feat: FaQUAD-NLI

faquad-nli.py

@@ -0,0 +1,124 @@
+"""FaQUAD-NLI dataset"""
+
+import datasets
+import pandas as pd
+import json
+
+_CITATION = """
+"""
+
+_DESCRIPTION = """
+"""
+
+_URLS = {
+    "data": "https://raw.githubusercontent.com/liafacom/faquad/6ad978f20672bb41625b3b71fbe4a88b893d0a86/data/dataset.json"
+}
+
+def check_overlap(interval1, interval2):
+    """Check for overlap between two integer intervals"""
+    return not (interval1[1] < interval2[0] or interval2[1] < interval1[0])
+
+
+class Boun(datasets.GeneratorBasedBuilder):
+
+    VERSION = datasets.Version("1.0.0")
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "document_index": datasets.Value("int32"),
+                    "document_title": datasets.Value("string"),
+                    "paragraph_index": datasets.Value("int32"),
+                    "question": datasets.Value("string"),
+                    "answer": datasets.Value("string"),
+                    "label": datasets.Value("int32")
+                }),
+            supervised_keys=None,
+            homepage="https://github.com/franciellevargas/HateBR",
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        downloaded_files = dl_manager.download(_URLS)
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "data": downloaded_files["data"],
+                    "split": "train"
+                }
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={
+                    "data": downloaded_files["data"],
+                    "split": "validation"
+                }
+            ),    
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={
+                    "data": downloaded_files["data"],
+                    "split": "test"
+                }
+            )        
+        ]
+
+    def _generate_examples(self, data, split):
+
+        with open(data, 'r') as f:
+            json_data = json.load(f)
+        
+        spans = pd.read_csv("spans.csv").to_dict("records")
+        counter = 0
+        for span_row in spans:
+            if span_row["split"] != split:
+                continue
+
+            document_title = json_data["data"][
+                span_row["document_index"]
+            ]["title"]
+
+            sentence = json_data["data"][
+                span_row["document_index"]
+            ]["paragraphs"][
+                span_row["paragraph_index"]
+            ]["context"][
+                span_row["sentence_start_char"]:span_row["sentence_end_char"]
+            ]
+            sentence_interval = (span_row["sentence_start_char"], span_row["sentence_end_char"])
+
+            for qas_row in json_data["data"][
+                span_row["document_index"]
+            ]["paragraphs"][
+                span_row["paragraph_index"]
+            ]["qas"]:
+                question = qas_row["question"]
+                question_spans = []
+                for qas_answer in qas_row["answers"]:
+                    qas_answer_start_span = qas_answer["answer_start"]
+                    qas_answer_end_span = qas_answer["answer_start"] + len(qas_answer["text"])
+                    question_spans.append((qas_answer_start_span, qas_answer_end_span))
+                for question_interval in question_spans:
+                    if check_overlap(sentence_interval, question_interval):
+                        yield counter, {
+                            "document_index": span_row["document_index"],
+                            "document_title": document_title,
+                            "paragraph_index": span_row["paragraph_index"],
+                            "question": question,
+                            "answer": sentence,
+                            "label": 1
+                        }
