diff --git "a/attnserver.run_attnserver.slurm.sh.343248.out.log" "b/attnserver.run_attnserver.slurm.sh.343248.out.log" new file mode 100644--- /dev/null +++ "b/attnserver.run_attnserver.slurm.sh.343248.out.log" @@ -0,0 +1,4621 @@ +Running ctx_length=1024, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 1024 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +INFO:megatron.training.initialize:Setting logging level to 0 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 1024 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 1024 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 1024 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.044 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.458 seconds +time to initialize megatron (seconds): 7.523 +[after megatron is initialized] datetime: 2025-06-21 22:03:32 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 283719680 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer +>>> embedding>>> embedding + +>>> decoder +>>> decoder>>> output_layer + +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 283719680 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 283719680 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 283719680 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 283719680 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 283719680 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 283719680 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 283719680 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (283719680 elements, 283719680 padded size): + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.bias + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight + module.decoder.final_layernorm.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.layers.0.mlp.linear_fc1.bias + module.embedding.position_embeddings.weight + module.embedding.word_embeddings.weight + module.decoder.final_layernorm.weight + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.decoder.layers.0.self_attention.linear_proj.weight +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (10.70, 10.74) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:03:33 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=1024, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.006474 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 66592 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.003639 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 66562 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.003403 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 66686 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:03:33 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (251.15, 256.62) + train/valid/test-data-iterators-setup ..........: (33.52, 153.60) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:03:33 +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor: tokens torch.Size([32, 32768]) +batch tensor: labels torch.Size([32, 32768]) +batch tensor: loss_mask torch.Size([32, 32768]) +batch tensor: attention_mask torch.Size([32, 1, 32768, 32768]) +batch tensor: position_ids torch.Size([32, 32768]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +batch tensor after cp: tokens torch.Size([32, 8192]) +batch tensor after cp: labels torch.Size([32, 8192]) +batch tensor after cp: loss_mask torch.Size([32, 8192]) +batch tensor after cp: attention_mask torch.Size([32, 1, 8192, 32768]) +batch tensor after cp: position_ids torch.Size([32, 8192]) +Start exporting trace 0 +Done exporting trace 0 +Running ctx_length=2048, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 2048 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 2048 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 2048 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 2048 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.053 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.135 seconds +time to initialize megatron (seconds): 7.443 +[after megatron is initialized] datetime: 2025-06-21 22:04:26 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer +>>> embedding>>> embedding + +>>> decoder>>> decoder + +>>> output_layer>>> output_layer + + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 287913984 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 287913984 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 287913984 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 287913984 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 287913984>>> decoder + +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 287913984 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 287913984 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 287913984 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (287913984 elements, 287913984 padded size): + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.weight + module.embedding.word_embeddings.weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.embedding.position_embeddings.weight + module.decoder.final_layernorm.bias + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.final_layernorm.weight + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.bias + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.bias +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (3.83, 3.85) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:04:27 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=2048, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.005495 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 33296 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.002633 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 33281 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.002547 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 33343 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:04:27 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (321.54, 322.77) + train/valid/test-data-iterators-setup ..........: (42.51, 167.04) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:04:27 +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +batch tensor: tokens torch.Size([32, 65536]) +batch tensor: labels torch.Size([32, 65536]) +batch tensor: loss_mask torch.Size([32, 65536]) +batch tensor: attention_mask torch.Size([32, 1, 65536, 65536]) +batch tensor: position_ids torch.Size([32, 65536]) +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 32.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 293, in get_batch\n batch = get_batch_on_this_cp_rank(batch)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/utils.py", line 1765, in get_batch_on_this_cp_rank\n val = val.index_select(seq_dim, index)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 8.10 GiB is free. Including non-PyTorch memory, this process has 131.70 GiB memory in use. Of the allocated memory 130.21 GiB is allocated by PyTorch, and 37.