compute_environment: LOCAL_MACHINE debug: false deepspeed_config: gradient_clipping: 1.0 offload_optimizer_device: cpu offload_param_device: cpu zero_stage: 2 distributed_type: DEEPSPEED gpu_ids: all machine_rank: 0 main_training_function: main num_machines: 1 num_processes: 8 rdzv_backend: static same_network: true tpu_env: [] tpu_use_cluster: false tpu_use_sudo: false use_cpu: false