CoolFace
Modelpublic

djsull/aha_sentence_classification

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face
0likes6downloads
config.json83 linesDownload Raw Back to root
1{2  "architectures": [3    "ModernBertForSequenceClassification"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "bos_token_id": 0,8  "classifier_activation": "gelu",9  "classifier_bias": false,10  "classifier_dropout": 0.0,11  "classifier_pooling": "mean",12  "cls_token_id": 0,13  "decoder_bias": true,14  "deterministic_flash_attn": false,15  "dtype": "float32",16  "embedding_dropout": 0.0,17  "eos_token_id": 1,18  "global_attn_every_n_layers": 3,19  "global_rope_theta": 160000,20  "gradient_checkpointing": false,21  "hidden_activation": "gelu",22  "hidden_size": 768,23  "id2label": {24    "0": "\ud575\uc2ec\uc9c8\ubb38",25    "1": "\uc11c\ube0c\uc9c8\ubb38",26    "2": "\ubc30\uacbd/\uc0c1\ud669",27    "3": "\uc2dc\ub3c4/\ud574\ubcf8\uac83",28    "4": "\uc624\ub958/\uc99d\uc0c1",29    "5": "\ud575\uc2ec\ub2f5\ubcc0/\uacb0\ub860",30    "6": "\uc778\uc6a9/\ucc38\uc870",31    "7": "\uadfc\uac70/\uc124\uba85",32    "8": "\uc808\ucc28/\ubc29\ubc95",33    "9": "\uc8fc\uc758/\ud55c\uacc4",34    "10": "\ub300\uc548/\ube44\uad50",35    "11": "\ucd94\uac00\uc815\ubcf4\uc694\uccad",36    "12": "\uc778\uc0ac/\uac10\uc0ac/\uc751\uc6d0",37    "13": "\ucd94\ucc9c/\ubcf4\uc0c1\uc720\ub3c4",38    "14": "\uc608\uc2dc\ub370\uc774\ud130/\uc99d\ube59",39    "15": "\uc758\ubbf8/\ud574\ub2f9\uc0ac\ud56d\uc5c6\uc74c"40  },41  "initializer_cutoff_factor": 2.0,42  "initializer_range": 0.02,43  "intermediate_size": 1152,44  "label2id": {45    "\uadfc\uac70/\uc124\uba85": 7,46    "\ub300\uc548/\ube44\uad50": 10,47    "\ubc30\uacbd/\uc0c1\ud669": 2,48    "\uc11c\ube0c\uc9c8\ubb38": 1,49    "\uc2dc\ub3c4/\ud574\ubcf8\uac83": 3,50    "\uc608\uc2dc\ub370\uc774\ud130/\uc99d\ube59": 14,51    "\uc624\ub958/\uc99d\uc0c1": 4,52    "\uc758\ubbf8/\ud574\ub2f9\uc0ac\ud56d\uc5c6\uc74c": 15,53    "\uc778\uc0ac/\uac10\uc0ac/\uc751\uc6d0": 12,54    "\uc778\uc6a9/\ucc38\uc870": 6,55    "\uc808\ucc28/\ubc29\ubc95": 8,56    "\uc8fc\uc758/\ud55c\uacc4": 9,57    "\ucd94\uac00\uc815\ubcf4\uc694\uccad": 11,58    "\ucd94\ucc9c/\ubcf4\uc0c1\uc720\ub3c4": 13,59    "\ud575\uc2ec\ub2f5\ubcc0/\uacb0\ub860": 5,60    "\ud575\uc2ec\uc9c8\ubb38": 061  },62  "layer_norm_eps": 1e-05,63  "local_attention": 128,64  "local_rope_theta": 10000.0,65  "max_position_embeddings": 16384,66  "mlp_bias": false,67  "mlp_dropout": 0.0,68  "model_type": "modernbert",69  "norm_bias": false,70  "norm_eps": 1e-05,71  "num_attention_heads": 12,72  "num_hidden_layers": 22,73  "pad_token_id": 49999,74  "position_embedding_type": "absolute",75  "problem_type": "single_label_classification",76  "repad_logits_with_grad": false,77  "sep_token_id": 1,78  "sparse_pred_ignore_index": -100,79  "sparse_prediction": false,80  "transformers_version": "4.56.1",81  "vocab_size": 5000082}83