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +Running ctx_length=4096, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 4096 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +INFO:megatron.training.initialize:Setting logging level to 0 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 4096 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 4096 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 4096 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.046 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.363 seconds +time to initialize megatron (seconds): 7.259 +[after megatron is initialized] datetime: 2025-06-21 22:05:03 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 296302592 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 296302592 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding>>> embedding>>> embedding + + +>>> decoder +>>> decoder +>>> output_layer>>> decoder>>> output_layer + + +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 296302592 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 296302592 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 296302592 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 296302592 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (296302592 elements, 296302592 padded size): + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.0.self_attention.linear_proj.weight + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.layers.0.mlp.linear_fc1.bias + module.decoder.final_layernorm.bias + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.embedding.word_embeddings.weight + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.embedding.position_embeddings.weight + module.decoder.final_layernorm.weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.bias +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 296302592 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 296302592 +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (3.04, 3.07) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:05:04 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=4096, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.005184 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 16648 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.002095 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 16640 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001998 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 16671 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:05:04 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (402.81, 426.13) + train/valid/test-data-iterators-setup ..........: (33.22, 152.86) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:05:04 +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 512.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.94 GiB is free. Including non-PyTorch memory, this process has 3.86 GiB memory in use. Of the allocated memory 2.29 GiB is allocated by PyTorch, and 119.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +Running ctx_length=8192, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 8192 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 8192 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 8192 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 8192 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.047 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.203 seconds +time to initialize megatron (seconds): 7.195 +[after megatron is initialized] datetime: 2025-06-21 22:05:39 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 313079808 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 313079808 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 313079808 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 313079808 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 313079808 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 313079808 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 313079808 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 313079808 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (313079808 elements, 313079808 padded size): + module.decoder.final_layernorm.bias + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.0.self_attention.linear_proj.weight + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.final_layernorm.weight + module.decoder.layers.0.mlp.linear_fc1.bias + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.embedding.position_embeddings.weight + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.embedding.word_embeddings.weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.bias + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (23.47, 23.51) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:05:40 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=8192, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.005350 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 8324 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001850 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 8320 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001714 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 8335 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:05:40 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (563.78, 565.59) + train/valid/test-data-iterators-setup ..........: (34.30, 168.04) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:05:40 +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 2048.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.73 GiB is free. Including non-PyTorch memory, this process has 4.08 GiB memory in use. Of the allocated memory 2.49 GiB is allocated by PyTorch, and 135.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +Running ctx_length=12288, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 12288 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 12288 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 12288 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 12288 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.086 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.275 seconds +time to initialize megatron (seconds): 7.353 +[after megatron is initialized] datetime: 2025-06-21 22:06:15 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> embedding>>> decoder + +>>> output_layer +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 329857024 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 329857024 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 329857024 +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 329857024 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 329857024 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 329857024 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 329857024 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 329857024 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (329857024 elements, 329857024 padded size): + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.decoder.final_layernorm.