{ "_name_or_path": "google/vit-base-patch16-224-in21k", "architectures": [ "ViTForImageClassification" ], "attention_probs_dropout_prob": 0.0, "encoder_stride": 16, "hidden_act": "gelu", "hidden_dropout_prob": 0.0, "hidden_size": 768, "id2label": { "0": "Ses02F_impro01", "1": "Ses02F_impro02", "2": "Ses02F_impro03", "3": "Ses02F_impro04", "4": "Ses02F_impro05", "5": "Ses02F_impro06", "6": "Ses02F_impro07", "7": "Ses02F_impro08", "8": "Ses02F_script01_1", "9": "Ses02F_script01_2", "10": "Ses02F_script01_3", "11": "Ses02F_script02_1", "12": "Ses02F_script02_2", "13": "Ses02F_script03_1", "14": "Ses02F_script03_2", "15": "Ses02M_impro01", "16": "Ses02M_impro02", "17": "Ses02M_impro03", "18": "Ses02M_impro04", "19": "Ses02M_impro05", "20": "Ses02M_impro06", "21": "Ses02M_impro07", "22": "Ses02M_impro08", "23": "Ses02M_script01_1", "24": "Ses02M_script01_2", "25": "Ses02M_script01_3", "26": "Ses02M_script02_1", "27": "Ses02M_script02_2", "28": "Ses02M_script03_1", "29": "Ses02M_script03_2" }, "image_size": 224, "initializer_range": 0.02, "intermediate_size": 3072, "label2id": { "Ses02F_impro01": 0, "Ses02F_impro02": 1, "Ses02F_impro03": 2, "Ses02F_impro04": 3, "Ses02F_impro05": 4, "Ses02F_impro06": 5, "Ses02F_impro07": 6, "Ses02F_impro08": 7, "Ses02F_script01_1": 8, "Ses02F_script01_2": 9, "Ses02F_script01_3": 10, "Ses02F_script02_1": 11, "Ses02F_script02_2": 12, "Ses02F_script03_1": 13, "Ses02F_script03_2": 14, "Ses02M_impro01": 15, "Ses02M_impro02": 16, "Ses02M_impro03": 17, "Ses02M_impro04": 18, "Ses02M_impro05": 19, "Ses02M_impro06": 20, "Ses02M_impro07": 21, "Ses02M_impro08": 22, "Ses02M_script01_1": 23, "Ses02M_script01_2": 24, "Ses02M_script01_3": 25, "Ses02M_script02_1": 26, "Ses02M_script02_2": 27, "Ses02M_script03_1": 28, "Ses02M_script03_2": 29 }, "layer_norm_eps": 1e-12, "model_type": "vit", "num_attention_heads": 12, "num_channels": 3, "num_hidden_layers": 12, "patch_size": 16, "problem_type": "single_label_classification", "qkv_bias": true, "torch_dtype": "float32", "transformers_version": "4.47.0" }