{ "architectures": [ "GPTNeoXJapaneseModel" ], "attention_dropout": 0.1, "bos_token_id": 31996, "dtype": "float32", "eos_token_id": 31999, "hidden_act": "gelu", "hidden_dropout": 0.0, "hidden_size": 32, "initializer_range": 0.02, "intermediate_multiple_size": 4, "is_decoder": false, "layer_norm_eps": 1e-05, "max_position_embeddings": 512, "model_type": "gpt_neox_japanese", "num_attention_heads": 4, "num_hidden_layers": 2, "pad_token_id": null, "rope_parameters": { "partial_rotary_factor": 1.0, "rope_theta": 10000.0, "rope_type": "default" }, "tie_word_embeddings": true, "transformers_version": "5.16.0.dev0", "type_vocab_size": 16, "use_cache": true, "vocab_size": 32000, "weight_tying": true }