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.bias + module.decoder.final_layernorm.weight + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.bias + module.embedding.position_embeddings.weight + module.embedding.word_embeddings.weight + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.weight +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (2.63, 2.75) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:06:16 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=12288, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.004793 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 5549 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001871 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 5546 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001747 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 5557 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:06:16 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (723.16, 740.31) + train/valid/test-data-iterators-setup ..........: (23.39, 152.84) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:06:16 +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4608.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.55 GiB is free. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 103.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +Running ctx_length=16384, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 16384 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 16384 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 16384 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 16384 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.046 seconds +> compiling and loading fused kernels ... +>>> done with compiling and loading fused kernels. Compilation time: 2.425 seconds +time to initialize megatron (seconds): 7.485 +[after megatron is initialized] datetime: 2025-06-21 22:06:52 +building GPT model ... +>>> embedding +>>> decoder +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 346634240 + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 346634240 +>>> embedding>>> embedding + +>>> decoder>>> decoder + +>>> output_layer +>>> output_layer +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 346634240 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 346634240 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 346634240 + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 346634240 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (1, 0): 346634240 +>>> embedding +>>> decoder +>>> output_layer + > number of parameters on (tensor, pipeline) model parallel rank (0, 0): 346634240 +INFO:megatron.core.distributed.distributed_data_parallel:Setting up DistributedDataParallel with config DistributedDataParallelConfig(grad_reduce_in_fp32=False, overlap_grad_reduce=False, overlap_param_gather=False, align_param_gather=False, use_distributed_optimizer=False, num_distributed_optimizer_instances=1, check_for_nan_in_grad=False, check_for_large_grads=False, bucket_size=None, pad_buckets_for_high_nccl_busbw=False, average_in_collective=False, fp8_param_gather=False, use_custom_fsdp=False, data_parallel_sharding_strategy='no_shard', gradient_reduce_div_fusion=True, suggested_communication_unit_size=None, preserve_fp32_weights=True, keep_fp8_transpose_cache_when_using_custom_fsdp=False, nccl_ub=False, fsdp_double_buffer=False) +INFO:megatron.core.distributed.param_and_grad_buffer:Number of buckets for gradient all-reduce / reduce-scatter: 1 +Params for bucket 1 (346634240 elements, 346634240 padded size): + module.decoder.final_layernorm.bias + module.decoder.layers.1.self_attention.linear_qkv.weight + module.decoder.layers.1.self_attention.linear_proj.weight + module.decoder.layers.0.self_attention.linear_qkv.weight + module.embedding.word_embeddings.weight + module.decoder.layers.1.mlp.linear_fc2.weight + module.decoder.layers.1.self_attention.linear_proj.bias + module.decoder.final_layernorm.weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.mlp.linear_fc2.weight + module.decoder.layers.0.mlp.linear_fc1.layer_norm_bias + module.decoder.layers.0.self_attention.linear_proj.weight + module.decoder.layers.1.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.bias + module.decoder.layers.0.mlp.linear_fc2.bias + module.decoder.layers.0.mlp.linear_fc1.layer_norm_weight + module.decoder.layers.0.self_attention.linear_qkv.bias + module.decoder.layers.0.self_attention.linear_proj.bias + module.decoder.layers.1.mlp.linear_fc1.weight + module.decoder.layers.0.mlp.linear_fc1.weight + module.decoder.layers.1.mlp.linear_fc2.bias + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_weight + module.decoder.layers.1.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_bias + module.decoder.layers.0.self_attention.linear_qkv.layer_norm_weight + module.embedding.position_embeddings.weight + module.decoder.layers.1.mlp.linear_fc1.bias + module.decoder.layers.0.mlp.linear_fc1.bias +INFO:megatron.core.optimizer:Setting up optimizer with config OptimizerConfig(optimizer='adam', lr=0.0005, min_lr=0.0, decoupled_lr=None, decoupled_min_lr=None, weight_decay=0.1, fp16=True, bf16=False, params_dtype=torch.float16, use_precision_aware_optimizer=False, store_param_remainders=True, main_grads_dtype=torch.float32, main_params_dtype=torch.float32, exp_avg_dtype=torch.float32, exp_avg_sq_dtype=torch.float32, loss_scale=None, initial_loss_scale=4294967296, min_loss_scale=1.0, loss_scale_window=1000, hysteresis=2, adam_beta1=0.9, adam_beta2=0.999, adam_eps=1e-08, sgd_momentum=0.9, use_distributed_optimizer=False, overlap_param_gather_with_optimizer_step=False, optimizer_cpu_offload=False, optimizer_offload_fraction=1.0, use_torch_optimizer_for_cpu_offload=False, overlap_cpu_optimizer_d2h_h2d=False, pin_cpu_grads=True, pin_cpu_params=True, clip_grad=1.0, log_num_zeros_in_grad=False, barrier_with_L1_time=True, timers=, config_logger_dir='') +INFO:megatron.core.optimizer_param_scheduler:> learning rate decay style: cosine +WARNING: could not find the metadata file gpt-checkpoint/latest_checkpointed_iteration.txt + will not load any checkpoints and will start from random +(min, max) time across ranks (ms): + load-checkpoint ................................: (2.84, 2.90) +[after model, optimizer, and learning rate scheduler are built] datetime: 2025-06-21 22:06:53 +> building train, validation, and test datasets ... + > datasets target sizes (minimum size): + train: 10 + validation: 1 + test: 1 +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let mock = True, as both blend and blend_per_split are None +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split = 1,1,1, an arbitrarily