{ "tgt_vocab_size": 20000, "n_sample": 0, "src_vocab_size": 20000, "valid_metrics": [ "BLEU" ], "tgt_vocab": "en.eole.vocab", "tensorboard_log_dir": "tensorboard", "tensorboard_log_dir_dated": "tensorboard/May-21_06-04-15", "seed": 1234, "transforms": [ "sentencepiece", "filtertoolong" ], "overwrite": true, "tensorboard": true, "src_vocab": "bg.eole.vocab", "share_vocab": false, "vocab_size_multiple": 8, "save_data": "data", "report_every": 100, "training": { "gpu_ranks": [ 0 ], "batch_size_multiple": 8, "world_size": 1, "accum_steps": [ 0 ], "compute_dtype": "torch.float16", "num_workers": 0, "dropout": [ 0.1 ], "learning_rate": 2.0, "save_checkpoint_steps": 5000, "max_grad_norm": 0.0, "valid_batch_size": 4096, "batch_size": 8000, "train_steps": 100000, "batch_type": "tokens", "average_decay": 0.0001, "prefetch_factor": 32, "accum_count": [ 10 ], "model_path": "quickmt-bg-en-eole-model", "valid_steps": 5000, "optim": "adamw", "attention_dropout": [ 0.1 ], "label_smoothing": 0.1, "warmup_steps": 4000, "dropout_steps": [ 0 ], "normalization": "tokens", "bucket_size": 128000, "keep_checkpoint": 4, "decay_method": "noam", "adam_beta2": 0.998, "param_init_method": "xavier_uniform" }, "transforms_configs": { "filtertoolong": { "tgt_seq_length": 256, "src_seq_length": 256 }, "sentencepiece": { "tgt_subword_model": "${MODEL_PATH}/en.spm.model", "src_subword_model": "${MODEL_PATH}/bg.spm.model" } }, "data": { "corpus_1": { "path_align": null, "transforms": [ "sentencepiece", "filtertoolong" ], "path_tgt": "train.en", "path_src": "train.bg" }, "valid": { "path_align": null, "transforms": [ "sentencepiece", "filtertoolong" ], "path_tgt": "dev.en", "path_src": "dev.bg" } }, "model": { "transformer_ff": 4096, "architecture": "transformer", "heads": 8, "share_decoder_embeddings": false, "share_embeddings": false, "position_encoding_type": "SinusoidalInterleaved", "hidden_size": 1024, "decoder": { "decoder_type": "transformer", "layers": 2, "transformer_ff": 4096, "heads": 8, "tgt_word_vec_size": 1024, "n_positions": null, "position_encoding_type": "SinusoidalInterleaved", "hidden_size": 1024 }, "encoder": { "layers": 8, "transformer_ff": 4096, "heads": 8, "src_word_vec_size": 1024, "encoder_type": "transformer", "n_positions": null, "position_encoding_type": "SinusoidalInterleaved", "hidden_size": 1024 }, "embeddings": { "word_vec_size": 1024, "src_word_vec_size": 1024, "position_encoding_type": "SinusoidalInterleaved", "tgt_word_vec_size": 1024 } } }