+                        counter += 1
+                        break
+                else:
+                    yield counter, {
+                        "document_index": span_row["document_index"],
+                        "document_title": document_title,
+                        "paragraph_index": span_row["paragraph_index"],
+                        "question": question,
+                        "answer": sentence,
+                        "label": 0
+                    }
+                    counter += 1

sentence_spans.py

@@ -0,0 +1,51 @@
+import spacy
+import json
+import requests
+import pandas as pd
+
+DATASET_URL = "https://raw.githubusercontent.com/liafacom/faquad/6ad978f20672bb41625b3b71fbe4a88b893d0a86/data/dataset.json"
+
+class SentenceModel(object):
+    def __init__(self, model_name):
+        self.nlp = spacy.load(model_name)
+
+    def parse(self, text):
+        with self.nlp.select_pipes(enable=['tok2vec', "parser", "senter"]):
+            doc = self.nlp(text)
+            sentences = [ (sentence.start_char, sentence.end_char) for sentence in doc.sents ]
+        return sentences
+
+def context_generator(url):
+    response = requests.get(url)
+    data = json.loads(response.text)["data"]
+    for idx, row in enumerate(data):
+        for paragraph_idx, paragraph_row in enumerate(row["paragraphs"]):
+            context = paragraph_row["context"]
+            yield idx, paragraph_idx, context
+
+def define_split(document_index):
+    if document_index % 5 == 0:
+        return "test"
+    elif document_index % 5 == 1:
+        return "validation"
+    else:
+        return "train"
+
+def main():
+    df = []
+    model = SentenceModel("pt_core_news_sm")
+    for idx, paragraph_idx, context in context_generator(DATASET_URL):
+        for start_char, end_char in model.parse(context):
+            row = {
+                "document_index": idx,
+                "paragraph_index": paragraph_idx,
+                "sentence_start_char": start_char,
+                "sentence_end_char": end_char,
+                "split": define_split(idx)
+            }
+            df.append(row)
+    df = pd.DataFrame(df)
+    df.to_csv("spans.csv", index=False)
+
+if __name__ == "__main__":
+    main()

spans.csv

@@ -0,0 +1,1175 @@
+document_index,paragraph_index,sentence_start_char,sentence_end_char,split
+0,0,0,317,test
+0,0,318,669,test
+0,0,670,935,test
+0,0,936,1155,test
+0,0,1155,1177,test
+0,1,0,183,test
+0,1,184,321,test
+0,1,322,475,test
+0,1,476,746,test
+0,1,747,909,test
+0,1,910,1328,test
+0,2,0,491,test
+0,2,492,658,test
+0,2,659,793,test
+0,2,794,959,test
+0,3,0,132,test
+0,3,133,243,test
+0,3,244,351,test
+0,3,352,538,test
+0,3,539,629,test
+0,3,630,813,test
+0,3,814,1071,test
+0,4,0,151,test
+0,4,152,409,test
+0,4,410,827,test
+0,4,828,1105,test
+0,4,1106,1312,test
+1,0,0,117,validation
+1,0,118,216,validation
+1,0,217,272,validation
+1,0,273,325,validation
+1,0,326,522,validation
+1,0,523,641,validation
+1,0,642,777,validation
+1,0,778,907,validation
+1,1,0,193,validation
+1,1,194,256,validation
+1,1,257,313,validation
+1,1,314,384,validation
+1,1,385,486,validation
+1,1,487,607,validation
+1,2,0,121,validation
+1,2,122,378,validation
+1,2,379,434,validation
+1,2,435,621,validation
+1,2,622,823,validation
+1,3,0,54,validation
+1,3,55,118,validation
+1,3,119,248,validation
+1,3,249,333,validation
+1,3,334,456,validation
+1,3,457,517,validation
+1,3,518,598,validation
+1,3,599,866,validation
+1,3,867,946,validation
+1,3,947,1021,validation
+1,3,1022,1143,validation
+1,3,1144,1219,validation
+1,4,0,110,validation
+1,4,111,191,validation
+1,4,192,319,validation
+1,4,320,490,validation
+1,4,491,549,validation
+1,4,550,760,validation
