| 2026-04-18 17:28:50,229 INFO MainThread:10295 [wandb_run.py:_config_callback():1385] config_cb None None {'peft_config': {'default': {'task_type': 'CAUSAL_LM', 'peft_type': 'LORA', 'auto_mapping': None, 'peft_version': '0.18.1', 'base_model_name_or_path': '/mnt/dhwfile/raise/user/panjiabao/huxiaobin/zhanghaobin/qwen3-asr-merged', 'revision': None, 'inference_mode': False, 'r': 8, 'target_modules': '^(thinker\\.model(?=\\.).*\\.(up_proj|q_proj|o_proj|gate_proj|k_proj|v_proj|down_proj)|thinker\\.audio_tower\\.layers(?=\\.).*\\.(q_proj|fc1|out_proj|k_proj|v_proj|fc2)|thinker\\.audio_tower\\.conv_out(?=\\.)|thinker\\.audio_tower\\.proj1(?=\\.)|thinker\\.audio_tower\\.proj2(?=\\.))$', 'exclude_modules': None, 'lora_alpha': 32, 'lora_dropout': 0.05, 'fan_in_fan_out': False, 'bias': 'none', 'use_rslora': False, 'modules_to_save': [], 'init_lora_weights': True, 'layers_to_transform': None, 'layers_pattern': None, 'rank_pattern': {}, 'alpha_pattern': {}, 'megatron_config': None, 'megatron_core': 'megatron.core', 'trainable_token_indices': None, 'loftq_config': {}, 'eva_config': None, 'corda_config': None, 'use_dora': False, 'alora_invocation_tokens': None, 'use_qalora': False, 'qalora_group_size': 16, 'layer_replication': None, 'runtime_config': {'ephemeral_gpu_offload': False}, 'lora_bias': False, 'target_parameters': None, 'arrow_config': None, 'ensure_weight_tying': False, 'lora_dtype': None, 'lorap_lr_ratio': None, 'lorap_emb_lr': 1e-06}}, 'return_dict': True, 'output_hidden_states': False, 'torchscript': False, 'dtype': 'bfloat16', 'pruned_heads': {}, 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'architectures': ['Qwen3ASRForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 151643, 'eos_token_id': 151645, 'sep_token_id': None, 'decoder_start_token_id': None, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'temperature': 0.5, 'top_k': 50, 'top_p': 0.95, 'typical_p': 1.0, 'repetition_penalty': 1.08, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'num_beam_groups': 1, 'diversity_penalty': 0.0, '_name_or_path': '/mnt/dhwfile/raise/user/panjiabao/huxiaobin/zhanghaobin/qwen3-asr-merged', 'transformers_version': '4.57.6', 'model_type': 'qwen3_asr', 'tf_legacy_loss': False, 'use_bfloat16': False, 'thinker_config': {'return_dict': True, 'output_hidden_states': False, 'torchscript': False, 'dtype': 'bfloat16', 'pruned_heads': {}, 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'architectures': ['Qwen3ASRForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'typical_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'num_beam_groups': 1, 'diversity_penalty': 0.0, '_name_or_path': '', 'audio_end_token_id': 151670, 'model_type': 'qwen3_asr_thinker', 'tf_legacy_loss': False, 'use_bfloat16': False, 'user_token_id': 872, 'audio_start_token_id': 151669, 'initializer_range': 0.02, 'audio_config': {'return_dict': True, 'output_hidden_states': False, 'torchscript': False, 'dtype': None, 'pruned_heads': {}, 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': None, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'typical_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'num_beam_groups': 1, 'diversity_penalty': 0.0, '_name_or_path': '', 'model_type': 'qwen3_asr_audio_encoder', 'num_hidden_layers': 24, 'tf_legacy_loss': False, 'use_bfloat16': False, 'num_mel_bins': 128, 'd_model': 1024, 'encoder_layers': 24, 'encoder_attention_heads': 16, 'encoder_ffn_dim': 4096, 'dropout': 