even split, as mock is True +INFO:megatron.core.datasets.blended_megatron_dataset_config:Let split_matrix = [(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)] +> building train, validation, and test datasets for GPT ... +INFO:megatron.core.datasets.blended_megatron_dataset_builder:Building MockGPTDataset splits with sizes=(10, 1, 1) and config=GPTDatasetConfig(random_seed=1234, sequence_length=16384, blend=None, blend_per_split=None, split='1,1,1', split_matrix=[(0, 0.3333333333333333), (0.3333333333333333, 0.6666666666666666), (0.6666666666666666, 1.0)], num_dataset_builder_threads=1, path_to_cache=None, mmap_bin_files=True, mock=True, tokenizer=, mid_level_dataset_surplus=0.005, reset_position_ids=False, reset_attention_mask=False, eod_mask_loss=False, create_attention_mask=True, drop_last_partial_validation_sequence=True, add_extra_token_to_sequence=True, object_storage_cache_path=None) +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset train indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.004614 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 4162 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset valid indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001692 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 4160 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +INFO:megatron.core.datasets.gpt_dataset:Build and save the MockGPTDataset test indices +DEBUG:megatron.core.datasets.gpt_dataset:> separate_final_epoch: False +WARNING:megatron.core.datasets.gpt_dataset:Unable to save MockGPTDataset indexes because path_to_cache is None +DEBUG:megatron.core.datasets.gpt_dataset: > time elapsed: 0.001473 seconds +INFO:megatron.core.datasets.gpt_dataset:> total number of samples: 4167 +INFO:megatron.core.datasets.gpt_dataset:> total number of epochs: 1 +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2025-06-21 22:06:53 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (863.57, 875.80) + train/valid/test-data-iterators-setup ..........: (31.59, 152.21) +training ... +Setting rerun_state_machine.current_iteration to 0... +[before the start of training step] datetime: 2025-06-21 22:06:53 +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 5 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 3 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 0 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 6 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 1 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 4 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 7 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +WARNING:megatron.core.utils:CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +['Traceback (most recent call last):\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 446, in forward_step\n (tokens, labels, loss_mask, attention_mask, position_ids), token_lens = get_batch(data_iterator)\n ^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 284, in get_batch\n batch = next(global_batches)\n ^^^^^^^^^^^^^^^^^^^^\n', ' File "/mnt/weka/home/hao.zhang/junda/attnserver-megatron/./pretrain_gpt_profile.py", line 226, in setup_batches\n attention_mask = torch.ones(\n ^^^^^^^^^^^\n', 'torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 8192.00 GiB. GPU 2 has a total capacity of 139.81 GiB of which 135.36 GiB is free. Including non-PyTorch memory, this process has 4.45 GiB memory in use. Of the allocated memory 2.90 GiB is allocated by PyTorch, and 97.54 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)\n'] +Running ctx_length=24576, TP_SIZE=2, CP_SIZE=4, BATCH_SIZE=32 +Cleaning up checkpoint directory: gpt-checkpoint +-------------------------------- +CTX_LENGTH: 24576 +TP_SIZE: 2 +CP_SIZE: 4 +CHECKPOINT_PATH: gpt-checkpoint +PWD: /mnt/weka/home/hao.zhang/junda/attnserver-megatron +-------------------------------- +/mnt/weka/home/hao.zhang/conda/miniconda/envs/junda-attnserver/bin/python3 +WARNING: TensorBoard writing requested but is not available (are you using PyTorch 1.1.0 or later?), no TensorBoard logs will be written. +WARNING: one_logger package is required to enable e2e metrics tracking. please go to https://confluence.nvidia.com/display/MLWFO/Package+Repositories for details to install it +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +using world size: 8, data-parallel size: 1, context-parallel size: 4, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 2, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0 +Number of virtual stages per pipeline stage: None +WARNING: Setting args.check_for_nan_in_loss_and_grad to False since dynamic loss scaling is being used +using torch.float16 for parameters ... +------------------------ arguments ------------------------ + account_for_embedding_in_pipeline_split ......... False + account_for_loss_in_pipeline_split .............. False + accumulate_allreduce_grads_in_fp32 .............. False + adam_beta1 ...................................... 0.9 + adam_beta2 ...................................... 0.999 + adam_eps ........................................ 1e-08 + add_bias_linear ................................. True + add_position_embedding .......................... True + add_qkv_bias .................................... True + adlr_autoresume ................................. False + adlr_autoresume_interval ........................ 1000 + align_grad_reduce ............................... True + align_param_gather .............................. False + app_tag_run_name ................................ None + app_tag_run_version ............................. 0.0.0 + apply_layernorm_1p .............................. False + apply_query_key_layer_scaling ................... False + apply_residual_connection_post_layernorm ........ False + apply_rope_fusion ............................... False + async_save ...................................... None + async_tensor_model_parallel_allreduce ........... True + attention_backend ............................... AttnBackend.auto + attention_dropout ............................... 0.1 + attention_softmax_in_fp32 ....................... False + auto_detect_ckpt_format ......................... False + barrier_with_L1_time ............................ True + bert_binary_head ................................ True + bert_embedder_type .............................. megatron + bert_load ....................................... None + bf16 ............................................ False + bias_dropout_fusion ............................. True + bias_gelu_fusion ................................ True + bias_swiglu_fusion .............................. True + biencoder_projection_dim ........................ 0 + biencoder_shared_query_context_model ............ False + block_data_path ................................. None + calc_ft_timeouts ................................ False + calculate_per_token_loss ........................ False + check_for_large_grads ........................... False + check_for_nan_in_loss_and_grad .................. False + check_for_spiky_loss ............................ False + check_weight_hash_across_dp_replicas_interval ... None + ckpt_assume_constant_structure .................. False + ckpt_convert_format ............................. None + ckpt_convert_save ............................... None + ckpt_convert_update_legacy_dist_opt_format ...... False + ckpt_format ..................................... torch_dist + ckpt_fully_parallel_load ........................ False + ckpt_fully_parallel_save ........................ True + ckpt_fully_parallel_save_deprecated ............. False + ckpt_step ....................................... None + classes_fraction ................................ 1.0 + clip_grad ....................................... 1.0 + clone_scatter_output_in_embedding ............... True + config_logger_dir ............................... + consumed_train_samples .......................... 0 + consumed_valid_samples .......................... 0 + context_parallel_size ........................... 4 + cp_comm_type .................................... ['p2p'] + create_attention_mask_in_dataloader ............. True + cross_entropy_fusion_impl ....................... native + cross_entropy_loss_fusion ....................... False + cuda_graph_scope ................................ full + cuda_graph_warmup_steps ......................... 3 + data_args_path .................................. None + data_cache_path ................................. None + data_parallel_random_init ....................... False + data_parallel_sharding_strategy ................. no_shard + data_parallel_size .............................. 1 + data_path ....................................... None + data_per_class_fraction ......................... 1.0 + data_sharding ................................... True + dataloader_type ................................. single + ddp_average_in_collective ....................... False + ddp_bucket_size ................................. None + ddp_num_buckets ................................. None + ddp_pad_buckets_for_high_nccl_busbw ............. False + decoder_first_pipeline_num_layers ............... None + decoder_last_pipeline_num_layers ................ None + decoder_num_layers .............................. None + decoder_seq_length .............................. None + decoupled_lr .................................... None + decoupled_min_lr ................................ None + decrease_batch_size_if_needed ................... False + defer_embedding_wgrad_compute ................... False + deprecated_use_mcore_models ..................... False + deterministic_mode .............................. False + dino_bottleneck_size ............................ 256 + dino_freeze_last_layer .......................... 1 + dino_head_hidden_size ........................... 2048 + dino_local_crops_number ......................... 10 + dino_local_img_size ............................. 96 + dino_norm_last_layer ............................ False + dino_teacher_temp ............................... 0.07 + dino_warmup_teacher_temp ........................ 0.04 + dino_warmup_teacher_temp_epochs ................. 30 + disable_bf16_reduced_precision_matmul ........... False + disable_mamba_mem_eff_path ...................... False + disable_straggler_on_startup .................... False + dist_ckpt_format_deprecated ..................... None + dist_ckpt_strictness ............................ assume_ok_unexpected + distribute_saved_activations .................... False + distributed_backend ............................. nccl + distributed_timeout_minutes ..................... 10 + embedding_path .................................. None + empty_unused_memory_level ....................... 0 + enable_cuda_graph ............................... False + enable_ft_package ............................... False + enable_gloo_process_groups ...................... True + enable_msc ...................................... True + enable_one_logger ............................... True + encoder_num_layers .............................. 2 + encoder_pipeline_model_parallel_size ............ 0 + encoder_seq_length .............................. 24576 + encoder_tensor_model_parallel_size .............. 0 + end_weight_decay ................................ 0.1 + eod_mask_loss ................................... False + error_injection_rate ............................ 0 + error_injection_type ............................ transient_error + eval_interval ................................... 16 + eval_iters ...................................... 1 + evidence_data_path .............................. None + exit_duration_in_mins ........................... None + exit_interval ................................... None + exit_on_missing_checkpoint ...................... False + exit_signal_handler ............................. False + exp_avg_dtype ................................... torch.float32 + exp_avg_sq_dtype ................................ torch.float32 + expert_model_parallel_size ...................... 1 + expert_tensor_parallel_size ..................... 2 + external_cuda_graph ............................. False + ffn_hidden_size ................................. 16384 + finetune ........................................ False + first_last_layers_bf16 .......................... False + flash_decode .................................... False + fp16 ............................................ True + fp16_lm_cross_entropy ........................... False + fp32_residual_connection ........................ False + fp8 ............................................. None + fp8_amax_compute_algo ........................... most_recent + fp8_amax_history_len ............................ 1 + fp8_interval .................................... 1 + fp8_margin ...................................... 0 + fp8_param_gather ................................ False + fp8_recipe ...................................... delayed + fp8_wgrad ....................................... True + fsdp_double_buffer .............................. False + global_batch_size ............................... 1 + grad_reduce_in_bf16 ............................. False + gradient_accumulation_fusion .................... True + gradient_reduce_div_fusion ...................... True + group_query_attention ........................... True + head_lr_mult .................................... 1.0 + heterogeneous_layers_config_encoded_json ........ None + heterogeneous_layers_config_path ................ None + hidden_dropout .................................. 0.1 + hidden_size ..................................... 