| { | |
| "architectures": [ | |
| "ToyAutoEncoder" | |
| ], | |
| "decoder_specs": { | |
| "activation": "gelu", | |
| "model_type": "", | |
| "residual": true, | |
| "single_layer": false, | |
| "type": "mlp", | |
| "with_bias": false | |
| }, | |
| "emb_structure": "all_onehot", | |
| "hidden_dim": 32, | |
| "intermediate_ratio": 1, | |
| "max_pos": 30, | |
| "model_type": "toy_autoencoder", | |
| "num_attention_heads": 1, | |
| "num_decoder_layers": 1, | |
| "num_encoder_layers": 3, | |
| "shared": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.37.2", | |
| "vocab_size": 2 | |
| } | |