Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
0: string
1: string
2: string
3: string
4: string
5: string
6: string
7: string
8: string
9: string
10: string
11: string
12: string
13: string
14: string
15: string
16: string
17: string
18: string
19: string
20: string
21: string
22: string
23: string
24: string
25: string
26: string
27: string
28: string
29: string
30: string
31: string
32: string
33: string
34: string
35: string
36: string
37: string
38: string
39: string
40: string
41: string
42: string
43: string
44: string
45: string
46: string
47: string
48: string
49: string
50: string
51: string
52: string
53: string
54: string
55: string
56: string
57: string
58: string
59: string
60: string
61: string
62: string
63: string
64: string
65: string
66: string
67: string
68: string
69: string
70: string
71: string
72: string
73: string
74: string
75: string
76: string
77: string
78: string
79: string
80: string
81: string
82: string
83: string
84: string
85: string
86: string
87: string
88: string
89: string
90: string
91: string
92: string
93: string
94: string
95: string
96: string
97: string
98: string
99: string
100: string
101: string
102: string
103: string
104: string
105: string
106: string
107: string
108: string
109: string
110: string
111: string
112: string
113: string
114: string
115: string
116: string
117: string
118: string
119: string
120: string
121: string
122: string
123: string
124: string
125: string
126: string
127: string
128: string
129: string
130: string
131: string
132: string
133: string
13
...
76: string
477: string
478: string
479: string
480: string
481: string
482: string
483: string
484: string
485: string
486: string
487: string
488: string
489: string
490: string
491: string
492: string
493: string
494: string
495: string
496: string
497: string
498: string
499: string
questions: int64
generation: struct<base_model: string, do_sample: bool, max_input_tokens: int64, max_new_tokens: int64>
child 0, base_model: string
child 1, do_sample: bool
child 2, max_input_tokens: int64
child 3, max_new_tokens: int64
prompt: struct<source: string, source_commit: string, sha256: string, transport: string>
child 0, source: string
child 1, source_commit: string
child 2, sha256: string
child 3, transport: string
runs: list<item: struct<name: string, adapter: string, responses: int64, correct: int64, ex: double, by_di (... 127 chars omitted)
child 0, item: struct<name: string, adapter: string, responses: int64, correct: int64, ex: double, by_difficulty: s (... 115 chars omitted)
child 0, name: string
child 1, adapter: string
child 2, responses: int64
child 3, correct: int64
child 4, ex: double
child 5, by_difficulty: struct<simple: double, moderate: double, challenging: double>
child 0, simple: double
child 1, moderate: double
child 2, challenging: double
child 6, remote_output: string
child 7, adapter_config_sha256: string
comparison_caveat: string
benchmark: string
evaluator: string
to
{'benchmark': Value('string'), 'evaluator': Value('string'), 'questions': Value('int64'), 'prompt': {'source': Value('string'), 'source_commit': Value('string'), 'sha256': Value('string'), 'transport': Value('string')}, 'generation': {'base_model': Value('string'), 'do_sample': Value('bool'), 'max_input_tokens': Value('int64'), 'max_new_tokens': Value('int64')}, 'runs': List({'name': Value('string'), 'adapter': Value('string'), 'responses': Value('int64'), 'correct': Value('int64'), 'ex': Value('float64'), 'by_difficulty': {'simple': Value('float64'), 'moderate': Value('float64'), 'challenging': Value('float64')}, 'remote_output': Value('string'), 'adapter_config_sha256': Value('string')}), 'comparison_caveat': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 489, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2818, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
0: string
1: string
2: string
3: string
4: string
5: string
6: string
7: string
8: string
9: string
10: string
11: string
12: string
13: string
14: string
15: string
16: string
17: string
18: string
19: string
20: string
21: string
22: string
23: string
24: string
25: string
26: string
27: string
28: string
29: string
30: string
31: string
32: string
33: string
34: string
35: string
36: string
37: string
38: string
39: string
40: string
41: string
42: string
43: string
44: string
45: string
46: string
47: string
48: string
49: string
50: string
51: string
52: string
53: string
54: string
55: string
56: string
57: string
58: string
59: string
60: string
61: string
62: string
63: string
64: string
65: string
66: string
67: string
68: string
69: string
70: string
71: string
72: string
73: string
74: string
75: string
76: string
77: string
78: string
79: string
80: string
81: string
82: string
83: string
84: string
85: string
86: string
87: string
88: string
89: string
90: string
91: string
92: string
93: string
94: string
95: string
96: string
97: string
98: string
99: string
100: string
101: string
102: string
103: string
104: string
105: string
106: string
107: string
108: string
109: string
110: string
111: string
112: string
113: string
114: string
115: string
116: string
117: string
118: string
119: string
120: string
121: string
122: string
123: string
124: string
125: string
126: string
127: string
128: string
129: string
130: string
131: string
132: string
133: string
13
...
