Using the `WANDB_DISABLED` environment variable is deprecated and will be removed in v5. Use the --report_to flag to control the integrations used for logging result (for instance --report_to none). 02/26/2024 18:37:58 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1, distributed training: False, 16-bits training: False 02/26/2024 18:37:58 - INFO - __main__ - Training/evaluation parameters Seq2SeqTrainingArguments( _n_gpu=1, adafactor=False, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, auto_find_batch_size=False, bf16=False, bf16_full_eval=False, data_seed=None, dataloader_drop_last=False, dataloader_num_workers=0, dataloader_persistent_workers=False, dataloader_pin_memory=True, 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, dispatch_batches=None, do_eval=False, do_predict=True, do_train=False, eval_accumulation_steps=None, eval_delay=0, eval_steps=None, evaluation_strategy=no, fp16=False, fp16_backend=auto, fp16_full_eval=False, fp16_opt_level=O1, fsdp=[], fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_grad_ckpt': False}, fsdp_min_num_params=0, fsdp_transformer_layer_cls_to_wrap=None, full_determinism=False, generation_config=None, generation_max_length=None, generation_num_beams=2, gradient_accumulation_steps=1, 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_strategy=every_save, hub_token=, ignore_data_skip=False, include_inputs_for_metrics=False, include_num_input_tokens_seen=False, include_tokens_per_second=False, jit_mode_eval=False, label_names=None, label_smoothing_factor=0.0, learning_rate=5e-05, length_column_name=input_length, load_best_model_at_end=False, local_rank=0, log_level=passive, log_level_replica=warning, log_on_each_node=True, logging_dir=/beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/eval/cascaded_SLU_wo_num2word/runs/Feb26_18-37-57_chasma-01.int.europe.naverlabs.com, logging_first_step=False, logging_nan_inf_filter=True, logging_steps=500, logging_strategy=steps, lr_scheduler_kwargs={}, lr_scheduler_type=linear, max_grad_norm=1.0, max_steps=-1, metric_for_best_model=None, mp_parameters=, neftune_noise_alpha=None, no_cuda=False, num_train_epochs=3.0, optim=adamw_torch, optim_args=None, output_dir=/beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/eval/cascaded_SLU_wo_num2word, overwrite_output_dir=False, past_index=-1, per_device_eval_batch_size=32, per_device_train_batch_size=8, predict_with_generate=True, prediction_loss_only=False, 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=[], resume_from_checkpoint=None, run_name=/beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/eval/cascaded_SLU_wo_num2word, save_on_each_node=False, save_only_model=False, save_safetensors=True, save_steps=500, save_strategy=steps, save_total_limit=None, seed=42, skip_memory_metrics=True, sortish_sampler=False, split_batches=False, tf32=None, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, torchdynamo=None, tpu_metrics_debug=False, tpu_num_cores=None, use_cpu=False, use_ipex=False, use_legacy_prediction_loop=False, use_mps_device=False, warmup_ratio=0.0, warmup_steps=0, weight_decay=0.0, ) Loading Dataset Infos from /beegfs/scratch/user/blee/hugging-face/models/modules/datasets_modules/datasets/speech_massive_cascaded_v2/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f 02/26/2024 18:37:58 - INFO - datasets.info - Loading Dataset Infos from /beegfs/scratch/user/blee/hugging-face/models/modules/datasets_modules/datasets/speech_massive_cascaded_v2/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f Overwrite dataset info from restored data version if exists. 02/26/2024 18:37:58 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists. Loading Dataset info from /beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f 02/26/2024 18:37:58 - INFO - datasets.info - Loading Dataset info from /beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f Found cached dataset speech_massive_cascaded_v2 (/beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f) 02/26/2024 18:37:58 - INFO - datasets.builder - Found cached dataset speech_massive_cascaded_v2 (/beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f) Loading Dataset info from /beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f 02/26/2024 18:37:58 - INFO - datasets.info - Loading Dataset info from /beegfs/scratch/user/blee/hugging-face/models/datasets/speech_massive_cascaded_v2/multilingual-test/1.0.0/87fd9810b74e996705ff0d1428302d3938e96498ebd6c48c6f7ddbb89ac91d6f [INFO|configuration_utils.py:737] 2024-02-26 18:37:58,259 >> loading configuration file /beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/config.json [INFO|configuration_utils.py:802] 2024-02-26 18:37:58,270 >> Model config MT5Config { "_name_or_path": "/beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407", "architectures": [ "MT5ForConditionalGeneration" ], "classifier_dropout": 0.0, "d_ff": 2048, "d_kv": 64, "d_model": 768, "decoder_start_token_id": 0, "dense_act_fn": "gelu_new", "dropout": 0.2, "dropout_rate": 0.1, "eos_token_id": 1, "feed_forward_proj": "gated-gelu", "initializer_factor": 1.0, "is_encoder_decoder": true, "is_gated_act": true, "layer_norm_epsilon": 