Spaces:
Runtime error
Runtime error
Update data_utils.py
Browse files- data_utils.py +7 -0
data_utils.py
CHANGED
|
@@ -120,6 +120,7 @@ def sample_an_eval_result(eval_results, model_list=[]):
|
|
| 120 |
random.shuffle(eval_results)
|
| 121 |
|
| 122 |
for eval_item in eval_results:
|
|
|
|
| 123 |
model = eval_item['model']
|
| 124 |
task_type = eval_item['task_type'] # primary task type
|
| 125 |
if model not in model_list:
|
|
@@ -135,6 +136,9 @@ def sample_an_eval_result(eval_results, model_list=[]):
|
|
| 135 |
"task_type": task_type,
|
| 136 |
"plan_history": plan_history,
|
| 137 |
"ground_history": ground_history,
|
|
|
|
|
|
|
|
|
|
| 138 |
"image": eval_item['image'].replace("eval/aokvqa/images/val2017/", "file/data_dir/test_images/")
|
| 139 |
}
|
| 140 |
else:
|
|
@@ -144,6 +148,9 @@ def sample_an_eval_result(eval_results, model_list=[]):
|
|
| 144 |
"task_type": task_type,
|
| 145 |
"plan_history": plan_history,
|
| 146 |
"ground_history": ground_history,
|
|
|
|
|
|
|
|
|
|
| 147 |
"image": None
|
| 148 |
}
|
| 149 |
break
|
|
|
|
| 120 |
random.shuffle(eval_results)
|
| 121 |
|
| 122 |
for eval_item in eval_results:
|
| 123 |
+
print(eval_item.keys())
|
| 124 |
model = eval_item['model']
|
| 125 |
task_type = eval_item['task_type'] # primary task type
|
| 126 |
if model not in model_list:
|
|
|
|
| 136 |
"task_type": task_type,
|
| 137 |
"plan_history": plan_history,
|
| 138 |
"ground_history": ground_history,
|
| 139 |
+
"pred": eval_item['pred'],
|
| 140 |
+
"answer": eval_item['answer'],
|
| 141 |
+
"correctness": eval_item['correctness'],
|
| 142 |
"image": eval_item['image'].replace("eval/aokvqa/images/val2017/", "file/data_dir/test_images/")
|
| 143 |
}
|
| 144 |
else:
|
|
|
|
| 148 |
"task_type": task_type,
|
| 149 |
"plan_history": plan_history,
|
| 150 |
"ground_history": ground_history,
|
| 151 |
+
"pred": eval_item['pred'],
|
| 152 |
+
"answer": eval_item['answer'],
|
| 153 |
+
"correctness": eval_item['correctness'],
|
| 154 |
"image": None
|
| 155 |
}
|
| 156 |
break
|