Deactivate eval ACT on Aloha (policy is None)

.github/workflows/test.yml

@@ -154,15 +154,16 @@ jobs:
             device=cpu \
             policy.pretrained_model_path=tests/outputs/act/models/2.pt
 
-      - name: Test eval ACT on ALOHA end-to-end (policy is None)
-        run: |
-          source .venv/bin/activate
-          python lerobot/scripts/eval.py \
-            --config lerobot/configs/default.yaml \
-            policy=act \
-            env=aloha \
-            eval_episodes=1 \
-            device=cpu
+      # TODO(aliberts): This takes ~2mn to run, needs to be improved
+      # - name: Test eval ACT on ALOHA end-to-end (policy is None)
+      #   run: |
+      #     source .venv/bin/activate
+      #     python lerobot/scripts/eval.py \
+      #       --config lerobot/configs/default.yaml \
+      #       policy=act \
+      #       env=aloha \
+      #       eval_episodes=1 \
+      #       device=cpu
 
       - name: Test train Diffusion on PushT end-to-end
         run: |