazy lora parameters:')