jjkim
commited on
Commit
·
747a38b
1
Parent(s):
96c951b
fix metric info
Browse files- code_eval.py +7 -10
code_eval.py
CHANGED
@@ -145,8 +145,11 @@ class CodeEval(evaluate.Metric):
|
|
145 |
# This defines the format of each prediction and reference
|
146 |
features=datasets.Features(
|
147 |
{
|
148 |
-
"
|
149 |
-
"
|
|
|
|
|
|
|
150 |
}
|
151 |
),
|
152 |
homepage="https://github.com/openai/human-eval",
|
@@ -157,11 +160,10 @@ class CodeEval(evaluate.Metric):
|
|
157 |
|
158 |
def _compute(
|
159 |
self,
|
|
|
160 |
predictions,
|
161 |
-
pred_key,
|
162 |
pred_template,
|
163 |
references,
|
164 |
-
ref_key,
|
165 |
ref_template,
|
166 |
k=[1, 10, 100],
|
167 |
num_workers=4,
|
@@ -179,14 +181,9 @@ class CodeEval(evaluate.Metric):
|
|
179 |
"This metric is currently not supported on Windows."
|
180 |
)
|
181 |
|
182 |
-
predictions = sorted(predictions, key=lambda x: x["id"])
|
183 |
-
references = sorted(references, key=lambda x: x["id"])
|
184 |
with ThreadPoolExecutor(max_workers=num_workers) as executor:
|
185 |
results = {}
|
186 |
-
for pred_d, ref_d in zip(predictions, references):
|
187 |
-
assert pred_d["id"] == ref_d["id"]
|
188 |
-
tid = pred_d["id"]
|
189 |
-
|
190 |
results[tid] = []
|
191 |
pred = pred_d[pred_key]
|
192 |
ref = ref_d[ref_key]
|
|
|
145 |
# This defines the format of each prediction and reference
|
146 |
features=datasets.Features(
|
147 |
{
|
148 |
+
"ids": datasets.Value("string"),
|
149 |
+
"predictions": datasets.Sequence(datasets.Value("string")),
|
150 |
+
"pred_template": datasets.Value("string"),
|
151 |
+
"references": datasets.Sequence(datasets.Value("string")),
|
152 |
+
"ref_template": datasets.Value("string"),
|
153 |
}
|
154 |
),
|
155 |
homepage="https://github.com/openai/human-eval",
|
|
|
160 |
|
161 |
def _compute(
|
162 |
self,
|
163 |
+
ids,
|
164 |
predictions,
|
|
|
165 |
pred_template,
|
166 |
references,
|
|
|
167 |
ref_template,
|
168 |
k=[1, 10, 100],
|
169 |
num_workers=4,
|
|
|
181 |
"This metric is currently not supported on Windows."
|
182 |
)
|
183 |
|
|
|
|
|
184 |
with ThreadPoolExecutor(max_workers=num_workers) as executor:
|
185 |
results = {}
|
186 |
+
for tid, pred_d, ref_d in zip(ids, predictions, references):
|
|
|
|
|
|
|
187 |
results[tid] = []
|
188 |
pred = pred_d[pred_key]
|
189 |
ref = ref_d[ref_key]
|