[BUG]: ValueError: The global train batch size (1 x 1) must be evenly divisible by the number of generations per prompt (8). Given the current train batch size, the valid values for the number of generations are: [].

[franztao]

Description of the bug

[WARNING|logging.py:329] 2025-03-14 16:34:32,585 >> Sliding Window Attention is enabled but not implemented for sdpa; unexpected results may be encountered.
Traceback (most recent call last):
File "/home/hengtao/miniconda3/envs/openmanus-rl/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/hengtao/miniconda3/envs/openmanus-rl/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/hengtao/code/OpenManus-RL/openmanus_rl/grpo.py", line 275, in
main(script_args, training_args, model_args)
File "/home/hengtao/code/OpenManus-RL/openmanus_rl/grpo.py", line 233, in main
trainer = GRPOTrainer(
File "/home/hengtao/miniconda3/envs/openmanus-rl/lib/python3.10/site-packages/trl/trainer/grpo_trainer.py", line 418, in init
raise ValueError(
ValueError: The global train batch size (1 x 1) must be evenly divisible by the number of generations per prompt (8). Given the current train batch size, the valid values for the number of generations are: [].

Steps To Reproduce

pls

Additional Information

No response