{ "architectures": [ "Transformer" ], "bilinear": false, "d_hidden": 2048, "d_model": 512, "gate": true, "modifier": null, "n_ctx": 256, "n_head": 8, "n_layer": 6, "n_vocab": 4096, "noise": null, "normalization": true, "torch_dtype": "float32", "transformers_version": "4.39.0" }