update to include the KenLMModel example in README

README.md

@@ -2,7 +2,8 @@
 Compute surprisal from language models!
 
 `surprisal` supports most Causal Language Models (`GPT2`- and `GPTneo`-like models) from Huggingface or local checkpoint, 
-as well as `GPT3` models from OpenAI using their API!
+as well as `GPT3` models from OpenAI using their API! We also support `KenLM` N-gram based language models using the
+KenLM Python interface.
 
 Masked Language Models (`BERT`-like models) are in the pipeline and will be supported at a future time. 
 
@@ -12,6 +13,10 @@ The snippet below computes per-token surprisals for a list of sentences
 ```python
 from surprisal import AutoHuggingFaceModel
 
+from surprisal import KenLMModel
+k = KenLMModel(model_path='./literature.arpa')
+
+
 sentences = [
     "The cat is on the mat",
     "The cat is on the hat",
@@ -26,6 +31,9 @@ m.to('cuda') # optionally move your model to GPU!
 
 for result in m.surprise(sentences):
     print(result)
+
+for result in k.surprise(sentences):
+    print(result)
 ```
 and produces output of this sort:
 ```