0, 'attention_dropout': 0, 'activation_function': 'gelu', 'activation_dropout': 0, 'initializer_range': 0.02, 'scale_embedding': False, 'max_source_positions': 1500, 'n_window': 50, 'output_dim': 2048, 'n_window_infer': 800, 'conv_chunksize': 500, 'downsample_hidden_size': 480, 'output_attentions': False}, 'text_config': {'vocab_size': 151936, 'max_position_embeddings': 65536, 'hidden_size': 2048, 'intermediate_size': 6144, 'num_hidden_layers': 28, 'num_attention_heads': 16, 'num_key_value_heads': 8, 'head_dim': 128, 'hidden_act': 'silu', 'initializer_range': 0.02, 'rms_norm_eps': 1e-06, 'use_cache': False, 'rope_theta': 1000000, 'rope_scaling': {'interleaved': True, 'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_type': 'default', 'type': 'default'}, 'attention_bias': False, 'attention_dropout': 0.0, 'return_dict': True, 'output_hidden_states': False, 'torchscript': False, 'dtype': None, 'pruned_heads': {}, 'tie_word_embeddings': True, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'architectures': None, 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'task_specific_params': None, 'problem_type': None, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 151643, 'eos_token_id': None, 'sep_token_id': None, 'decoder_start_token_id': None, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'typical_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'exponential_decay_length_penalty': None, 'suppress_tokens': None, 'begin_suppress_tokens': None, 'num_beam_groups': 1, 'diversity_penalty': 0.0, '_name_or_path': '', 'model_type': 'qwen3_asr_text', 'tf_legacy_loss': False, 'use_bfloat16': False, 'output_attentions': False}, 'audio_token_id': 151676, 'output_attentions': False}, 'support_languages': ['Chinese', 'English', 'Cantonese', 'Arabic', 'German', 'French', 'Spanish', 'Portuguese', 'Indonesian', 'Italian', 'Korean', 'Russian', 'Thai', 'Vietnamese', 'Japanese', 'Turkish', 'Hindi', 'Malay', 'Dutch', 'Swedish', 'Danish', 'Finnish', 'Polish', 'Czech', 'Filipino', 'Persian', 'Greek', 'Romanian', 'Hungarian', 'Macedonian'], 'output_attentions': False, 'output_dir': '/mnt/dhwfile/raise/user/panjiabao/huxiaobin/zhanghaobin/0416upload/out/qwen3asr_dapo_reward56_4x4x12_12gen_4GPU/v5-20260418-172837', 'overwrite_output_dir': False, 'do_train': False, 'do_eval': True, 'do_predict': False, 'eval_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 4, 'per_device_eval_batch_size': 4, 'per_gpu_train_batch_size': None, 'per_gpu_eval_batch_size': None, 'gradient_accumulation_steps': 12, 'eval_accumulation_steps': None, 'eval_delay': 0, 'torch_empty_cache_steps': None, 'learning_rate': 5e-05, 'weight_decay': 0.1, 'adam_beta1': 0.9, 'adam_beta2': 0.95, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 3.0, 'max_steps': -1, 'lr_scheduler_type': 'cosine', 'lr_scheduler_kwargs': None, 'warmup_ratio': 0.03, 'warmup_steps': 0, 'log_level': 'passive', 'log_level_replica': 'warning', 'log_on_each_node': True, 'logging_dir': '/mnt/dhwfile/raise/user/panjiabao/huxiaobin/zhanghaobin/0416upload/out/qwen3asr_dapo_reward56_4x4x12_12gen_4GPU/v5-20260418-172837/runs', 'logging_strategy': 'steps', 'logging_first_step': True, 'logging_steps': 5, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 20, 'save_total_limit': None, 'save_safetensors': True, 'save_on_each_node': False, 'save_only_model': False, 'restore_callback_states_from_checkpoint': False, 'no_cuda': False, 'use_cpu': False, 