update readme

Author: WendyH1108

README.md

@@ -1 +1,23 @@
 # SnapKV
+## Quick Start
+### Use SnapKV-optimized Models
+SnapKV-optimized models are all under models file, which could be directly imported and used the same like baseline models.
+For example: 
+```python
+from models.modeling_mistral import MistralForCausalLM as SnapKVMistralForCausalLM
+model = SnapKVMistralForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16,
+    low_cpu_mem_usage=True,
+    device_map="auto",
+    use_flash_attention_2=True
+)
+tokenizer = transformers.AutoTokenizer.from_pretrained(
+    model_name,
+    padding_side="right",
+    use_fast=False,
+)
+```
+
+### Customize Your SnapKV-optimized Models
+SnapKV can be easily integrate with other models. You can follow the comment marked with `[SnapKV]` in [existing models](./models) to constrcut your own models. The algorithm of SnapKV is in [snapkv_utils.py](./snapkv_utils.py)