{ "activation_dropout": 0.0, "activation_function": "relu", "add_cross_attention": true, "architectures": [ "TrOCRForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 0, "classifier_dropout": 0.0, "cross_attention_hidden_size": 8, "d_model": 8, "decoder_attention_heads": 2, "decoder_ffn_dim": 8, "decoder_layerdrop": 0.0, "decoder_layers": 2, "decoder_start_token_id": 2, "dropout": 0.1, "eos_token_id": 2, "init_std": 0.02, "is_decoder": true, "layernorm_embedding": true, "max_position_embeddings": 512, "model_type": "trocr", "pad_token_id": 1, "scale_embedding": true, "tie_word_embeddings": false, "torch_dtype": "float32", "transformers_version": "4.36.0", "use_cache": false, "use_learned_position_embeddings": true, "vocab_size": 64044 }