Keep K3 suite selection and report-schema scoring in bash, merge K3/vision dataset_args into dpv4 yamls, and pin EvalScope at 735d920ee911 with local patches. Co-authored-by: Cursor <cursoragent@cursor.com>
4.3 KiB
4.3 KiB
BoolQ-Indic
Overview
BoolQ-Indic is a translation of the BoolQ yes/no reading-comprehension benchmark into 10 Indic languages plus English, for evaluating multilingual passage understanding.
Task Description
- Task Type: Multilingual Yes/No Reading Comprehension
- Input: Passage + yes/no question in one of 11 languages
- Output:
YesorNo - Languages: Bengali, English, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu
Evaluation Notes
- Default configuration uses 0-shot evaluation (validation split)
- Use
subset_listto evaluate specific languages (e.g.,['hi', 'ta']), orlimitto cap sample count — the full default run is 35,970 samples across all 11 languages - Set
few_shot_num> 0 to enable few-shot prompting; examples are drawn from thetrainsplit - All languages ship in a single dataset config; this adapter reformats by the
languagefield
Properties
| Property | Value |
|---|---|
| Benchmark Name | indic_boolq |
| Dataset ID | sarvamai/boolq-indic |
| Paper | N/A |
| Tags | MCQ, MultiLingual, ReadingComprehension |
| Metrics | accuracy |
| Default Shots | 0-shot |
| Evaluation Split | validation |
| Train Split | train |
Data Statistics
| Metric | Value |
|---|---|
| Total Samples | 35,970 |
| Prompt Length (Mean) | 822.66 chars |
| Prompt Length (Min/Max) | 275 / 5035 chars |
Per-Subset Statistics:
| Subset | Samples | Prompt Mean | Prompt Min | Prompt Max |
|---|---|---|---|---|
bn |
3,270 | 801.95 | 294 | 2308 |
en |
3,270 | 814.26 | 292 | 5035 |
gu |
3,270 | 793.37 | 283 | 2105 |
hi |
3,270 | 818.47 | 297 | 3078 |
kn |
3,270 | 833.99 | 275 | 2920 |
ml |
3,270 | 869.99 | 294 | 3558 |
mr |
3,270 | 806.68 | 289 | 2593 |
or |
3,270 | 787.68 | 306 | 1482 |
pa |
3,270 | 804.93 | 295 | 1975 |
ta |
3,270 | 904.01 | 297 | 3570 |
te |
3,270 | 813.95 | 284 | 3312 |
Sample Example
Subset: bn
{
"input": [
{
"id": "0f04f0f7",
"content": "Answer the following multiple choice question. The entire content of your response should be of the following format: 'ANSWER: [LETTER]' (without quotes) where [LETTER] is one of A,B.\n\nসকল জৈববস্তুই কমপক্ষে এই ধাপগুলোর মধ্য দিয়ে যায়: এগুলো চা ... [TRUNCATED 1099 chars] ... বার্কলেতে ছয়টি পৃথক গবেষণা বিশ্লেষণ করার পর, একটি গবেষণায় উপসংহারে আসা গেছে যে, ভুট্টা থেকে ইথানল উৎপাদনে পেট্রোলিয়ামের ব্যবহার গ্যাসোলিন উৎপাদনের তুলনায় অনেক কম।\n\nQuestion: ইথানল উৎপাদনের চেয়ে তৈরিতে কি বেশি শক্তি লাগে??\n\nA) Yes\nB) No"
}
],
"choices": [
"Yes",
"No"
],
"target": "B",
"id": 0,
"group_id": 0,
"subset_key": "bn",
"metadata": {
"language": "Bengali"
}
}
Note: Some content was truncated for display.
Prompt Template
Prompt Template:
Answer the following multiple choice question. The entire content of your response should be of the following format: 'ANSWER: [LETTER]' (without quotes) where [LETTER] is one of {letters}.
{question}
{choices}
Usage
Using CLI
evalscope eval \
--model YOUR_MODEL \
--api-url OPENAI_API_COMPAT_URL \
--api-key EMPTY_TOKEN \
--datasets indic_boolq \
--limit 10 # Remove this line for formal evaluation
Using Python
from evalscope import run_task
from evalscope.config import TaskConfig
task_cfg = TaskConfig(
model='YOUR_MODEL',
api_url='OPENAI_API_COMPAT_URL',
api_key='EMPTY_TOKEN',
datasets=['indic_boolq'],
dataset_args={
'indic_boolq': {
# subset_list: ['bn', 'en', 'gu'] # optional, evaluate specific subsets
}
},
limit=10, # Remove this line for formal evaluation
)
run_task(task_cfg=task_cfg)