|
{ |
|
"_name_or_path": "openai-community/gpt2", |
|
"activation_function": "gelu_new", |
|
"architectures": [ |
|
"GPT2ForSequenceClassification" |
|
], |
|
"attn_pdrop": 0.1, |
|
"bos_token_id": 50256, |
|
"embd_pdrop": 0.1, |
|
"eos_token_id": 50256, |
|
"id2label": {"0": "af", |
|
"1": "am", |
|
"2": "ar", |
|
"3": "as", |
|
"4": "az", |
|
"5": "ba", |
|
"6": "be", |
|
"7": "bg", |
|
"8": "bn", |
|
"9": "ca", |
|
"10": "ceb", |
|
"11": "ckb", |
|
"12": "cs", |
|
"13": "cy", |
|
"14": "da", |
|
"15": "de", |
|
"16": "dv", |
|
"17": "el", |
|
"18": "en", |
|
"19": "eo", |
|
"20": "es", |
|
"21": "et", |
|
"22": "eu", |
|
"23": "fa", |
|
"24": "fi", |
|
"25": "fr", |
|
"26": "fy", |
|
"27": "ga", |
|
"28": "gd", |
|
"29": "gl", |
|
"30": "gu", |
|
"31": "he", |
|
"32": "hi", |
|
"33": "hr", |
|
"34": "hu", |
|
"35": "hy", |
|
"36": "id", |
|
"37": "is", |
|
"38": "it", |
|
"39": "ja", |
|
"40": "ka", |
|
"41": "kk", |
|
"42": "kn", |
|
"43": "ku", |
|
"44": "ky", |
|
"45": "la", |
|
"46": "lb", |
|
"47": "lt", |
|
"48": "lv", |
|
"49": "mg", |
|
"50": "mk", |
|
"51": "ml", |
|
"52": "mn", |
|
"53": "mr", |
|
"54": "mt", |
|
"55": "my", |
|
"56": "nds", |
|
"57": "ne", |
|
"58": "nl", |
|
"59": "nn", |
|
"60": "no", |
|
"61": "or", |
|
"62": "pa", |
|
"63": "pl", |
|
"64": "ps", |
|
"65": "pt", |
|
"66": "ro", |
|
"67": "ru", |
|
"68": "sah", |
|
"69": "sd", |
|
"70": "si", |
|
"71": "sk", |
|
"72": "sl", |
|
"73": "sq", |
|
"74": "sr", |
|
"75": "sv", |
|
"76": "sw", |
|
"77": "ta", |
|
"78": "te", |
|
"79": "tg", |
|
"80": "th", |
|
"81": "tk", |
|
"82": "tl", |
|
"83": "tr", |
|
"84": "tt", |
|
"85": "ug", |
|
"86": "uk", |
|
"87": "ur", |
|
"88": "vi", |
|
"89": "yi"}, |
|
"initializer_range": 0.02, |
|
"label2id": { |
|
"af": 0, |
|
"am": 1, |
|
"ar": 2, |
|
"as": 3, |
|
"az": 4, |
|
"ba": 5, |
|
"be": 6, |
|
"bg": 7, |
|
"bn": 8, |
|
"ca": 9, |
|
"ceb": 10, |
|
"ckb": 11, |
|
"cs": 12, |
|
"cy": 13, |
|
"da": 14, |
|
"de": 15, |
|
"dv": 16, |
|
"el": 17, |
|
"en": 18, |
|
"eo": 19, |
|
"es": 20, |
|
"et": 21, |
|
"eu": 22, |
|
"fa": 23, |
|
"fi": 24, |
|
"fr": 25, |
|
"fy": 26, |
|
"ga": 27, |
|
"gd": 28, |
|
"gl": 29, |
|
"gu": 30, |
|
"he": 31, |
|
"hi": 32, |
|
"hr": 33, |
|
"hu": 34, |
|
"hy": 35, |
|
"id": 36, |
|
"is": 37, |
|
"it": 38, |
|
"ja": 39, |
|
"ka": 40, |
|
"kk": 41, |
|
"kn": 42, |
|
"ku": 43, |
|
"ky": 44, |
|
"la": 45, |
|
"lb": 46, |
|
"lt": 47, |
|
"lv": 48, |
|
"mg": 49, |
|
"mk": 50, |
|
"ml": 51, |
|
"mn": 52, |
|
"mr": 53, |
|
"mt": 54, |
|
"my": 55, |
|
"nds": 56, |
|
"ne": 57, |
|
"nl": 58, |
|
"nn": 59, |
|
"no": 60, |
|
"or": 61, |
|
"pa": 62, |
|
"pl": 63, |
|
"ps": 64, |
|
"pt": 65, |
|
"ro": 66, |
|
"ru": 67, |
|
"sah": 68, |
|
"sd": 69, |
|
"si": 70, |
|
"sk": 71, |
|
"sl": 72, |
|
"sq": 73, |
|
"sr": 74, |
|
"sv": 75, |
|
"sw": 76, |
|
"ta": 77, |
|
"te": 78, |
|
"tg": 79, |
|
"th": 80, |
|
"tk": 81, |
|
"tl": 82, |
|
"tr": 83, |
|
"tt": 84, |
|
"ug": 85, |
|
"uk": 86, |
|
"ur": 87, |
|
"vi": 88, |
|
"yi": 89 |
|
}, |
|
"layer_norm_epsilon": 1e-05, |
|
"model_type": "gpt2", |
|
"n_ctx": 1024, |
|
"n_embd": 768, |
|
"n_head": 12, |
|
"n_inner": null, |
|
"n_layer": 12, |
|
"n_positions": 1024, |
|
"pad_token_id": 50256, |
|
"problem_type": "single_label_classification", |
|
"reorder_and_upcast_attn": false, |
|
"resid_pdrop": 0.1, |
|
"scale_attn_by_inverse_layer_idx": false, |
|
"scale_attn_weights": true, |
|
"summary_activation": null, |
|
"summary_first_dropout": 0.1, |
|
"summary_proj_to_labels": true, |
|
"summary_type": "cls_index", |
|
"summary_use_proj": true, |
|
"task_specific_params": { |
|
"text-generation": { |
|
"do_sample": true, |
|
"max_length": 50 |
|
} |
|
}, |
|
"torch_dtype": "float32", |
|
"transformers_version": "4.36.2", |
|
"use_cache": true, |
|
"vocab_size": 50257 |
|
} |
|
|