| { |
| "_mlp_class": "LLaMAMLP", |
| "_norm_class": "FusedRMSNorm", |
| "average_2": false, |
| "betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "bias": false, |
| "block_size": 8224, |
| "block_size_val": 2080, |
| "condense_ratio": 1, |
| "decay_lr": true, |
| "discount": true, |
| "eval_iters": 100, |
| "eval_step_interval": 1000, |
| "forecasting_patch": 1, |
| "global_batch_size": 512, |
| "grad_clip": 1.0, |
| "group": "70m-test_all", |
| "haar_loss_match": false, |
| "haar_trans": true, |
| "haar_trans_inv": true, |
| "haar_trans_norm": "backward", |
| "half_diff": false, |
| "imputation": false, |
| "inner_norm": false, |
| "inter_control": false, |
| "intermediate_size": 3072, |
| "is_diff": false, |
| "is_smape": false, |
| "learning_rate": 0.0005, |
| "log_step_interval": 10, |
| "max_step": 100000, |
| "mean_replace": false, |
| "micro_batch_size": 64, |
| "micro_batch_size_val": 256, |
| "min_lr": 1e-05, |
| "mix_train": false, |
| "multi_loss": false, |
| "n_cot": 1, |
| "n_embd": 768, |
| "n_head": 12, |
| "n_layer": 12, |
| "n_query_groups": 4, |
| "name": "small-100k", |
| "new_arch": false, |
| "new_tokenizer": false, |
| "norm_eps": 1e-05, |
| "num_of_devices": 4, |
| "num_of_nodes": 2, |
| "org": "Ali-Could", |
| "ou": false, |
| "ou_mean": false, |
| "ou_prod": false, |
| "padded_vocab_size": "None", |
| "padding_multiple": 1, |
| "parallel_residual": false, |
| "patch_size": 32, |
| "pid": false, |
| "quantitle": true, |
| "rollback_win": 256, |
| "rolling_patch": 6, |
| "rope_base": 10000, |
| "rotary_percentage": 1.0, |
| "save_step_interval": 10000, |
| "scaling": true, |
| "seed0": 3407, |
| "shared_attention_norm": false, |
| "stats_encoding": false, |
| "stats_encoding_new": false, |
| "sum_divided": false, |
| "triple_diff": false, |
| "triple_diff_new": false, |
| "unet": true, |
| "vocab_size": 1, |
| "vq": false, |
| "warmup_steps": 2000, |
| "weight_decay": 0.1, |
| "yj_trans": false |
| } |
