Update

sergiopaniego · sergiopaniego · commit de73f9202b54 · 2026-03-11T17:01:25.000+01:00
diff --git a/examples/scripts/bco.py b/examples/scripts/bco.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "einops",
 #     "scikit-learn",
 #     "joblib",
diff --git a/examples/scripts/cpo.py b/examples/scripts/cpo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/dpo_vlm.py b/examples/scripts/dpo_vlm.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "Pillow>=9.4.0",
 #     "torchvision",
 #     "trackio",
diff --git a/examples/scripts/gkd.py b/examples/scripts/gkd.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/grpo_agent.py b/examples/scripts/grpo_agent.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/grpo_vlm.py b/examples/scripts/grpo_vlm.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow",
-#     "peft",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "torchvision",
diff --git a/examples/scripts/gspo.py b/examples/scripts/gspo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "trackio",
diff --git a/examples/scripts/gspo_vlm.py b/examples/scripts/gspo_vlm.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow",
-#     "peft",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "torchvision",
diff --git a/examples/scripts/kto.py b/examples/scripts/kto.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/mpo_vlm.py b/examples/scripts/mpo_vlm.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow",
-#     "peft",
 #     "torchvision",
 #     "trackio",
 #     "kernels",
diff --git a/examples/scripts/online_dpo.py b/examples/scripts/online_dpo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/online_dpo_vlm.py b/examples/scripts/online_dpo_vlm.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "trackio",
diff --git a/examples/scripts/openenv/browsergym.py b/examples/scripts/openenv/browsergym.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl[vllm]",
-#     "peft",
+#     "trl[vllm,peft]",
 #     "trackio",
 #     "kernels",
 #     "openenv-browsergym @ git+https://huggingface.co/spaces/openenv/browsergym_env",
diff --git a/examples/scripts/openenv/browsergym_llm.py b/examples/scripts/openenv/browsergym_llm.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl[vllm]",
-#     "peft",
+#     "trl[vllm,peft]",
 #     "trackio",
 #     "kernels",
 #     "openenv-browsergym @ git+https://huggingface.co/spaces/openenv/browsergym_env",
@@ -291,10 +290,26 @@ def __init__(self):
             self._done = False
             self._step_count = 0
 
+        def _ensure_large_max_size(self):
+            """Raise WebSocket max message size for large observations (e.g. accessibility trees).
+
+            openenv-core<=0.2.1 does not pass max_size to ws_connect, so the websockets library
+            defaults to 1MB. We force a connection and patch it to 100MB before any messages are sent.
+            """
+            self.client.connect()
+            ws = self.client._ws
+            if ws is not None and hasattr(ws, "protocol"):
+                proto = ws.protocol
+                # websockets <16: max_size; websockets >=16: max_message_size
+                attr = "max_size" if hasattr(proto, "max_size") else "max_message_size"
+                if getattr(proto, attr) == 2**20:
+                    setattr(proto, attr, 100 * 1024 * 1024)
+
         def reset(self, **kwargs) -> str:
             self.reward = 0.0
             self._done = False
             self._step_count = 0
+            self._ensure_large_max_size()
             result = self.client.reset()
             self._done = result.done
             return self._format_observation(result.observation)
diff --git a/examples/scripts/openenv/catch.py b/examples/scripts/openenv/catch.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl[vllm]",
-#     "peft",
+#     "trl[vllm,peft]",
 #     "trackio",
 #     "kernels",
 #     "openenv-openspiel-env @ git+https://huggingface.co/spaces/openenv/openspiel_env",
diff --git a/examples/scripts/openenv/sudoku.py b/examples/scripts/openenv/sudoku.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl[vllm]",
-#     "peft",
+#     "trl[vllm,peft]",
 #     "trackio",
 #     "kernels",
 #     "openenv-textarena @ git+https://huggingface.co/spaces/openenv/sudoku",
@@ -131,7 +130,7 @@ def parse_args() -> argparse.Namespace:
         "--difficulty",
         type=str,
         choices=["easy", "medium", "hard"],
-        default="hard",
+        default="easy",
         help="Training difficulty: easy=guaranteed+options, medium=only options, hard=no hints",
     )
     parser.add_argument(
@@ -178,7 +177,7 @@ def parse_args() -> argparse.Namespace:
     # vLLM
     parser.add_argument("--vllm-mode", choices=("colocate", "server"), default="colocate")
     parser.add_argument("--vllm-server-url", type=str, default="http://localhost:8000")
-    parser.add_argument("--vllm-gpu-memory-utilization", type=float, default=0.1)
+    parser.add_argument("--vllm-gpu-memory-utilization", type=float, default=0.15)
 
     return parser.parse_args()
 
diff --git a/examples/scripts/orpo.py b/examples/scripts/orpo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/ppo/ppo.py b/examples/scripts/ppo/ppo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/ppo/ppo_tldr.py b/examples/scripts/ppo/ppo_tldr.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/rloo.py b/examples/scripts/rloo.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl[vllm]",
-#     "peft",
+#     "trl[vllm,peft]",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "trackio",
diff --git a/examples/scripts/rloo_vlm.py b/examples/scripts/rloo_vlm.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow",
-#     "peft",
 #     "math-verify",
 #     "latex2sympy2_extended",
 #     "torchvision",
diff --git a/examples/scripts/sft_video_llm.py b/examples/scripts/sft_video_llm.py
@@ -14,8 +14,7 @@
 
 # /// script
 # dependencies = [
-#     "trl",
-#     "peft",
+#     "trl[peft]",
 #     "qwen-vl-utils",
 #     "torchvision",
 #     "bitsandbytes",
diff --git a/examples/scripts/sft_vlm.py b/examples/scripts/sft_vlm.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow>=9.4.0",
-#     "peft",
 #     "trackio",
 #     "kernels",
 # ]
diff --git a/examples/scripts/sft_vlm_gemma3.py b/examples/scripts/sft_vlm_gemma3.py
@@ -14,9 +14,8 @@
 
 # /// script
 # dependencies = [
-#     "trl",
+#     "trl[peft]",
 #     "Pillow>=9.4.0",
-#     "peft",
 #     "trackio",
 #     "kernels",
 # ]
