fix grpo

Jintao-Huang · Jintao-Huang · commit 479fb102419a · 2025-11-29T23:52:29.000+08:00
diff --git a/swift/trainers/rlhf_trainer/grpo_trainer.py b/swift/trainers/rlhf_trainer/grpo_trainer.py
@@ -1351,7 +1351,7 @@ def _get_per_token_logps_and_entropies_single(self,
                 k: v
                 for k, v in inputs.items() if k not in [
                     'logits_to_keep', 'completion_mask', 'ref_per_token_logps', 'advantages', 'old_per_token_logps',
-                    'truncated_mask', 'num_items_in_batch'
+                    'truncated_mask', 'seq_lengths', 'num_items_in_batch'
                 ]
             }
             if 'logits_to_keep' in self.model_kwarg_keys:

Original file line number	Diff line number	Diff line change
`@@ -1351,7 +1351,7 @@ def _get_per_token_logps_and_entropies_single(self,`
`1351`	`1351`	`k: v`
`1352`	`1352`	`for k, v in inputs.items() if k not in [`
`1353`	`1353`	`'logits_to_keep', 'completion_mask', 'ref_per_token_logps', 'advantages', 'old_per_token_logps',`
`1354`		`- 'truncated_mask', 'num_items_in_batch'`
	`1354`	`+ 'truncated_mask', 'seq_lengths', 'num_items_in_batch'`
`1355`	`1355`	`]`
`1356`	`1356`	`}`
`1357`	`1357`	`if 'logits_to_keep' in self.model_kwarg_keys:`