'use_mps_device': False, 'seed': 42, 'data_seed': 42, 'jit_mode_eval': False, 'bf16': True, 'fp16': False, 'fp16_opt_level': 'O1', 'half_precision_backend': 'auto', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'local_rank': 0, 'ddp_backend': None, 'tpu_num_cores': None, 'tpu_metrics_debug': False, 'debug': [], 'dataloader_drop_last': True, 'eval_steps': 20, 'dataloader_num_workers': 1, 'dataloader_prefetch_factor': 2, 'past_index': -1, 'run_name': 'qwen3asr_dapo_reward56_4x4x12_12gen_4GPU', 'disable_tqdm': False, 'remove_unused_columns': False, 'label_names': None, 'load_best_model_at_end': False, 'metric_for_best_model': 'loss', 'greater_is_better': False, 'ignore_data_skip': False, 'fsdp': [], 'fsdp_min_num_params': 0, 'fsdp_config': {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, 'fsdp_transformer_layer_cls_to_wrap': None, 'accelerator_config': {'split_batches': False, 'dispatch_batches': False, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'deepspeed': None, 'label_smoothing_factor': 0.0, 'optim': 'adamw_torch', 'optim_args': None, 'adafactor': False, 'group_by_length': False, 'length_column_name': 'length', 'report_to': ['wandb'], 'project': 'huggingface', 'trackio_space_id': 'trackio', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': False, 'resume_from_checkpoint': None, 'hub_model_id': None, 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_always_push': False, 'hub_revision': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'include_inputs_for_metrics': False, 'include_for_metrics': [], 'eval_do_concat_batches': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': None, 'push_to_hub_organization': None, 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', 'mp_parameters': '', 'auto_find_batch_size': False, 'full_determinism': False, 'torchdynamo': None, 'ray_scope': 'last', 'ddp_timeout': 18000000, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'include_tokens_per_second': False, 'include_num_input_tokens_seen': 'no', 'neftune_noise_alpha': None, 'optim_target_modules': None, 'batch_eval_metrics': False, 'eval_on_start': False, 'use_liger_kernel': False, 'liger_kernel_config': None, 'eval_use_gather_object': False, 'average_tokens_across_devices': True, 'model_init_kwargs': None, 'disable_dropout': False, 'cast_lm_head_to_fp32': False, 'num_generations': 12, 'num_generations_eval': 4, 'max_completion_length': 256, 'ds3_gather_for_generation': True, 'shuffle_dataset': True, 'generation_batch_size': 48, 'steps_per_generation': 3, 'min_p': None, 'generation_kwargs': None, 'chat_template_kwargs': None, 'use_transformers_paged': False, 'cache_implementation': None, 'use_vllm': False, 'vllm_mode': None, 'vllm_model_impl': 'vllm', 'vllm_enable_sleep_mode': False, 'vllm_structured_outputs_regex': None, 'vllm_server_base_url': None, 'vllm_server_host': None, 'vllm_server_port': [8000], 'vllm_server_timeout': 240.0, 'vllm_group_port': 51216, 'vllm_gpu_memory_utilization': 0.9, 'vllm_max_model_length': None, 'vllm_tensor_parallel_size': 1, 'beta': 0.04, 'num_iterations': 2, 'epsilon': 0.2, 'delta': None, 'epsilon_high': 0.28, 'sapo_temperature_neg': 1.05, 'sapo_temperature_pos': 1.0, 'importance_sampling_level': 'token', 'reward_weights': None, 'multi_objective_aggregation': 'sum_then_normalize', 'scale_rewards': 'group', 'loss_type': 'dapo', 'mask_truncated_completions': False, 'sync_ref_model': False, 'ref_model_mixup_alpha': 0.6, 'ref_model_sync_steps': 