| { | |
| "_name_or_path": "facebook/bart-base", | |
| "accelerator": null, | |
| "accumulate_grad_batches": 2, | |
| "activation_dropout": 0.1, | |
| "activation_function": "gelu", | |
| "add_bias_logits": false, | |
| "add_context": true, | |
| "add_final_layer_norm": false, | |
| "amp_backend": null, | |
| "amp_level": null, | |
| "architectures": [ | |
| "BartForContextualGeneration" | |
| ], | |
| "attention_dropout": 0.1, | |
| "auto_lr_find": false, | |
| "auto_scale_batch_size": false, | |
| "auto_select_gpus": null, | |
| "batch_size": 8, | |
| "benchmark": null, | |
| "bos_token_id": 0, | |
| "check_val_every_n_epoch": 1, | |
| "checkpoint": null, | |
| "class_labels": { | |
| "delete": 4, | |
| "ignore": 0, | |
| "merge": 3, | |
| "rephrase": 1, | |
| "split": 2 | |
| }, | |
| "classif_dropout": 0.1, | |
| "classifier_dropout": 0.0, | |
| "context_dir": "context/complex", | |
| "context_doc_id": "pair_id", | |
| "context_window": 5, | |
| "d_model": 768, | |
| "decoder_attention_heads": 12, | |
| "decoder_ffn_dim": 3072, | |
| "decoder_layerdrop": 0.0, | |
| "decoder_layers": 6, | |
| "decoder_start_token_id": 2, | |
| "default_root_dir": null, | |
| "detect_anomaly": false, | |
| "devices": "2", | |
| "dropout": 0.1, | |
| "early_stopping": true, | |
| "enable_checkpointing": true, | |
| "enable_model_summary": true, | |
| "enable_progress_bar": true, | |
| "encoder_attention_heads": 12, | |
| "encoder_ffn_dim": 3072, | |
| "encoder_layerdrop": 0.0, | |
| "encoder_layers": 6, | |
| "eos_token_id": 2, | |
| "eval_beams": null, | |
| "eval_max_length": null, | |
| "fast_dev_run": false, | |
| "forced_bos_token_id": 0, | |
| "forced_eos_token_id": 2, | |
| "gpus": null, | |
| "gradient_checkpointing": false, | |
| "gradient_clip_algorithm": null, | |
| "gradient_clip_val": null, | |
| "id2label": { | |
| "0": "LABEL_0", | |
| "1": "LABEL_1", | |
| "2": "LABEL_2" | |
| }, | |
| "inference_mode": true, | |
| "init_std": 0.02, | |
| "ipus": null, | |
| "is_encoder_decoder": true, | |
| "label2id": { | |
| "LABEL_0": 0, | |
| "LABEL_1": 1, | |
| "LABEL_2": 2 | |
| }, | |
| "limit_predict_batches": null, | |
| "limit_test_batches": null, | |
| "limit_train_batches": null, | |
| "limit_val_batches": null, | |
| "log_every_n_steps": 50, | |
| "logger": true, | |
| "longformer": false, | |
| "lr": 2e-05, | |
| "lr_scheduler": false, | |
| "max_epochs": null, | |
| "max_length": 1024, | |
| "max_position_embeddings": 1024, | |
| "max_samples": -1, | |
| "max_steps": -1, | |
| "max_time": null, | |
| "min_epochs": null, | |
| "min_steps": null, | |
| "model_type": "context-bart", | |
| "move_metrics_to_cpu": false, | |
| "multiple_trainloader_mode": "max_size_cycle", | |
| "name": "o-conbart", | |
| "no_repeat_ngram_size": 3, | |
| "normalize_before": false, | |
| "normalize_embedding": true, | |
| "num_beams": 4, | |
| "num_hidden_layers": 6, | |
| "num_nodes": 1, | |
| "num_processes": null, | |
| "num_sanity_val_steps": 2, | |
| "op_col": "label", | |
| "op_tokens": [ | |
| "<COPY>", | |
| "<REPHRASE>", | |
| "<SPLIT>", | |
| "<MERGE>", | |
| "<DELETE>" | |
| ], | |
| "overfit_batches": 0.0, | |
| "pad_token_id": 1, | |
| "plan_col": "labels", | |
| "plan_loss": null, | |
| "plan_prefix": false, | |
| "plan_sep": false, | |
| "plugins": null, | |
| "precision": 32, | |
| "prefix_only": false, | |
| "profiler": null, | |
| "project": "simplification", | |
| "reading_lvl": null, | |
| "reload_dataloaders_every_n_epochs": 0, | |
| "replace_sampler_ddp": true, | |
| "resume_from_checkpoint": null, | |
| "save_dir": null, | |
| "scale_embedding": false, | |
| "sent_level": true, | |
| "simple_context_dir": "context/simple", | |
| "simple_context_doc_id": "pair_id", | |
| "skip_val_gen": true, | |
| "strategy": null, | |
| "sync_batchnorm": false, | |
| "task_specific_params": { | |
| "summarization": { | |
| "length_penalty": 1.0, | |
| "max_length": 128, | |
| "min_length": 12, | |
| "num_beams": 4 | |
| }, | |
| "summarization_cnn": { | |
| "length_penalty": 2.0, | |
| "max_length": 142, | |
| "min_length": 56, | |
| "num_beams": 4 | |
| }, | |
| "summarization_xsum": { | |
| "length_penalty": 1.0, | |
| "max_length": 62, | |
| "min_length": 11, | |
| "num_beams": 6 | |
| } | |
| }, | |
| "torch_dtype": "float32", | |
| "tpu_cores": null, | |
| "track_grad_norm": -1, | |
| "train_check_interval": 0.2, | |
| "train_data_dir": null, | |
| "train_file": "merge/cochrane_sents_train.csv", | |
| "train_split": 0.9, | |
| "train_workers": 8, | |
| "transformers_version": "4.29.1", | |
| "use_cache": true, | |
| "val_check_interval": null, | |
| "val_file": "merge/cochrane_sents_val.csv", | |
| "val_metric": "bleu", | |
| "val_split": 0.05, | |
| "val_workers": 8, | |
| "valid_data_dir": null, | |
| "vocab_size": 50275, | |
| "wandb_id": null, | |
| "x_col": "complex", | |
| "y_col": "simple" | |
| } | |