diff --git a/environments/eval_environments/gsm8k_eval.py b/environments/eval_environments/gsm8k_eval.py
index 8dec68a5..85faedd9 100644
--- a/environments/eval_environments/gsm8k_eval.py
+++ b/environments/eval_environments/gsm8k_eval.py
@@ -133,10 +133,10 @@ class GSM8KEvalEnv(BaseEnv):
         self,
         config: GSM8KEvalConfig,
         server_configs: List[APIServerConfig],
-        slurm_job_id: Optional[str] = None,
+        slurm=False,
         testing: bool = False,
     ):
-        super().__init__(config, server_configs, slurm_job_id, testing)
+        super().__init__(config, server_configs, slurm, testing)
         self.config: GSM8KEvalConfig = config
         self.eval_items: List[Dict] = []
         self._dataset_loaded = False
diff --git a/example_trainer/grpo.py b/example_trainer/grpo.py
index 73201600..9cd935ad 100644
--- a/example_trainer/grpo.py
+++ b/example_trainer/grpo.py
@@ -679,12 +679,15 @@ def _attach_to_vllm_shared_tensors(
     # This catches obvious mapping failures before we try to load
     # =========================================================================
     hf_param_count = len(list(model.named_parameters()))
+    # Note: attached_count may include fused tensors that map to multiple HF params
+    # So coverage can exceed 100% - that's OK
     mapping_coverage = attached_count / hf_param_count if hf_param_count > 0 else 0
 
-    print(f"[Setup] Mapping coverage: {attached_count}/{hf_param_count} ({mapping_coverage:.1%})")
+    print(f"[Setup] Mapping coverage: {attached_count} tensors for {hf_param_count} parameters")
 
     # Expect at least 90% coverage for a valid mapping
-    # Some params like inv_freq buffers won't be in vLLM
+    # Note: with fused tensors, we may have MORE mappings than params
+    # So we check if we have at least 90% of params covered
     if mapping_coverage < 0.90:
         unmapped_params = set(model.state_dict().keys()) - set(hf_state_dict.keys())
         warning_msg = f"[Setup] WARNING: Low mapping coverage ({mapping_coverage:.1%})\n"
@@ -701,6 +704,8 @@ def _attach_to_vllm_shared_tensors(
                 "  2. tensor-parallel-size=1 for single-copy mode\n"
                 "  3. vllm_bridge_config.json contains valid ipc_handles"
             )
+    else:
+        print(f"[Setup] ✓ Good mapping coverage ({mapping_coverage:.1%})")
 
     print(f"[Setup] ✓ Attached {attached_count} tensors to vLLM's shared memory")