+1,4,761,887,validation
+1,4,888,1048,validation
+1,4,1049,1342,validation
+1,5,0,131,validation
+1,5,132,278,validation
+1,5,279,427,validation
+1,5,428,533,validation
+1,5,534,724,validation
+1,5,725,838,validation
+1,6,0,151,validation
+1,6,152,251,validation
+1,6,252,429,validation
+1,7,0,72,validation
+1,7,73,307,validation
+1,7,308,494,validation
+1,7,495,619,validation
+1,7,620,640,validation
+1,8,0,51,validation
+1,8,52,108,validation
+1,8,109,183,validation
+1,8,184,241,validation
+1,8,242,300,validation
+1,8,301,502,validation
+1,8,503,715,validation
+1,8,716,761,validation
+1,8,762,868,validation
+1,8,869,914,validation
+2,0,0,145,train
+2,0,146,203,train
+2,0,204,280,train
+2,0,281,294,train
+2,0,295,459,train
+2,0,460,613,train
+2,0,614,766,train
+2,0,767,935,train
+2,1,0,127,train
+2,1,128,223,train
+2,1,224,451,train
+2,1,452,616,train
+2,1,617,758,train
+2,2,0,173,train
+2,2,174,343,train
+2,2,344,611,train
+2,2,612,778,train
+2,2,779,895,train
+2,3,0,101,train
+2,3,102,193,train
+2,3,193,288,train
+2,3,289,441,train
+2,3,442,667,train
+2,4,0,226,train
+2,4,227,287,train
+2,4,288,340,train
+2,4,341,528,train
+2,4,529,590,train
+2,4,591,718,train
+2,5,0,190,train
+2,5,191,557,train
+2,5,558,733,train
+2,6,0,158,train
+2,6,159,237,train
+2,6,238,518,train
+2,6,519,694,train
+2,7,0,435,train
+2,7,436,521,train
+2,7,522,755,train
+2,8,0,118,train
+2,8,119,294,train
+2,8,295,410,train
+2,8,411,555,train
+2,8,556,631,train
+2,9,0,125,train
+2,9,126,339,train
+2,9,340,411,train
+2,9,412,440,train
+2,9,441,580,train
+2,9,581,654,train
+2,10,0,214,train
+2,10,215,483,train
+2,11,0,140,train
+2,11,141,408,train
+2,11,409,623,train
+2,11,624,768,train
+2,11,769,943,train
+2,12,0,100,train
+2,12,101,251,train
+2,12,252,422,train
+2,12,423,614,train
+2,13,0,116,train
+2,13,116,248,train
+2,13,249,379,train
+2,13,380,492,train
+2,13,493,633,train
+2,13,634,761,train
+2,13,762,884,train
+2,14,0,99,train
+2,14,100,273,train
+2,14,274,385,train
+2,14,386,506,train
+2,14,507,693,train
+2,14,694,861,train
+2,14,862,980,train
+2,15,0,165,train
+2,15,166,333,train
+2,15,334,461,train
+2,15,462,665,train
+2,16,0,65,train
+2,16,66,264,train
+2,16,265,596,train
+2,16,597,741,train
+3,0,0,85,train
+3,0,86,256,train
+3,0,257,319,train
+3,0,320,458,train
+3,0,459,580,train
+3,0,581,665,train
+3,0,666,772,train
+3,0,773,862,train
+3,0,863,1007,train
+3,1,0,130,train
+3,1,131,262,train
+3,1,263,370,train
+3,1,371,635,train
+3,1,636,770,train
+3,1,771,992,train
+3,2,0,118,train
+3,2,119,218,train
+3,2,219,331,train
+3,2,332,403,train
+3,2,404,753,train
+3,3,0,332,train
+3,3,333,562,train
+3,3,563,742,train
+3,3,743,854,train
+3,4,0,115,train
+3,4,116,345,train
+3,4,346,423,train
+3,4,424,515,train
+3,4,516,595,train
+3,4,596,738,train
+3,4,739,941,train
+3,5,0,132,train
+3,5,133,301,train
+3,5,302,437,train
+3,5,438,519,train
+3,5,520,679,train
+3,6,0,110,train
+3,6,111,226,train
+3,6,227,470,train
+3,6,471,595,train
+3,6,596,761,train
+3,6,762,812,train
+3,6,813,1000,train
+3,6,1001,1159,train
+3,6,1160,1289,train
+3,6,1290,1413,train
+3,7,0,187,train
+3,7,188,309,train
+3,7,310,437,train
+3,7,438,624,train
+3,7,625,724,train
+3,7,725,976,train
+3,7,977,1082,train
+3,7,1083,1203,train
+3,8,0,247,train
+3,8,248,423,train
+3,8,424,473,train
+3,8,474,607,train
+3,8,608,738,train
+3,8,739,932,train
+3,9,0,80,train
+3,9,81,209,train
+3,9,210,328,train
+3,9,329,447,train
+3,9,448,524,train
+3,9,525,722,train
+3,9,723,832,train
+3,9,833,1151,train
+3,9,1152,1318,train
+3,10,0,154,train
+3,10,155,278,train
+3,10,279,487,train
+3,10,488,636,train
+3,10,637,777,train
+3,10,778,956,train
+3,10,957,1129,train
+3,11,0,201,train
+3,11,202,377,train
+3,11,378,476,train
