| { | |
| "exp_name":"mt_enfrde", | |
| "exp_id":"demo", | |
| "dump_path":"/content/dump_path", | |
| "encoder_only":"False", | |
| "data_path":"/content/processed", | |
| "lgs":"de-en-fr", | |
| "clm_steps":"", | |
| "mlm_steps":"", | |
| "mt_steps":"...", | |
| "ae_steps":"", | |
| "bt_steps":"", | |
| "pc_steps":"", | |
| "reload_emb":"", | |
| "reload_model":"/content/dump_path/mlm_enfrde/demo/best-valid_mlm_ppl.pth,/content/dump_path/mlm_enfrde/demo/best-valid_mlm_ppl.pth", | |
| "reload_checkpoint":"", | |
| "epoch_size":"100", | |
| "max_epoch":"1", | |
| "stopping_criterion":"valid_en-fr_mt_bleu,10", | |
| "validation_metrics":"valid_en-fr_mt_bleu", | |
| "train_n_samples":"-1", | |
| "valid_n_samples":"-1", | |
| "test_n_samples":"-1", | |
| "remove_long_sentences_train":"False", | |
| "remove_long_sentences_valid":"False", | |
| "remove_long_sentences_test":"False", | |
| "eval_bleu":"True", | |
| "eval_only":"False", | |
| "max_len":"100", | |
| "group_by_size":"True", | |
| "batch_size":"32", | |
| "save_periodic":"0", | |
| "fp16":"False", | |
| "amp":"-1", | |
| "emb_dim":"1024", | |
| "n_layers":"6", | |
| "n_heads":"8", | |
| "dropout":"0.1", | |
| "attention_dropout":"0.1", | |
| "gelu_activation":"True", | |
| "share_inout_emb":"True", | |
| "sinusoidal_embeddings":"False", | |
| "use_lang_emb":"True", | |
| "use_memory":"False", | |
| "asm":"False", | |
| "context_size":"0", | |
| "word_pred":"0.15", | |
| "word_mask_keep_rand":"0.8,0.1,0.1", | |
| "word_shuffle":"3", | |
| "word_dropout":"0.1", | |
| "word_blank":"0.1", | |
| "max_vocab":"-1", | |
| "min_count":"0", | |
| "lg_sampling_factor":"-1", | |
| "bptt":"256", | |
| "max_batch_size":"0", | |
| "tokens_per_batch":"2000", | |
| "split_data":"False", | |
| "optimizer":"adam_inverse_sqrt,beta1=0.9,beta2=0.98,lr=0.0001", | |
| "clip_grad_norm":"5", | |
| "accumulate_gradients":"1", | |
| "lambda_mlm":"1", | |
| "lambda_clm":"1", | |
| "lambda_pc":"1", | |
| "lambda_ae":"0:1,100000:0.1,300000:0", | |
| "lambda_mt":"1", | |
| "lambda_bt":"1", | |
| "beam_size":"1", | |
| "length_penalty":"1", | |
| "early_stopping":"False", | |
| "debug_train":"False", | |
| "debug_slurm":"False", | |
| "local_rank":"-1", | |
| "master_port":"-1", | |
| "mem_enc_positions":"", | |
| "mem_dec_positions":"", | |
| "asm_cutoffs":"8000,20000", | |
| "asm_div_value":"4" | |
| } |