| { | |
| "n_head": 32, | |
| "n_vocab": 50257, | |
| "embed_dropout": 0, | |
| "lr": 0.00016, | |
| "lr_decay": "cosine", | |
| "warmup_steps": 3000, | |
| "beta1": 0.9, | |
| "beta2": 0.95, | |
| "epsilon": 1e-8, | |
| "ada_epsilon1": 1e-30, | |
| "ada_epsilon2": 1e-3, | |
| "opt_name": "adam", | |
| "weight_decay": 0.10, | |
| "train_batch_size": 512, | |
| "attn_dropout": 0, | |
| "train_steps": 286150, | |
| "eval_steps": 0, | |
| "predict_steps": 1, | |
| "res_dropout": 0, | |
| "eval_batch_size": 128, | |
| "predict_batch_size": 1, | |
| "iterations": 500, | |
| "n_embd": 2560, | |
| "datasets": [["openwebtext-documents", 25, "documents_random", 1.0]], | |
| "model_path": "gs://neo-models/GPT3_2-7B", | |
| "n_ctx": 2048, | |
| "n_layer": 32, | |
| "scale_by_depth": true, | |
| "scale_by_in": false, | |
| "attention_types" : [[["global"],32]], | |
| "mesh_shape": "x:128,y:2", | |
| "layout": "embd:y,batch:x", | |
| "activation_function": "gelu", | |
| "recompute_grad": true, | |
| "gradient_clipping": 1.0 | |
| } | |