+3,11,477,571,train
+3,12,0,98,train
+3,12,99,221,train
+3,12,222,300,train
+3,12,301,492,train
+3,12,493,610,train
+3,12,611,714,train
+3,12,715,801,train
+3,12,802,994,train
+3,13,0,166,train
+3,13,167,460,train
+3,13,461,635,train
+3,13,636,710,train
+3,13,711,946,train
+3,14,0,110,train
+3,14,111,228,train
+3,14,229,345,train
+3,14,346,446,train
+3,14,447,650,train
+3,14,651,766,train
+3,14,767,835,train
+4,0,0,192,train
+4,0,193,281,train
+4,0,282,404,train
+4,0,405,614,train
+4,0,615,813,train
+4,0,814,1257,train
+4,1,0,177,train
+4,1,178,494,train
+4,2,0,305,train
+4,2,306,360,train
+4,2,361,509,train
+4,2,510,867,train
+4,2,868,1048,train
+4,3,0,216,train
+4,3,217,259,train
+4,3,260,294,train
+4,3,295,746,train
+4,3,747,1372,train
+4,3,1373,1426,train
+4,3,1427,1676,train
+4,3,1677,1887,train
+4,4,0,374,train
+4,4,375,600,train
+4,4,601,816,train
+4,5,0,202,train
+4,5,203,362,train
+4,6,0,168,train
+4,6,169,309,train
+4,6,310,395,train
+4,6,396,474,train
+4,6,475,549,train
+4,6,550,645,train
+4,6,646,773,train
+4,6,774,909,train
+4,6,910,1260,train
+4,7,0,115,train
+4,7,116,297,train
+4,8,0,178,train
+4,8,179,374,train
+4,8,375,677,train
+4,8,678,885,train
+4,8,886,1117,train
+4,9,0,415,train
+4,9,416,451,train
+4,9,452,691,train
+4,9,692,853,train
+4,10,0,642,train
+4,10,643,834,train
+4,10,835,1002,train
+4,10,1003,1152,train
+5,0,0,81,test
+5,0,82,211,test
+5,0,212,352,test
+5,0,353,508,test
+5,0,509,589,test
+5,1,0,259,test
+5,1,260,469,test
+5,1,470,692,test
+5,1,693,929,test
+5,1,930,1130,test
+5,2,0,225,test
+5,2,226,304,test
+5,2,305,424,test
+5,2,425,616,test
+5,2,617,713,test
+5,2,714,925,test
+5,2,926,1216,test
+5,3,0,107,test
+5,3,108,357,test
+5,3,358,547,test
+5,3,548,706,test
+5,4,0,230,test
+5,4,231,327,test
+5,4,328,516,test
+5,4,517,693,test
+5,4,694,1214,test
+5,5,0,153,test
+5,5,154,304,test
+5,5,305,403,test
+5,5,404,543,test
+5,5,544,722,test
+5,5,723,830,test
+5,5,831,999,test
+6,0,0,810,validation
+6,1,0,105,validation
+6,1,106,141,validation
+6,1,142,452,validation
+6,1,453,672,validation
+6,2,0,125,validation
+6,2,126,203,validation
+6,2,204,263,validation
+6,2,264,335,validation
+6,2,336,638,validation
+6,2,639,849,validation
+6,3,0,53,validation
+6,3,54,108,validation
+6,3,109,244,validation
+6,3,245,405,validation
+6,3,406,555,validation
+6,3,556,636,validation
+6,3,637,753,validation
+6,3,754,915,validation
+6,3,916,955,validation
+7,0,0,156,train
+7,0,157,432,train
+7,0,433,536,train
+7,0,537,703,train
+7,1,0,209,train
+7,1,210,379,train
+7,1,380,537,train
+7,1,538,718,train
+7,1,719,923,train
+7,1,924,1084,train
+7,2,0,192,train
+7,2,193,383,train
+7,2,384,514,train
+7,3,0,161,train
+7,3,162,319,train
+7,3,320,494,train
+7,4,0,304,train
+7,5,0,249,train
+7,5,250,514,train
+7,5,515,636,train
+7,5,637,840,train
+7,5,841,1005,train
+7,5,1006,1134,train
+7,5,1135,1217,train
+7,6,0,424,train
+7,6,425,768,train
+7,6,769,872,train
+7,6,873,985,train
+8,0,0,92,train
+8,0,93,187,train
+8,0,188,266,train
+8,0,267,298,train
+8,1,0,374,train
+8,2,0,452,train
+8,2,453,816,train
+8,3,0,108,train
+8,3,109,373,train
+8,3,374,375,train
+8,3,375,646,train
+8,4,0,221,train
+8,4,222,272,train
+8,4,273,437,train
+8,4,438,547,train
+8,4,548,639,train
+8,4,640,899,train
+8,4,900,962,train
+8,5,0,86,train
+8,5,87,231,train
+8,5,232,400,train
+8,6,0,98,train
+8,6,99,100,train
+8,6,100,255,train
+8,6,256,433,train
+8,7,0,137,train
+8,7,138,286,train
+8,7,287,466,train
+8,8,0,153,train
+8,8,154,230,train
+8,8,231,232,train
+8,8,232,349,train
+9,0,0,194,train
+9,0,195,358,train
+9,0,359,671,train
+9,1,0,187,train
+9,1,188,291,train
+9,1,292,615,train
+9,2,0,314,train
+9,2,315,426,train
+9,3,0,269,train
+9,3,270,446,train
+9,4,0,123,train
