diff --git a/README.md b/README.md
index 7e34319..c4f0f6e 100644
--- a/README.md
+++ b/README.md
@@ -32,7 +32,7 @@ Vectorized environments allow batching beam search planning and select actions i
 if you need to evaluate agent on number of episodes (or seeds) during training.  
 
 
-# Training
+## Training
 
 I trained it on D4RL medium datasets to validate that everything is OK. Scores seem to be very close to the original.
 Pretrained models are [available](pretrained).
@@ -44,7 +44,7 @@ Also, all datasets for [D4RL](https://sites.google.com/view/d4rl/home) gym tasks
 python scripts/train.py --config="configs/medium/halfcheetah_medium" --device="cuda" --seed="42"
 ```
 
-# Evaluation
+## Evaluation
 
 Available evaluation parameters can be seen in validation [config](configs/eval_vase.yaml).
 Here parameters are set to match evaluation configs from original implementation by [@jannerm](https://github.com/jannerm). 
@@ -77,7 +77,7 @@ python scripts/eval.py \
     beam_width=128
 ```
 
-# References
+## References
 ```
 @inproceedings{janner2021sequence,
   title = {Offline Reinforcement Learning as One Big Sequence Modeling Problem},
diff --git a/scripts/eval.py b/scripts/eval.py
index cf65178..057ef9e 100644
--- a/scripts/eval.py
+++ b/scripts/eval.py
@@ -25,8 +25,6 @@ def create_argparser():
 def run_experiment(config, seed, device):
     set_seed(seed=seed)
 
-    print(config)
-
     run_config = OmegaConf.load(os.path.join(config.checkpoints_path, "config.yaml"))
     discretizer = torch.load(os.path.join(config.checkpoints_path, "discretizer.pt"), map_location=device)