1e-06, "model_type": "mt5", "num_decoder_layers": 12, "num_heads": 12, "num_layers": 12, "output_past": true, "pad_token_id": 0, "relative_attention_max_distance": 128, "relative_attention_num_buckets": 32, "tie_word_embeddings": false, "tokenizer_class": "T5Tokenizer", "torch_dtype": "float32", "transformers_version": "4.37.0.dev0", "use_cache": true, "vocab_size": 250112 } [INFO|tokenization_utils_base.py:2024] 2024-02-26 18:37:58,272 >> loading file spiece.model [INFO|tokenization_utils_base.py:2024] 2024-02-26 18:37:58,272 >> loading file tokenizer.json [INFO|tokenization_utils_base.py:2024] 2024-02-26 18:37:58,273 >> loading file added_tokens.json [INFO|tokenization_utils_base.py:2024] 2024-02-26 18:37:58,273 >> loading file special_tokens_map.json [INFO|tokenization_utils_base.py:2024] 2024-02-26 18:37:58,273 >> loading file tokenizer_config.json [INFO|modeling_utils.py:3373] 2024-02-26 18:37:58,770 >> loading weights file /beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/model.safetensors [INFO|configuration_utils.py:826] 2024-02-26 18:37:58,879 >> Generate config GenerationConfig { "decoder_start_token_id": 0, "eos_token_id": 1, "pad_token_id": 0 } [INFO|modeling_utils.py:4224] 2024-02-26 18:38:03,296 >> All model checkpoint weights were used when initializing MT5ForConditionalGeneration. [INFO|modeling_utils.py:4232] 2024-02-26 18:38:03,296 >> All the weights of MT5ForConditionalGeneration were initialized from the model checkpoint at /beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407. If your task is similar to the task the model of the checkpoint was trained on, you can already use MT5ForConditionalGeneration for predictions without further training. [INFO|configuration_utils.py:779] 2024-02-26 18:38:03,303 >> loading configuration file /beegfs/scratch/user/blee/project_3/models/NLU.mt5-base.task_type-1.fine_tune.gpu_a100-40g+.node-1x1.bsz-64.epochs-22.metric-ema.metric_lang-all/checkpoint-30407/generation_config.json [INFO|configuration_utils.py:826] 2024-02-26 18:38:03,303 >> Generate config GenerationConfig { "decoder_start_token_id": 0, "eos_token_id": 1, "pad_token_id": 0 } Running tokenizer on prediction dataset: 0%| | 0/2974 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:38:08,885 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:38:08,885 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:38:08,886 >> Batch size = 32 [WARNING|logging.py:314] 2024-02-26 18:38:08,894 >> You're using a T5TokenizerFast tokenizer. Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the text followed by a call to the `pad` method to get a padded encoding. 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:38:57,771 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:38:57,771 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:38:57,771 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:39:49,067 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:39:49,067 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:39:49,067 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:40:46,775 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:40:46,775 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:40:46,775 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:41:46,289 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:41:46,289 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:41:46,289 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:42:34,801 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:42:34,801 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:42:34,801 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:43:17,325 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:43:17,325 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:43:17,325 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:44:10,057 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:44:10,057 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:44:10,057 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:44:59,741 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:44:59,741 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:44:59,741 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:45:53,791 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:45:53,791 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:45:53,791 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:46:43,394 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:46:43,394 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:46:43,394 >> Batch size = 32 0%| | 0/93 [00:00> The following columns in the test set don't have a corresponding argument in `MT5ForConditionalGeneration.forward` and have been ignored: id, intent_str, annot_utt. If id, intent_str, annot_utt are not expected by `MT5ForConditionalGeneration.forward`, you can safely ignore this message. [INFO|trainer.py:3199] 2024-02-26 18:47:32,579 >> ***** Running Prediction ***** [INFO|trainer.py:3201] 2024-02-26 18:47:32,579 >> Num examples = 2974 [INFO|trainer.py:3204] 2024-02-26 18:47:32,579 >> Batch size = 32 0%| | 0/93 [00:00