+9,4,124,211,train
+9,4,212,351,train
+9,4,352,656,train
+10,0,0,138,test
+10,0,139,297,test
+10,0,298,372,test
+10,0,373,603,test
+10,1,0,151,test
+10,1,152,257,test
+10,1,258,292,test
+10,1,293,383,test
+10,1,384,466,test
+10,1,467,664,test
+10,2,0,353,test
+11,0,0,375,validation
+11,0,376,642,validation
+11,1,0,203,validation
+11,1,204,396,validation
+11,1,397,483,validation
+11,2,0,146,validation
+11,2,147,499,validation
+11,3,0,242,validation
+11,3,243,394,validation
+11,3,395,511,validation
+11,3,512,627,validation
+11,3,628,886,validation
+11,4,0,234,validation
+11,4,235,428,validation
+11,4,429,560,validation
+12,0,0,175,train
+12,0,176,216,train
+12,0,217,256,train
+12,0,257,369,train
+12,1,0,171,train
+12,1,172,312,train
+12,2,0,121,train
+12,2,122,400,train
+12,2,401,656,train
+12,2,657,820,train
+12,2,821,1082,train
+13,0,0,226,train
+13,0,227,371,train
+13,0,372,577,train
+13,0,578,690,train
+13,0,691,828,train
+13,0,829,1015,train
+13,1,0,479,train
+13,1,480,570,train
+13,2,0,796,train
+13,3,0,175,train
+13,3,176,443,train
+13,3,444,678,train
+13,3,679,872,train
+13,4,0,313,train
+13,4,314,576,train
+13,5,0,140,train
+13,5,141,268,train
+13,5,269,517,train
+13,5,518,674,train
+13,5,675,871,train
+13,5,872,943,train
+13,6,0,252,train
+13,6,253,451,train
+13,6,452,623,train
+13,6,624,836,train
+13,7,0,169,train
+13,7,170,338,train
+13,7,339,490,train
+14,0,0,234,train
+14,0,235,338,train
+14,0,339,481,train
+14,0,482,593,train
+14,1,0,166,train
+14,1,167,285,train
+14,1,286,449,train
+14,1,450,569,train
+14,1,570,708,train
+14,1,709,964,train
+14,2,0,298,train
+14,2,299,632,train
+14,3,0,83,train
+14,3,83,212,train
+14,3,213,302,train
+14,3,303,437,train
+14,3,438,510,train
+14,4,0,199,train
+14,4,200,281,train
+14,4,282,438,train
+14,4,439,667,train
+14,4,668,794,train
+14,4,795,1024,train
+14,5,0,96,train
+14,5,97,285,train
+14,5,286,443,train
+14,5,444,665,train
+14,5,666,939,train
+14,6,0,85,train
+14,6,86,264,train
+14,6,265,443,train
+14,6,444,636,train
+14,7,0,309,train
+14,7,310,421,train
+14,7,422,459,train
+14,7,460,595,train
+14,7,596,753,train
+14,7,754,856,train
+14,8,0,385,train
+14,8,386,529,train
+14,8,530,622,train
+15,0,0,120,test
+15,0,121,222,test
+15,0,223,409,test
+15,0,410,540,test
+15,1,0,325,test
+15,1,326,586,test
+15,2,0,47,test
+15,2,48,145,test
+15,2,146,345,test
+15,2,346,465,test
+15,3,0,259,test
+15,3,260,409,test
+15,3,410,569,test
+15,4,0,109,test
+15,4,110,293,test
+15,4,294,477,test
+16,0,0,201,validation
+16,0,202,504,validation
+16,1,0,124,validation
+16,1,125,464,validation
+16,1,465,562,validation
+16,1,563,633,validation
+16,2,0,165,validation
+16,2,166,293,validation
+16,2,294,537,validation
+16,3,0,126,validation
+16,3,127,321,validation
+16,3,322,507,validation
+16,3,508,690,validation
+16,3,691,768,validation
+16,4,0,88,validation
+16,4,89,440,validation
+16,4,441,538,validation
+16,4,539,580,validation
+16,4,581,737,validation
+16,5,0,119,validation
+16,5,120,400,validation
+16,6,0,392,validation
+16,6,393,657,validation
+16,7,0,84,validation
+16,7,85,286,validation
+16,7,287,552,validation
+16,7,553,684,validation
+16,7,685,1014,validation
+16,8,0,142,validation
+16,8,143,228,validation
+16,8,229,413,validation
+16,8,414,814,validation
+16,9,0,111,validation
+16,9,112,183,validation
+16,9,184,313,validation
+16,9,314,351,validation
+16,10,0,70,validation
+16,10,71,157,validation
+16,10,158,357,validation
+16,10,358,543,validation
+16,11,0,81,validation
+16,11,82,191,validation
+16,11,192,469,validation
+16,11,470,772,validation
+16,11,773,1067,validation
+16,11,1068,1138,validation
+16,12,0,96,validation
+16,12,97,287,validation
+16,13,0,68,validation
+16,13,69,207,validation
