Update README.md

README.md

@@ -10,6 +10,10 @@ language:
 
 ## Inference example
 ```python
+!pip install -U transformers
+!pip install git+https://github.com/huggingface/peft
+!pip install -U bitsandbytes
+
 import torch
 from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 
@@ -22,6 +26,12 @@ quantization_config = BitsAndBytesConfig(
 model = AutoModelForCausalLM.from_pretrained(repo_name, quantization_config=quantization_config, device_map='cuda')
 tokenizer = AutoTokenizer.from_pretrained(repo_name)
 
+pipe = pipeline(
+    'text-generation',
+    model=model,
+    tokenizer=tokenizer,
+    device_map='cuda',
+)
 
 def ask(x, context='', is_input_full=False):
     ans = pipe(