Upload folder using huggingface_hub

code/autoslm/engine/disaggregated.py

@@ -276,8 +276,14 @@ def build_accelerate_launch_cmd(
             "FULL_SHARD",
             "--fsdp_auto_wrap_policy",
             "TRANSFORMER_BASED_WRAP",
+            # FULL_STATE_DICT (not SHARDED): transformers' Trainer rejects save_only_model — which
+            # GRPOConfig sets — alongside SHARDED_STATE_DICT ("save_only_model is not compatible with
+            # FSDP state dict type 'SHARDED_STATE_DICT'"). FULL gathers the (small LoRA) adapter on
+            # rank 0 at save time, which is what trainer.save_model needs anyway. Fine for the dense
+            # 1-9B models here; a 70B+ base would want SHARDED + save_only_model off, but those route
+            # to TP inference, not a sharded full-state save.
             "--fsdp_state_dict_type",
-            "SHARDED_STATE_DICT",
+            "FULL_STATE_DICT",
         ]
     else:
         # Plain DDP across the train GPUs (replicate, no sharding) — needs the explicit --multi_gpu.

code/autoslm/engine/worker.py

@@ -1514,6 +1514,11 @@ def run_rl():
     )
     _rollout_split = None
     _disagg_base_env: dict | None = None
+    # Defaults for the colocate (inference_gpus==0) and train_gpus==1 paths — these MUST be defined
+    # before the disaggregated branch so the trainer-build block's `if _is_fsdp_launcher:` never hits
+    # an UnboundLocalError on the colocate path (only the train_gpus>1 launcher sets it True).
+    _trainer_only = False
+    _is_fsdp_launcher = False
     if _inference_gpus > 0:
         _total_gpus = _disagg.detect_total_gpus()
         _rollout_split = select_rollout_split(_total_gpus, _inference_gpus)