+16,13,208,413,validation
+16,13,414,514,validation
+16,13,515,634,validation
+16,13,635,715,validation
+16,13,716,801,validation
+16,13,802,1038,validation
+16,14,0,66,validation
+16,14,67,132,validation
+16,14,133,265,validation
+16,14,266,509,validation
+16,14,510,630,validation
+16,15,0,208,validation
+16,15,209,355,validation
+16,15,356,541,validation
+16,15,542,724,validation
+16,16,0,10,validation
+16,16,11,147,validation
+16,16,148,533,validation
+16,17,0,13,validation
+16,17,14,96,validation
+16,17,97,169,validation
+16,17,170,344,validation
+16,17,345,482,validation
+17,0,0,148,train
+17,0,149,542,train
+17,0,543,715,train
+17,1,0,369,train
+17,1,370,485,train
+17,1,486,561,train
+17,2,0,206,train
+17,2,207,255,train
+17,2,256,375,train
+17,2,376,537,train
+17,2,538,656,train
+17,3,0,31,train
+17,3,32,151,train
+17,3,152,306,train
+17,3,307,534,train
+17,3,535,748,train
+17,3,749,915,train
+17,3,916,1041,train
+17,4,0,27,train
+17,4,28,149,train
+17,4,150,218,train
+17,4,219,399,train
+17,4,400,605,train
+17,4,606,792,train
+17,4,793,1149,train
+17,5,0,92,train
+17,5,93,267,train
+17,5,268,367,train
+17,5,368,477,train
+17,5,478,534,train
+17,5,535,601,train
+18,0,0,149,train
+18,0,150,365,train
+18,0,366,522,train
+18,1,0,98,train
+18,1,99,283,train
+18,1,284,375,train
+18,1,376,512,train
+18,1,513,656,train
+18,1,657,795,train
+18,1,796,916,train
+18,1,917,1072,train
+18,2,0,149,train
+18,2,150,317,train
+18,2,318,446,train
+18,2,447,719,train
+18,2,720,839,train
+18,2,840,1093,train
+18,2,1094,1173,train
+18,2,1174,1326,train
+18,2,1327,1452,train
+18,3,0,273,train
+18,3,274,519,train
+18,3,520,744,train
+18,3,745,838,train
+18,4,0,113,train
+18,4,114,322,train
+18,4,323,513,train
+18,4,514,667,train
+18,4,668,817,train
+18,5,0,71,train
+18,5,72,225,train
+18,5,226,397,train
+18,5,398,514,train
+18,5,515,640,train
+18,6,0,154,train
+18,6,155,315,train
+18,6,316,461,train
+18,6,462,646,train
+18,7,0,133,train
+18,7,134,224,train
+18,7,225,360,train
+18,7,361,650,train
+18,8,0,250,train
+18,8,251,351,train
+18,8,352,503,train
+18,8,504,760,train
+18,8,761,959,train
+18,8,960,1127,train
+18,8,1128,1269,train
+18,9,0,225,train
+18,9,226,367,train
+18,9,368,602,train
+18,9,603,896,train
+18,10,0,196,train
+18,10,197,364,train
+18,10,365,600,train
+18,10,601,750,train
+18,10,751,900,train
+18,10,901,1148,train
+18,11,0,302,train
+18,11,303,388,train
+18,11,389,476,train
+18,11,477,637,train
+18,11,638,825,train
+18,11,826,1108,train
+18,11,1109,1257,train
+18,12,0,138,train
+18,12,139,292,train
+18,12,293,336,train
+18,12,337,610,train
+18,12,611,858,train
+18,12,859,990,train
+18,12,991,1150,train
+18,13,0,164,train
+18,13,165,222,train
+18,13,223,386,train
+18,13,387,591,train
+18,13,592,797,train
+18,13,798,907,train
+18,13,908,978,train
+18,14,0,226,train
+18,14,227,627,train
+18,14,628,762,train
+18,15,0,364,train
+18,16,0,335,train
+18,16,336,453,train
+18,16,454,550,train
+18,16,551,715,train
+18,17,0,203,train
+18,17,204,434,train
+18,17,435,567,train
+18,18,0,39,train
+18,18,40,283,train
+18,18,284,583,train
+18,19,0,539,train
+18,19,540,691,train
+19,0,0,136,train
+19,0,137,275,train
+19,0,276,383,train
+19,0,384,537,train
+19,0,538,703,train
+19,1,0,259,train
+19,1,260,375,train
+19,1,376,499,train
+19,1,500,766,train
+19,2,0,151,train
+19,2,152,1126,train
+19,3,0,475,train
+19,3,476,550,train
+19,4,0,617,train
+19,5,0,171,train
+19,5,172,350,train
+19,5,351,457,train
+19,5,458,513,train
+19,5,514,640,train
+19,6,0,194,train
+19,6,195,475,train
+20,0,0,185,test
+20,0,186,505,test
+20,0,506,671,test
+20,0,672,825,test
+20,0,826,1012,test
+20,1,0,134,test
+20,1,135,287,test
+20,1,288,405,test
+20,1,406,464,test
+20,2,0,271,test
+20,2,272,622,test
