Update README.md

README.md

@@ -4,8 +4,20 @@ language:
 - zh
 pipeline_tag: text-generation
 widget:
-- text: "生命、宇宙以及一切的终极答案是"
+- text: "中国的首都是"
+
 ---
-The model is based on [bigscience/bloom-1b7](https://huggingface.co/bigscience/bloom-1b7). 
 
-To reduce GPU memory usage, we pruned its vocabulary from 250880 to 46145 with Chinese corpus. So the total parameter is 1.4b now.
+This model is based on [bigscience/bloom-1b7](https://huggingface.co/bigscience/bloom-1b7). 
+
+We pruned its vocabulary from 250880 to 46145 with Chinese corpus to reduce GPU memory usage. So the total parameter is 1.4b now.
+
+# How to use
+```python
+from transformers import BloomTokenizerFast, BloomForCausalLM
+
+tokenizer = BloomTokenizerFast.from_pretrained('Langboat/bloom-1.4b-zh')
+model = BloomForCausalLM.from_pretrained('Langboat/bloom-1.4b-zh')
+
+print(tokenizer.batch_decode(model.generate(tokenizer.encode('中国的首都是', return_tensors='pt'))))
+```