Update config
Browse files
README.md
CHANGED
@@ -39,7 +39,7 @@ This dataset repo contains only the queries and relevance judgements. The corpus
|
|
39 |
```
|
40 |
ciral_dataset = load_dataset("ciral/ciral", "hausa") #or swahili, somali, yoruba
|
41 |
|
42 |
-
for data in
|
43 |
query_id = data['query_id']
|
44 |
query = data['query']
|
45 |
pos_qrels = data['positive_passages']
|
|
|
39 |
```
|
40 |
ciral_dataset = load_dataset("ciral/ciral", "hausa") #or swahili, somali, yoruba
|
41 |
|
42 |
+
for data in ciral_dataset['dev']: # or 'testA' or 'testB'
|
43 |
query_id = data['query_id']
|
44 |
query = data['query']
|
45 |
pos_qrels = data['positive_passages']
|
ciral.py
CHANGED
@@ -108,6 +108,15 @@ class CIRAL(datasets.GeneratorBasedBuilder):
|
|
108 |
'text': datasets.Value("string"),
|
109 |
}],
|
110 |
"negative_passages": [{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
111 |
"docid": datasets.Value("string"),
|
112 |
"text": datasets.Value("string")}]
|
113 |
}
|
@@ -144,7 +153,7 @@ class CIRAL(datasets.GeneratorBasedBuilder):
|
|
144 |
datasets.SplitGenerator(
|
145 |
name='testB',
|
146 |
gen_kwargs={
|
147 |
-
'filepaths': downloaded_files['
|
148 |
},
|
149 |
),
|
150 |
]
|
@@ -152,7 +161,7 @@ class CIRAL(datasets.GeneratorBasedBuilder):
|
|
152 |
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
153 |
def _generate_examples(self, filepaths):
|
154 |
lang = self.config.name
|
155 |
-
corpus = datasets.load_dataset('ciral/ciral-corpus', lang)
|
156 |
docid2doc = {doc['docid']: doc['text'] for doc in corpus}
|
157 |
|
158 |
query_file, qrel_file, pools_file = (filepaths) if len(filepaths) == 3 else (filepaths[0], filepaths[1], None)
|
|
|
108 |
'text': datasets.Value("string"),
|
109 |
}],
|
110 |
"negative_passages": [{
|
111 |
+
"docid": datasets.Value("string"),
|
112 |
+
"text": datasets.Value("string")
|
113 |
+
}],
|
114 |
+
|
115 |
+
"pools_positive_passages": [{
|
116 |
+
'docid': datasets.Value("string"),
|
117 |
+
'text': datasets.Value("string"),
|
118 |
+
}],
|
119 |
+
"pools_negative_passages": [{
|
120 |
"docid": datasets.Value("string"),
|
121 |
"text": datasets.Value("string")}]
|
122 |
}
|
|
|
153 |
datasets.SplitGenerator(
|
154 |
name='testB',
|
155 |
gen_kwargs={
|
156 |
+
'filepaths': downloaded_files['testB'],
|
157 |
},
|
158 |
),
|
159 |
]
|
|
|
161 |
# method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
|
162 |
def _generate_examples(self, filepaths):
|
163 |
lang = self.config.name
|
164 |
+
corpus = datasets.load_dataset('ciral/ciral-corpus', lang)['train']
|
165 |
docid2doc = {doc['docid']: doc['text'] for doc in corpus}
|
166 |
|
167 |
query_file, qrel_file, pools_file = (filepaths) if len(filepaths) == 3 else (filepaths[0], filepaths[1], None)
|