+20,3,0,222,test
+20,3,223,346,test
+20,3,347,471,test
+20,3,472,599,test
+20,3,600,744,test
+20,3,745,985,test
+20,3,986,1500,test
+20,4,0,163,test
+20,4,164,351,test
+20,4,352,466,test
+20,4,467,700,test
+20,4,701,912,test
+21,0,0,222,validation
+21,0,223,756,validation
+21,0,757,850,validation
+21,0,851,1003,validation
+21,1,0,127,validation
+21,1,128,205,validation
+21,1,206,309,validation
+21,1,310,357,validation
+21,1,358,419,validation
+21,1,420,454,validation
+21,1,455,532,validation
+21,1,533,596,validation
+21,1,597,776,validation
+21,2,0,441,validation
+22,0,0,216,train
+22,0,217,604,train
+22,0,605,1001,train
+22,0,1002,1310,train
+22,1,0,227,train
+22,1,228,468,train
+22,1,469,803,train
+22,2,0,170,train
+22,2,171,279,train
+22,2,280,466,train
+22,3,0,109,train
+22,3,110,282,train
+22,3,283,417,train
+22,3,418,624,train
+22,4,0,72,train
+22,4,73,250,train
+22,4,251,405,train
+22,4,406,525,train
+22,4,526,597,train
+22,5,0,162,train
+22,5,163,396,train
+22,5,397,628,train
+22,5,629,915,train
+22,5,916,1008,train
+22,6,0,167,train
+22,6,168,390,train
+22,6,391,483,train
+22,6,484,846,train
+22,7,0,262,train
+22,7,263,400,train
+22,7,401,583,train
+22,7,584,739,train
+22,7,739,740,train
+22,7,741,905,train
+22,7,906,984,train
+22,7,985,1056,train
+22,8,0,148,train
+22,8,149,271,train
+22,8,272,336,train
+22,8,337,338,train
+22,8,338,570,train
+22,8,571,660,train
+22,8,661,831,train
+22,8,832,902,train
+22,9,0,105,train
+22,9,106,252,train
+22,9,253,394,train
+22,9,395,491,train
+22,9,492,621,train
+22,10,0,232,train
+22,10,233,408,train
+22,10,409,555,train
+22,10,556,698,train
+23,0,0,83,train
+23,0,84,187,train
+23,0,188,418,train
+23,0,419,631,train
+23,0,632,955,train
+23,1,0,233,train
+23,1,234,469,train
+23,1,470,533,train
+23,1,534,719,train
+23,1,720,928,train
+23,1,929,1065,train
+23,1,1066,1263,train
+23,2,0,258,train
+23,2,259,459,train
+23,3,0,215,train
+23,3,216,393,train
+23,3,394,659,train
+23,3,660,758,train
+23,3,759,908,train
+23,3,909,1051,train
+23,4,0,235,train
+23,4,236,518,train
+23,5,0,172,train
+23,5,173,494,train
+23,5,495,746,train
+23,6,0,344,train
+24,0,0,274,train
+24,0,275,443,train
+24,1,0,163,train
+24,1,164,419,train
+24,2,0,229,train
+24,2,230,465,train
+24,2,466,684,train
+24,2,685,1021,train
+24,2,1022,1144,train
+24,3,0,178,train
+24,3,179,286,train
+24,3,287,529,train
+24,3,530,736,train
+24,4,0,189,train
+24,4,190,392,train
+24,5,0,9,train
+24,5,9,173,train
+24,5,174,335,train
+24,5,336,455,train
+24,5,456,644,train
+24,6,0,158,train
+24,6,159,303,train
+24,6,304,386,train
+24,6,386,408,train
+24,6,409,508,train
+24,6,509,572,train
+24,6,573,758,train
+24,7,0,173,train
+24,7,174,214,train
+24,7,215,304,train
+24,7,305,435,train
+24,7,436,516,train
+24,7,517,590,train
+24,7,591,734,train
+24,7,735,963,train
+24,7,964,1220,train
+24,7,1221,1365,train
+24,7,1366,1368,train
+24,7,1368,1503,train
+24,8,0,159,train
+24,8,160,372,train
+24,9,0,196,train
+24,9,197,417,train
+24,9,418,575,train
+24,9,576,671,train
+24,9,672,1043,train
+24,10,0,296,train
+24,10,297,520,train
+24,10,521,814,train
+24,10,815,1013,train
+24,10,1014,1212,train
+24,11,0,89,train
+24,11,90,283,train
+24,11,284,466,train
+24,12,0,197,train
+24,12,198,267,train
+24,12,268,522,train
+24,12,522,523,train
+25,0,0,159,test
+25,0,160,408,test
+25,0,409,578,test
+25,0,579,733,test
+25,1,0,274,test
+25,1,275,526,test
+25,1,527,725,test
+25,1,726,803,test
+25,2,0,361,test
+25,2,362,485,test
+25,2,486,632,test
+25,2,633,709,test
+25,3,0,392,test
+25,3,393,538,test
+25,3,539,586,test
+25,3,587,835,test
+26,0,0,192,validation
+26,0,193,371,validation
+26,0,372,521,validation
+26,0,522,682,validation
+27,0,0,226,train
+27,0,227,371,train