76: string
477: string
478: string
479: string
480: string
481: string
482: string
483: string
484: string
485: string
486: string
487: string
488: string
489: string
490: string
491: string
492: string
493: string
494: string
495: string
496: string
497: string
498: string
499: string
questions: int64
generation: struct<base_model: string, do_sample: bool, max_input_tokens: int64, max_new_tokens: int64>
child 0, base_model: string
child 1, do_sample: bool
child 2, max_input_tokens: int64
child 3, max_new_tokens: int64
prompt: struct<source: string, source_commit: string, sha256: string, transport: string>
child 0, source: string
child 1, source_commit: string
child 2, sha256: string
child 3, transport: string
runs: list<item: struct<name: string, adapter: string, responses: int64, correct: int64, ex: double, by_di (... 127 chars omitted)
child 0, item: struct<name: string, adapter: string, responses: int64, correct: int64, ex: double, by_difficulty: s (... 115 chars omitted)
child 0, name: string
child 1, adapter: string
child 2, responses: int64
child 3, correct: int64
child 4, ex: double
child 5, by_difficulty: struct<simple: double, moderate: double, challenging: double>
child 0, simple: double
child 1, moderate: double
child 2, challenging: double
child 6, remote_output: string
child 7, adapter_config_sha256: string
comparison_caveat: string
benchmark: string
evaluator: string
to
{'benchmark': Value('string'), 'evaluator': Value('string'), 'questions': Value('int64'), 'prompt': {'source': Value('string'), 'source_commit': Value('string'), 'sha256': Value('string'), 'transport': Value('string')}, 'generation': {'base_model': Value('string'), 'do_sample': Value('bool'), 'max_input_tokens': Value('int64'), 'max_new_tokens': Value('int64')}, 'runs': List({'name': Value('string'), 'adapter': Value('string'), 'responses': Value('int64'), 'correct': Value('int64'), 'ex': Value('float64'), 'by_difficulty': {'simple': Value('float64'), 'moderate': Value('float64'), 'challenging': Value('float64')}, 'remote_output': Value('string'), 'adapter_config_sha256': Value('string')}), 'comparison_caveat': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Qwen3-8B Text-to-SQL 评测结果
本仓库发布 Qwen3-8B 真实推理输出、作者 EX 评测输出和不可变清单。
已核验结论
在 BIRD Mini-Dev SQLite 的 500 题、作者集合型 EX 标准下：
| 模型 | 推理 prompt | EX | 正确数 |
|---|---|---|---|
| Qwen3-8B 基座 | 本项目构造的结构化 prompt | 41.80% | 209 / 500 |
| Qwen3-8B + LoRA checkpoint-969 | 本项目构造的结构化 prompt | 50.40% | 252 / 500 |
正式 checkpoint-969 运行
已发布的 experiment_summaries/author_prompt_qwen3_8b_minidev_ex.json 记录了完整 500 题的作者 prompt 对照实验：Qwen3-8B 基座为 49.40%（247/500），checkpoint-969 为 49.20%（246/500）。该文件只含评测汇总和运行元数据，不含预测 SQL、数据库或 gold SQL。
runs/qwen3_8b_dbdev4_checkpoint-969/ 包含取得 50.40% 的真实运行产物：
predictions.jsonl：自定义结构化 prompt 的 498 条question_id/db_id/sql预测。author_predictions.json：按 BIRD 作者格式适配后的 500 题预测；缺失的 119、120 两题填入空 SQL。author_ex.txt：作者evaluation_ex.py的 SQLite 集合型 EX 输出，50.40%（252/500）。SHA256SUMS：上述三个文件的 SHA-256。
作者的 prompt 文件在此运行中仅用于题目顺序和输出格式适配，不参与模型推理。模型推理使用的是本项目构造的结构化 prompt。BIRD SQLite 数据库、gold SQL 和每题的 gold 执行结果不在本仓库发布。
来源：BIRD Mini-Dev https://github.com/bird-bench/mini_dev，派生发布遵守 CC BY-SA 4.0。
- Downloads last month
- 29