Update README.md

README.md

@@ -28,7 +28,7 @@ These checkpoints use the same architecture and starting checkpoint as the offic
 ## Inference
 You can access these checkpoints using the standard Hugging Face Transformers library:
 
-```
+```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
 olmo_early_training = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0425-1B-early-training")
 tokenizer = AutoTokenizer.from_pretrained("allenai/OLMo-2-0425-1B-early-training")
@@ -40,7 +40,7 @@ print(tokenizer.batch_decode(response, skip_special_tokens=True)[0])
 ```
 
 To access a specific checkpoint, you can specify the revision:
-```
+```python
 olmo_early_training = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0425-1B-early-training", revision="stage1-step20000-tokens42B")
 ```