+27,0,372,577,train
+27,0,578,690,train
+27,0,691,828,train
+27,0,829,1015,train
+27,1,0,899,train
+27,1,900,1073,train
+27,2,0,140,train
+27,2,141,268,train
+27,2,269,517,train
+27,2,518,674,train
+27,2,675,871,train
+27,2,872,943,train
+28,0,0,88,train
+28,0,89,101,train
+28,0,102,210,train
+28,0,211,340,train
+28,0,341,719,train
+28,0,720,875,train
+28,1,0,161,train
+28,1,162,295,train
+28,1,296,609,train
+28,1,610,745,train
+29,0,0,379,train
+29,0,380,630,train
+29,1,0,246,train
+29,1,247,311,train
+29,1,311,556,train
+29,2,0,252,train
+29,2,253,579,train
+30,0,0,242,test
+30,0,243,414,test
+30,0,415,612,test
+30,0,613,742,test
+30,0,743,1001,test
+30,1,0,198,test
+30,1,198,233,test
+30,1,234,291,test
+30,1,292,320,test
+30,1,321,359,test
+30,1,360,456,test
+30,1,457,569,test
+30,1,570,611,test
+30,1,612,685,test
+30,1,686,725,test
+30,1,726,827,test
+30,1,828,1055,test
+30,1,1056,1208,test
+30,1,1209,1343,test
+30,1,1344,1538,test
+31,0,0,209,validation
+31,0,210,379,validation
+31,0,380,537,validation
+31,0,538,718,validation
+31,0,719,923,validation
+31,0,924,1084,validation
+31,1,0,257,validation
+31,1,258,563,validation
+31,2,0,57,validation
+31,2,57,58,validation
+31,2,59,483,validation
+31,2,484,835,validation
+31,2,836,938,validation
+31,2,939,1051,validation
+31,3,0,249,validation
+31,3,250,514,validation
+31,3,515,636,validation
+31,3,637,840,validation
+31,3,841,1005,validation
+31,3,1006,1134,validation
+31,3,1135,1217,validation
+31,3,1218,1410,validation
+31,3,1411,1601,validation
+31,3,1602,1732,validation
+32,0,0,227,train
+32,0,228,730,train
+32,0,731,909,train
+32,0,910,974,train
+33,0,0,223,train
+33,0,224,352,train
+33,0,353,935,train
+33,0,936,1266,train
+33,1,0,178,train
+33,1,179,285,train
+33,1,286,341,train
+33,1,342,468,train
+33,1,469,769,train
+33,1,770,845,train
+33,1,846,969,train
+33,1,970,1236,train
+34,0,0,30,train
+34,0,31,311,train
+34,0,312,514,train
+34,0,515,688,train
+34,1,0,259,train
+34,1,260,610,train
+34,1,611,828,train
+34,1,829,997,train
+34,1,998,1103,train
+34,2,0,112,train
+34,2,113,332,train
+34,2,333,456,train
+34,2,457,731,train
+34,2,732,882,train
+34,2,883,994,train
+34,2,995,1127,train
+34,2,1128,1346,train
+34,2,1347,1527,train
+34,2,1528,1617,train
+34,2,1618,1775,train
+35,0,0,98,test
+35,0,99,283,test
+35,0,284,375,test
+35,0,376,512,test
+35,0,513,656,test
+35,0,657,795,test
+35,0,796,916,test
+35,0,917,1072,test
+35,1,0,112,test
+35,1,113,452,test
+35,1,453,558,test
+35,1,559,894,test
+35,1,895,1013,test
+35,1,1014,1110,test
+35,1,1111,1275,test
+35,2,0,166,test
+35,2,166,612,test
+35,2,613,791,test
+35,3,0,112,test
+35,3,112,113,test
+35,3,114,322,test
+35,3,323,513,test
+35,3,514,667,test
+35,3,668,817,test
+35,3,818,961,test
+35,4,0,302,test
+35,4,303,388,test
+35,4,389,476,test
+35,4,477,637,test
+35,4,638,825,test
+35,4,826,1108,test
+35,4,1109,1257,test
+35,5,0,138,test
+35,5,139,196,test
+35,5,197,350,test
+35,5,351,668,test
+35,5,669,916,test
+35,5,917,1048,test
+35,5,1049,1208,test
+35,6,0,164,test
+35,6,165,222,test
+35,6,223,386,test
+35,6,387,591,test
+35,6,592,797,test
+35,6,798,907,test
+35,6,908,978,test
+35,7,0,230,test
+35,7,231,363,test
+35,7,364,420,test
+35,7,421,570,test
+35,7,571,727,test
+35,8,0,341,test
+35,8,342,429,test
+35,8,430,922,test
+35,9,0,84,test
+35,9,85,181,test
+35,9,182,733,test
+35,9,734,895,test
+35,9,896,1064,test
+35,9,1065,1193,test
+35,9,1194,1508,test
+35,9,1509,1607,test
+36,0,0,111,validation
+36,0,112,196,validation
+36,0,197,495,validation
+36,0,496,604,validation
+36,0,605,807,validation
+36,0,808,924,validation
+37,0,0,246,train
+37,0,247,432,train
+37,0,433,489,train
+38,0,0,150,train
+38,0,151,280,train
+38,0,281,414,train
+38,0,415,628,train