2025-11-11 11:24:34,344 - INFO - SFT Config: SFTConfig( _n_gpu=1, accelerator_config={'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False}, activation_offloading=False, adafactor=False, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, assistant_only_loss=False, auto_find_batch_size=False, average_tokens_across_devices=True, batch_eval_metrics=True, bf16=False, bf16_full_eval=False, chat_template_path=None, completion_only_loss=True, data_seed=42, dataloader_drop_last=False, dataloader_num_workers=0, dataloader_persistent_workers=False, dataloader_pin_memory=True, dataloader_prefetch_factor=None, dataset_kwargs=None, dataset_num_proc=None, dataset_text_field=text, ddp_backend=None, ddp_broadcast_buffers=None, ddp_bucket_cap_mb=None, ddp_find_unused_parameters=None, ddp_timeout=1800, debug=[], deepspeed=None, disable_tqdm=False, do_eval=True, do_predict=False, do_train=False, eos_token=, eval_accumulation_steps=None, eval_delay=0, eval_do_concat_batches=True, eval_on_start=True, eval_packing=None, eval_steps=0.1, eval_strategy=IntervalStrategy.STEPS, eval_use_gather_object=False, fp16=False, fp16_backend=auto, fp16_full_eval=False, fp16_opt_level=O1, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_min_num_params=0, fsdp_transformer_layer_cls_to_wrap=None, full_determinism=False, gradient_accumulation_steps=2, gradient_checkpointing=False, gradient_checkpointing_kwargs=None, greater_is_better=None, group_by_length=False, half_precision_backend=auto, hub_always_push=False, hub_model_id=None, hub_private_repo=False, hub_revision=None, hub_strategy=HubStrategy.EVERY_SAVE, hub_token=, ignore_data_skip=False, include_for_metrics=[], include_inputs_for_metrics=False, include_num_input_tokens_seen=no, include_tokens_per_second=False, jit_mode_eval=False, label_names=None, label_smoothing_factor=0.0, learning_rate=2e-05, length_column_name=length, liger_kernel_config=None, load_best_model_at_end=False, local_rank=0, log_level=info, log_level_replica=warning, log_on_each_node=True, logging_dir=./runs, logging_first_step=True, logging_nan_inf_filter=True, logging_steps=5, logging_strategy=IntervalStrategy.STEPS, loss_type=nll, lr_scheduler_kwargs={}, lr_scheduler_type=SchedulerType.COSINE, max_grad_norm=1.0, max_length=1024, max_steps=-1, metric_for_best_model=None, model_init_kwargs=None, mp_parameters=, neftune_noise_alpha=None, no_cuda=False, num_train_epochs=2, optim=OptimizerNames.ADAMW_TORCH_FUSED, optim_args=None, optim_target_modules=None, output_dir=./models, overwrite_output_dir=False, packing=False, packing_strategy=bfd, pad_to_multiple_of=None, pad_token=, padding_free=False, parallelism_config=None, past_index=-1, per_device_eval_batch_size=6, per_device_train_batch_size=6, prediction_loss_only=False, project=huggingface, push_to_hub=False, push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=, ray_scope=last, remove_unused_columns=True, report_to=['tensorboard'], restore_callback_states_from_checkpoint=False, resume_from_checkpoint=None, run_name=None, save_on_each_node=False, save_only_model=False, save_safetensors=True, save_steps=0.1, save_strategy=SaveStrategy.STEPS, save_total_limit=3, seed=42, skip_memory_metrics=True, tf32=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, torch_empty_cache_steps=None, torchdynamo=None, tpu_metrics_debug=False, tpu_num_cores=None, trackio_space_id=trackio, use_cpu=False, use_legacy_prediction_loop=False, use_liger_kernel=False, use_mps_device=False, warmup_ratio=0.05, warmup_steps=0, weight_decay=0.01, ) 2025-11-11 11:24:34,344 - INFO - Data Config: data_train_path='./data/train_sft_data.jsonl' data_dev_path='./data/dev_preprocessed_human.jsonl' data_test_path='./data/test_preprocessed_human.jsonl' db_train_path='./data/train.db' db_dev_path='./data/dev.db' db_test_path='./data/test.db' do_sample_train=False do_sample_dev=False do_sample_test=False n_sample_train=200 n_sample_dev=50 n_sample_test=50 2025-11-11 11:25:42,178 - INFO - Starting full pipeline. 2025-11-11 11:25:42,178 - INFO - Training model... 2025-11-11 13:23:39,551 - INFO - Training completed. 2025-11-11 13:47:01,198 - INFO - Final Dev Metrics: {'lf_accuracy': 0.7649813858532485, 'execution_accuracy': 0.8046115047436051} 2025-11-11 13:47:01,198 - INFO - Running final test evaluation... 2025-11-11 14:31:11,980 - INFO - Test Metrics: {'lf_accuracy': 0.7681140965236215, 'execution_accuracy': 0.807143766713358} 2025-11-11 14:31:11,980 - INFO - Final Test Metrics: {'lf_accuracy': 0.7681140965236215, 'execution_accuracy': 0.807143766713358}