| { | |
| "architectures": [ | |
| "Transformer" | |
| ], | |
| "aux_loss_alpha": 0.01, | |
| "dim": 1024, | |
| "dropout": 0.0, | |
| "flash_attn": true, | |
| "hidden_dim": null, | |
| "max_seq_len": 1024, | |
| "model_type": "minimind", | |
| "multiple_of": 64, | |
| "n_heads": 16, | |
| "n_kv_heads": 8, | |
| "n_layers": 16, | |
| "n_routed_experts": 4, | |
| "n_shared_experts": true, | |
| "norm_eps": 1e-05, | |
| "norm_topk_prob": true, | |
| "num_experts_per_tok": 2, | |
| "scoring_func": "softmax", | |
| "seq_aux": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.45.2", | |
| "use_moe": false, | |
| "vocab_size": 6400 | |
| } | |