Adjusted the batch size to enable llama2-70b to run on v5p-128

Author: hsuan-lun-chiang

benchmarks/maxtext_v5p_model_configs.py

@@ -202,7 +202,7 @@
         model_type="llama2-70b",
         tuning_params={
             "ici_fsdp_parallelism": -1,
-            "per_device_batch_size": 4,
+            "per_device_batch_size": 2,
             "remat_policy": "save_dot_except_mlpwi",
             "max_target_length": 4096,
             "use_iota_embed": True,