| { | |
| "model_type": "frontd", | |
| "architectures": [ | |
| "FrontD11M" | |
| ], | |
| "model_name": "FrontD.11m", | |
| "vocab_size": 16000, | |
| "context_length": 1096, | |
| "hidden_size": 256, | |
| "num_hidden_layers": 6, | |
| "num_attention_heads": 16, | |
| "intermediate_size": 1152, | |
| "position_embedding_type": "rope", | |
| "rope_theta": 10000.0, | |
| "normalization": "rmsnorm", | |
| "activation_function": "silu", | |
| "attention_bias": false, | |
| "mlp_bias": false, | |
| "tie_word_embeddings": true, | |
| "dropout": 0.0, | |
| "bos_token": "<bos>", | |
| "eos_token": "<eos>", | |
| "pad_token": "<pad>", | |
| "unk_token": "<unk>", | |
| "parameter_target": "~11M", | |
| "parameter_count": 10996608, | |
| "dataset": "Plans11/Organized_PreTrain_1k_Context", | |
| "session_examples": 100000, | |
| "training_from_scratch": true, | |
| "license": "mit" | |
| } |