{ "name": "C2", "description": "RoPE instead of sinusoidal absolute", "model": { "dim_model": 128, "num_heads": 4, "num_encoder_layers": 2, "num_decoder_layers": 2, "dim_ff": 512, "dropout": 0.1, "norm_eps": 1e-06, "num_kv_heads": 2, "rope_base": 10000.0, "pos_encoding": "rope", "attention": "mha", "norm": "layernorm", "tokenization": "subword", "tie_output_embedding": false }, "data": { "cipher_path": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/dataset/brown_cipher.txt", "plain_path": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/dataset/brown_plain.txt", "split_ratios": [ 0.8, 0.1, 0.1 ], "split_seed": 1234, "max_src_len_subword": 768, "max_tgt_len_subword": 256, "max_src_len_byte": 3072, "max_tgt_len_byte": 1152, "length_policy": "drop", "effective_batch_size": 64, "micro_batch_size_subword": 64, "micro_batch_size_byte": 64, "num_workers": 0, "verbose": true, "tokenizer": { "src_vocab_size": 4096, "tgt_vocab_size": 4096, "src_seed_size": 20000, "tgt_seed_size": 50000, "src_max_segment": 16, "tgt_max_segment": 32, "max_piece_len": 16, "shrink_factor": 0.75, "em_iters": 2, "min_freq": 2, "sample_seed": 1234, "cache_dir": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/outputs/tokenizers", "retrain": false }, "window": { "enabled": true, "src_window": 1024, "src_stride": 768, "min_tgt_chars": 16, "eval_full_sentences": true } }, "train": { "epochs": 100, "lr": 0.0007, "betas": [ 0.9, 0.98 ], "adam_eps": 1e-09, "weight_decay": 0.01, "grad_clip": 1.0, "label_smoothing": 0.05, "scheduler": "cosine", "warmup_steps": 400, "min_lr": 1e-05, "seed": 26, "device": "auto", "amp": true, "eval_every_epochs": 1, "generate_every_epochs": 5, "max_eval_batches": null, "length_slack": 16, "gen_eval_examples": 128, "gen_batch_size": 32, "test_gen_examples": null, "log_every_steps": 50, "checkpoint_dir": "/home2/winterdew/ANLP_A1/ANLP-M26-Assignment-1/outputs/checkpoints", "keep_best_only": true, "early_stopping_patience": null, "wandb_project": "anlp-a1-transformers", "wandb_entity": null, "wandb_mode": "online" }, "blt": { "patch_size": 8, "local_dim": 128, "local_heads": 4, "local_dim_ff": 512, "local_encoder_layers": 1, "local_decoder_layers": 1, "local_dropout": 0.1 }, "derived": { "token_free": false, "max_src_len": 1026, "max_tgt_len": 1026, "micro_batch_size": 64, "windowed": true, "grad_accum_steps": 1, "pe_max_len": 1042, "head_dim": 32 } }