{ "src_vocab_size": 20000, "valid_metrics": [ "BLEU" ], "src_vocab": "hu.eole.vocab", "tensorboard": true, "tensorboard_log_dir": "tensorboard", "report_every": 100, "transforms": [ "sentencepiece", "filtertoolong" ], "tgt_vocab": "en.eole.vocab", "overwrite": true, "seed": 1234, "share_vocab": false, "save_data": "data", "tgt_vocab_size": 20000, "tensorboard_log_dir_dated": "tensorboard/Aug-24_09-05-34", "vocab_size_multiple": 8, "n_sample": 0, "training": { "adam_beta2": 0.998, "normalization": "tokens", "save_checkpoint_steps": 5000, "optim": "adamw", "dropout": [ 0.1 ], "param_init_method": "xavier_uniform", "world_size": 1, "valid_steps": 5000, "average_decay": 0.0001, "learning_rate": 2.0, "prefetch_factor": 32, "accum_steps": [ 0 ], "attention_dropout": [ 0.1 ], "decay_method": "noam", "bucket_size": 128000, "compute_dtype": "torch.float16", "batch_size": 8000, "accum_count": [ 10 ], "dropout_steps": [ 0 ], "keep_checkpoint": 4, "valid_batch_size": 4096, "num_workers": 0, "max_grad_norm": 0.0, "gpu_ranks": [ 0 ], "model_path": "quickmt-hu-en-eole-model", "batch_type": "tokens", "label_smoothing": 0.1, "batch_size_multiple": 8, "train_steps": 100000, "warmup_steps": 4000 }, "data": { "corpus_1": { "path_src": "train.hu", "transforms": [ "sentencepiece", "filtertoolong" ], "path_tgt": "train.en", "path_align": null }, "valid": { "path_src": "valid.hu", "transforms": [ "sentencepiece", "filtertoolong" ], "path_tgt": "valid.en", "path_align": null } }, "transforms_configs": { "filtertoolong": { "src_seq_length": 256, "tgt_seq_length": 256 }, "sentencepiece": { "tgt_subword_model": "${MODEL_PATH}/en.spm.model", "src_subword_model": "${MODEL_PATH}/hu.spm.model" } }, "model": { "share_decoder_embeddings": false, "heads": 8, "architecture": "transformer", "hidden_size": 1024, "transformer_ff": 4096, "share_embeddings": false, "position_encoding_type": "SinusoidalInterleaved", "encoder": { "heads": 8, "hidden_size": 1024, "encoder_type": "transformer", "n_positions": null, "position_encoding_type": "SinusoidalInterleaved", "src_word_vec_size": 1024, "layers": 8, "transformer_ff": 4096 }, "decoder": { "heads": 8, "hidden_size": 1024, "n_positions": null, "tgt_word_vec_size": 1024, "position_encoding_type": "SinusoidalInterleaved", "decoder_type": "transformer", "layers": 2, "transformer_ff": 4096 }, "embeddings": { "src_word_vec_size": 1024, "word_vec_size": 1024, "tgt_word_vec_size": 1024, "position_encoding_type": "SinusoidalInterleaved" } } }