4096 + hierarchical_context_parallel_sizes ............. None + high_priority_stream_groups ..................... [] + hybrid_attention_ratio .......................... 0.0 + hybrid_mlp_ratio ................................ 0.0 + hybrid_override_pattern ......................... None + hysteresis ...................................... 2 + ict_head_size ................................... None + ict_load ........................................ None + img_h ........................................... 224 + img_w ........................................... 224 + indexer_batch_size .............................. 128 + indexer_log_interval ............................ 1000 + inference_batch_times_seqlen_threshold .......... -1 + inference_dynamic_batching ...................... False + inference_dynamic_batching_buffer_guaranteed_fraction 0.2 + inference_dynamic_batching_buffer_overflow_factor None + inference_dynamic_batching_buffer_size_gb ....... 40.0 + inference_dynamic_batching_chunk_size ........... 256 + inference_dynamic_batching_max_requests_override None + inference_dynamic_batching_max_tokens_override .. None + inference_max_batch_size ........................ 8 + inference_max_seq_length ........................ 2560 + inference_rng_tracker ........................... False + init_method_std ................................. 0.02 + init_method_xavier_uniform ...................... False + init_model_with_meta_device ..................... False + initial_loss_scale .............................. 4294967296 + inprocess_active_world_size ..................... 8 + inprocess_barrier_timeout ....................... 120 + inprocess_completion_timeout .................... 120 + inprocess_empty_cuda_cache ...................... False + inprocess_granularity ........................... node + inprocess_hard_timeout .......................... 90 + inprocess_heartbeat_interval .................... 30 + inprocess_heartbeat_timeout ..................... 60 + inprocess_last_call_wait ........................ 1 + inprocess_max_iterations ........................ None + inprocess_monitor_process_interval .............. 1.0 + inprocess_monitor_thread_interval ............... 1.0 + inprocess_progress_watchdog_interval ............ 1.0 + inprocess_restart ............................... False + inprocess_soft_timeout .......................... 60 + inprocess_termination_grace_time ................ 1 + is_hybrid_model ................................. False + iter_per_epoch .................................. 1250 + iterations_to_skip .............................. [] + keep_fp8_transpose_cache_when_using_custom_fsdp . False + kv_channels ..................................... 64 + kv_lora_rank .................................... 32 + lazy_mpu_init ................................... None + load ............................................ gpt-checkpoint + load_model_opt_format ........................... False + local_rank ...................................... 0 + log_interval .................................... 1 + log_loss_scale_to_tensorboard ................... True + log_memory_to_tensorboard ....................... False + log_num_zeros_in_grad ........................... False + log_params_norm ................................. False + log_progress .................................... False + log_straggler ................................... False + log_throughput .................................. False + log_timers_to_tensorboard ....................... False + log_validation_ppl_to_tensorboard ............... False + log_world_size_to_tensorboard ................... False + logging_level ................................... 0 + loss_scale ...................................... None + loss_scale_window ............................... 1000 + lr .............................................. 0.0005 + lr_decay_iters .................................. 150000 + lr_decay_samples ................................ None + lr_decay_style .................................. cosine + lr_warmup_fraction .............................. None + lr_warmup_init .................................. 0.0 + lr_warmup_iters ................................. 2 + lr_warmup_samples ............................... 0 + lr_wsd_decay_iters .............................. None + lr_wsd_decay_samples ............................ None + lr_wsd_decay_style .............................. exponential + main_grads_dtype ................................ torch.float32 + main_params_dtype ............................... torch.float32 + make_vocab_size_divisible_by .................... 128 + mamba_head_dim .................................. 64 + mamba_num_groups ................................ 8 + mamba_num_heads ................................. None + mamba_state_dim ................................. 128 + manual_gc ....................................... False + manual_gc_eval .................................. True + manual_gc_interval .............................. 0 + mask_factor ..................................... 1.0 + mask_prob ....................................... 0.15 + mask_type ....................................... random + masked_softmax_fusion ........................... True + max_position_embeddings ......................... 24576 + max_tokens_to_oom ............................... 12000 + memory_snapshot_path ............................ snapshot.pickle + merge_file ...................................... merges.txt + micro_batch_size ................................ 1 + microbatch_group_size_per_vp_stage .............. None + mid_level_dataset_surplus ....................... 0.005 + min_loss_scale .................................. 1.0 + min_lr .......................................... 0.0 + mlp_chunks_for_prefill .......................... 1 + mmap_bin_files .................................. True + mock_data ....................................... True + moe_apply_probs_on_input ........................ False + moe_aux_loss_coeff .............................. 0.0 + moe_enable_deepep ............................... False + moe_expert_capacity_factor ...................... None + moe_extended_tp ................................. False + moe_ffn_hidden_size ............................. None + moe_grouped_gemm ................................ False + moe_input_jitter_eps ............................ None + moe_layer_freq .................................. 1 + moe_layer_recompute ............................. False + moe_pad_expert_input_to_capacity ................ False + moe_per_layer_logging ........................... False + moe_permute_fusion .............................. False + moe_router_bias_update_rate ..................... 0.001 + moe_router_dtype ................................ None + moe_router_enable_expert_bias ................... False + moe_router_force_load_balancing ................. False + moe_router_group_topk ........................... None + moe_router_load_balancing_type .................. aux_loss + moe_router_num_groups ........................... None + moe_router_padding_for_fp8 ...................... False + moe_router_pre_softmax .......................... False + moe_router_score_function ....................... softmax + moe_router_topk ................................. 2 + moe_router_topk_scaling_factor .................. None + moe_shared_expert_intermediate_size ............. None + moe_shared_expert_overlap ....................... False + moe_token_dispatcher_type ....................... allgather + moe_token_drop_policy ........................... probs + moe_use_legacy_grouped_gemm ..................... False + moe_use_upcycling ............................... False + moe_z_loss_coeff ................................ None + mrope_section ................................... None + mscale .......................................... 1.0 + mscale_all_dim .................................. 1.0 + mtp_loss_scaling_factor ......................... 0.1 + mtp_num_layers .................................. None + multi_latent_attention .......................... False + nccl_all_reduce_for_prefill ..................... False + nccl_communicator_config_path ................... None + nccl_ub ......................................... False + no_load_optim ................................... None + no_load_rng ..................................... None + no_persist_layer_norm ........................... False + no_rope_freq .................................... None + no_save_optim ................................... None + no_save_rng ..................................... None + non_persistent_ckpt_type ........................ None + non_persistent_global_ckpt_dir .................. None + non_persistent_local_ckpt_algo .................. fully_parallel + non_persistent_local_ckpt_dir ................... None + non_persistent_save_interval .................... None + norm_epsilon .................................... 1e-05 + normalization ................................... LayerNorm + num_attention_heads ............................. 64 + num_channels .................................... 3 + num_classes ..................................... 1000 + num_dataset_builder_threads ..................... 1 + num_distributed_optimizer_instances ............. 1 + num_experts ..................................... None + num_layers ...................................... 2 + num_layers_at_end_in_bf16 ....................... 1 + num_layers_at_start_in_bf16 ..................... 1 + num_layers_per_virtual_pipeline_stage ........... None + num_query_groups ................................ 16 + num_virtual_stages_per_pipeline_rank ............ None + num_workers ..................................... 2 + object_storage_cache_path ....................... None + one_logger_async ................................ False + one_logger_project .............................. megatron-lm + one_logger_run_name ............................. None + onnx_safe ....................................... None + openai_gelu ..................................... False + optimizer ....................................... adam + optimizer_cpu_offload ........................... False + optimizer_offload_fraction ...................... 1.0 + output_bert_embeddings .......................... False + overlap_cpu_optimizer_d2h_h2d ................... False + overlap_grad_reduce ............................. False + overlap_p2p_comm ................................ False + overlap_p2p_comm_warmup_flush ................... False + overlap_param_gather ............................ False + overlap_param_gather_with_optimizer_step ........ False + override_opt_param_scheduler .................... False + params_dtype .................................... torch.float16 + patch_dim ....................................... 16 + per_split_data_args_path ........................ None + perform_initialization .......................... True + pin_cpu_grads ................................... True + pin_cpu_params .................................. True + pipeline_model_parallel_comm_backend ............ None + pipeline_model_parallel_size .................... 1 + pipeline_model_parallel_split_rank .............. None + position_embedding_type ......................... learned_absolute + pretrained_checkpoint ........................... None + profile ......................................... False + profile_ranks ................................... [0] + profile_step_end ................................ 12 + profile_step_start .............................. 10 + q_lora_rank ..................................... None + qk_head_dim ..................................... 128 + qk_l2_norm ...................................... False + qk_layernorm .................................... False + qk_pos_emb_head_dim ............................. 64 + query_in_block_prob ............................. 0.1 + rampup_batch_size ............................... None + rank ............................................ 0 + recompute_granularity ........................... None + recompute_method ................................ None + recompute_modules ............................... None + recompute_num_layers ............................ None + record_memory_history ........................... False + relative_attention_max_distance ................. 128 + relative_attention_num_buckets .................. 32 + replication ..................................... False + replication_factor .............................. 2 + replication_jump ................................ None + rerun_mode ...................................... disabled + reset_attention_mask ............................ False + reset_position_ids .............................. False + result_rejected_tracker_filename ................ None + retriever_report_topk_accuracies ................ [] + retriever_score_scaling ......................... False + retriever_seq_length ............................ 256 + retro_add_retriever ............................. False + retro_attention_gate ............................ 1 + retro_cyclic_train_iters ........................ None + retro_encoder_attention_dropout ................. 0.1 + retro_encoder_hidden_dropout .................... 0.1 + retro_encoder_layers ............................ 2 + retro_num_neighbors ............................. 2 + retro_num_retrieved_chunks ...................... 2 + retro_project_dir ............................... None + retro_verify_neighbor_count ..................... True + rope_scaling_factor ............................. 8.0 + rotary_base ..................................... 