diff --git a/README.md b/README.md
index e19f6c7..4a8ced2 100644
--- a/README.md
+++ b/README.md
@@ -1 +1,21 @@
-# CodeGen2
\ No newline at end of file
+# CodeGen2
+
+Official research release for the **CodeGen2** models (`3B`, `7B`, `16B`) for **Program Synthesis** as presented in ICLR 2023:
+
+*Title*: [CodeGen2: Lessons for Training LLMs on Programming and Natural Languages](https://arxiv.org/abs/TBD)
+
+*Authors*: [Erik Nijkamp](https://enijkamp.github.io/)\*, [Hiroaki Hayashi](https://hiroakih.me/)\*, [Silvio Savarese](https://scholar.google.com/citations?user=ImpbxLsAAAAJ&hl=en), [Caiming Xiong](https://scholar.google.com/citations?user=vaSdahkAAAAJ&hl=en), and [Yingbo Zhou](https://scholar.google.com/citations?user=H_6RQ7oAAAAJ&hl=en) (* indicates equal contribution)
+
+## Sampling
+
+Program synthesis in the form of auto-regressive sampling can be performed as follows:
+
+```python
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM
+tokenizer = AutoTokenizer.from_pretrained()
+model = AutoModelForCausalLM.from_pretrained()
+inputs = tokenizer("# this function prints hello world", return_tensors="pt").to(0)
+sample = model.generate(**inputs, max_length=128)
+print(tokenizer.decode(sample[0], truncate_before_pattern=[r"\n\n^#", "^'''", "\n\n\n"]))
+```