512, 'top_entropy_quantile': 1.0, 'max_tool_calling_iterations': None, 'vllm_importance_sampling_correction': True, 'vllm_importance_sampling_mode': 'sequence_mask', 'vllm_importance_sampling_cap': 3.0, 'off_policy_mask_threshold': None, 'use_bias_correction_kl': False, 'log_completions': True, 'num_completions_to_print': None, 'log_unique_prompts': False, 'log_completions_hub_repo': None, 'tuner_backend': 'peft', 'vit_gradient_checkpointing': True, 'router_aux_loss_coef': 0.0, 'enable_dft_loss': False, 'enable_channel_loss': False, 'safe_serialization': True, 'max_shard_size': '5GB', 'check_model': True, 'acc_strategy': 'token', 'train_dataloader_shuffle': True, 'max_epochs': None, 'aligner_lr': None, 'vit_lr': None, 'use_logits_to_keep': None, 'resume_only_model': False, 'optimizer': None, 'eval_metric': None, 'callbacks': [], 'early_stop_interval': None, 'eval_use_evalscope': False, 'eval_dataset': [], 'eval_dataset_args': None, 'eval_limit': None, 'eval_generation_config': None, 'extra_eval_args': None, 'tuner_type': 'lora', 'use_galore': False, 'galore_target_modules': None, 'galore_rank': 128, 'galore_update_proj_gap': 50, 'galore_scale': 1.0, 'galore_proj_type': 'std', 'galore_optim_per_parameter': False, 'galore_with_embedding': False, 'galore_quantization': False, 'galore_proj_quant': False, 'galore_proj_bits': 4, 'galore_proj_group_size': 256, 'galore_cos_threshold': 0.4, 'galore_gamma_proj': 2, 'galore_queue_size': 5, 'lisa_activated_layers': 0, 'lisa_step_interval': 20, 'use_flash_ckpt': False, 'vllm_pipeline_parallel_size': 1, 'vllm_enable_expert_parallel': False, 'vllm_max_num_seqs': None, 'vllm_max_model_len': None, 'vllm_disable_custom_all_reduce': True, 'vllm_enforce_eager': False, 'vllm_limit_mm_per_prompt': None, 'vllm_max_lora_rank': 16, 'vllm_enable_prefix_caching': True, 'vllm_use_async_engine': None, 'vllm_quantization': None, 'vllm_reasoning_parser': None, 'vllm_disable_cascade_attn': False, 'vllm_mm_processor_cache_gb': None, 'vllm_speculative_config': None, 'vllm_engine_kwargs': {}, 'vllm_data_parallel_size': 1, 'stop_words': [], 'vllm_enable_lora': False, 'lora_rank': 8, 'vllm_server_group_port': None, 'enable_flattened_weight_sync': True, 'async_generate': False, 'structured_outputs_regex': None, 'sleep_level': 0, 'move_model_batches': None, 'offload_optimizer': False, 'offload_model': False, 'wandb_log_unique_prompts': None, 'cosine_min_len_value_wrong': -0.5, 'cosine_max_len_value_wrong': 0.0, 'cosine_min_len_value_correct': 1.0, 'cosine_max_len_value_correct': 0.5, 'cosine_max_len': 256, 'repetition_n_grams': 3, 'repetition_max_penalty': -1.0, 'reward_model': None, 'reward_model_plugin': None, 'chord_sft_dataset': [], 'chord_sft_per_device_train_batch_size': None, 'chord_enable_phi_function': False, 'chord_mu_warmup_steps': None, 'chord_mu_decay_steps': None, 'chord_mu_peak': None, 'chord_mu_valley': None, 'multi_turn_scheduler': None, 'max_turns': None, 'completion_length_limit_scope': 'per_round', 'vllm_server_pass_dataset': False, 'dynamic_sample': True, 'max_resample_times': 4, 'overlong_filter': True, 'soft_max_length': None, 'soft_cache_length': None, 'log_entropy': False, 'tau_pos': 1.0, 'tau_neg': 1.05, 'advantage_estimator': 'grpo', 'kl_in_reward': False, 'dataset_shuffle': True, 'rollout_importance_sampling_mode': None, 'rollout_importance_sampling_threshold': 2.0, 'log_rollout_offpolicy_metrics': False, 'off_policy_sequence_mask_delta': None} |