10000 + rotary_interleaved .............................. False + rotary_percent .................................. 1.0 + rotary_scaling_factor ........................... 1.0 + rotary_seq_len_interpolation_factor ............. None + run_workload_inspector_server ................... False + sample_rate ..................................... 1.0 + save ............................................ gpt-checkpoint + save_interval ................................... 16 + scatter_gather_tensors_in_pipeline .............. True + seed ............................................ 1234 + seq_length ...................................... 24576 + sequence_parallel ............................... False + sgd_momentum .................................... 0.9 + short_seq_prob .................................. 0.1 + skip_train ...................................... False + skipped_train_samples ........................... 0 + spec ............................................ None + split ........................................... None + squared_relu .................................... False + start_weight_decay .............................. 0.1 + straggler_ctrlr_port ............................ 65535 + straggler_minmax_count .......................... 1 + suggested_communication_unit_size ............... None + swiglu .......................................... False + swin_backbone_type .............................. tiny + symmetric_ar_type ............................... None + te_rng_tracker .................................. False + tensor_model_parallel_size ...................... 2 + tensorboard_dir ................................. tensorboard-logs/ + tensorboard_log_interval ........................ 1 + tensorboard_queue_size .......................... 1000 + test_data_path .................................. None + test_mode ....................................... False + tiktoken_num_special_tokens ..................... 1000 + tiktoken_pattern ................................ None + tiktoken_special_tokens ......................... None + timing_log_level ................................ 0 + timing_log_option ............................... minmax + titles_data_path ................................ None + tokenizer_model ................................. None + tokenizer_type .................................. GPT2BPETokenizer + torch_fsdp2_reshard_after_forward ............... True + tp_comm_bootstrap_backend ....................... nccl + tp_comm_bulk_dgrad .............................. True + tp_comm_bulk_wgrad .............................. True + tp_comm_overlap ................................. False + tp_comm_overlap_ag .............................. True + tp_comm_overlap_cfg ............................. None + tp_comm_overlap_rs .............................. True + tp_comm_overlap_rs_dgrad ........................ False + tp_comm_split_ag ................................ True + tp_comm_split_rs ................................ True + train_data_path ................................. None + train_iters ..................................... 10 + train_samples ................................... None + train_sync_interval ............................. None + transformer_impl ................................ transformer_engine + transformer_pipeline_model_parallel_size ........ 1 + untie_embeddings_and_output_weights ............. False + use_checkpoint_args ............................. False + use_checkpoint_opt_param_scheduler .............. False + use_cpu_initialization .......................... None + use_custom_fsdp ................................. False + use_dist_ckpt ................................... True + use_dist_ckpt_deprecated ........................ False + use_distributed_optimizer ....................... False + use_flash_attn .................................. False + use_legacy_models ............................... False + use_mp_args_from_checkpoint_args ................ False + use_one_sent_docs ............................... False + use_persistent_ckpt_worker ...................... False + use_precision_aware_optimizer ................... False + use_pytorch_profiler ............................ False + use_ring_exchange_p2p ........................... False + use_rope_scaling ................................ False + use_rotary_position_embeddings .................. False + use_sharp ....................................... False + use_tokenizer_model_from_checkpoint_args ........ True + use_torch_fsdp2 ................................. False + use_torch_optimizer_for_cpu_offload ............. False + use_tp_pp_dp_mapping ............................ False + v_head_dim ...................................... 128 + valid_data_path ................................. None + variable_seq_lengths ............................ False + virtual_pipeline_model_parallel_size ............ None + vision_backbone_type ............................ vit + vision_pretraining .............................. False + vision_pretraining_type ......................... classify + vocab_extra_ids ................................. 0 + vocab_file ...................................... vocab.json + vocab_size ...................................... None + wandb_exp_name .................................. + wandb_project ................................... + wandb_save_dir .................................. + weight_decay .................................... 0.1 + weight_decay_incr_style ......................... constant + wgrad_deferral_limit ............................ 0 + world_size ...................................... 8 + yaml_cfg ........................................ None +-------------------- end of arguments --------------------- +INFO:megatron.core.num_microbatches_calculator:setting number of microbatches to constant 1 +> building GPT2BPETokenizer tokenizer ... +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 +INFO:megatron.training.initialize:Setting logging level to 0 + > padded vocab (size: 50257) with 175 dummy tokens (new size: 50432) +INFO:megatron.training.initialize:Setting logging level to 0 +WARNING:megatron.core.rerun_state_machine:RerunStateMachine initialized in mode RerunMode.DISABLED +> initializing torch distributed ... +> initialized tensor model parallel with size 2 +> initialized pipeline model parallel with size 1 +> setting random seeds to 1234 ... +> compiling dataset index builder ... +make: Entering directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +make: Nothing to be done for 'default'. +make: Leaving directory '/mnt/weka/home/hao.zhang/junda/attnserver-megatron/megatron/core/datasets' +>>> done with dataset index builder. Compilation time: 0.040 seconds +WARNING: constraints for invoking optimized fused softmax kernel are not met. We default back to unfused kernel invocations. +> compiling and loading fused kernels ...