Datasets:
input_ids listlengths 23 11.5k | labels listlengths 23 11.5k | prompt_ids listlengths 19 3.63k | completion_ids listlengths 1 8.19k | messages listlengths 2 2 | domain stringclasses 3
values | source_dataset stringclasses 3
values | source_config stringclasses 2
values | source_split stringclasses 3
values | source_shard int64 0 139 | source_id stringlengths 32 64 | prompt_sha256 stringlengths 64 64 | source_row_index int64 0 100k | unified_order_index int64 0 60.3k | finish_reason stringclasses 2
values | generation_seed int64 42 60.4k |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
[
151644,
872,
198,
9885,
279,
1372,
315,
25780,
17767,
308,
1124,
701,
1380,
17767,
220,
16,
1124,
273,
80,
308,
1124,
273,
80,
220,
16,
15,
15,
15,
1124,
701,
1741,
429,
17767,
308,
1124,
8,
374,
537,
264,
4727,
9334,
11,
23739,
1... | [
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100... | [
151644,
872,
198,
9885,
279,
1372,
315,
25780,
17767,
308,
1124,
701,
1380,
17767,
220,
16,
1124,
273,
80,
308,
1124,
273,
80,
220,
16,
15,
15,
15,
1124,
701,
1741,
429,
17767,
308,
1124,
8,
374,
537,
264,
4727,
9334,
11,
23739,
1... | [
1654,
525,
4588,
311,
1477,
279,
1372,
315,
25780,
400,
308,
400,
1741,
429,
400,
220,
16,
1124,
273,
80,
308,
1124,
273,
80,
220,
16,
15,
15,
15,
400,
323,
400,
308,
400,
374,
3070,
1921,
334,
264,
4727,
9334,
11,
23739,
11,
47... | [
{
"role": "user",
"content": "Find the number of integers \\( n \\), where \\( 1 \\leq n \\leq 1000 \\), such that \\( n \\) is not a perfect square, cube, or fourth power."
},
{
"role": "assistant",
"content": "We are asked to find the number of integers $ n $ such that $ 1 \\leq n \\leq 1000 $... | math | nvidia/OpenMathReasoning | default | cot | 35 | ed7556298e542d5e91a293e5fa1484f2ca18617d7b4ff55b86d3be6a8c7dd451 | ed7556298e542d5e91a293e5fa1484f2ca18617d7b4ff55b86d3be6a8c7dd451 | 37,542 | 0 | stop | 42 |
[
151644,
872,
198,
78653,
25,
17143,
11,
633,
752,
847,
5777,
11016,
13,
1084,
594,
31034,
323,
7292,
4120,
2219,
3862,
374,
264,
1293,
1140,
315,
308,
7571,
304,
4065,
315,
93429,
11,
323,
566,
374,
311,
11047,
279,
22823,
330,
71,
... | [
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100... | [
151644,
872,
198,
78653,
25,
17143,
11,
633,
752,
847,
5777,
11016,
13,
1084,
594,
31034,
323,
7292,
4120,
2219,
3862,
374,
264,
1293,
1140,
315,
308,
7571,
304,
4065,
315,
93429,
11,
323,
566,
374,
311,
11047,
279,
22823,
330,
71,
... | [
1654,
525,
2661,
264,
3491,
1380,
93429,
6801,
311,
3293,
3070,
327,
32739,
220,
17,
7571,
334,
504,
264,
1140,
315,
1565,
77,
63,
7571,
311,
3070,
2810,
11853,
806,
23009,
334,
382,
4854,
3743,
702,
510,
12,
1565,
69,
25998,
44622,
... | [
{
"role": "user",
"content": "Ted: Robin, get me my legal pad. It's Pros and Cons Time!\n\nThere is a long list of n girls in front of Barney, and he is to calculate the optimal \"happiness\" he can find by selecting exactly 2 girls. (Why 2? No one knows!)\n\nTed, as a fan of pros and cons, suggests to make... | code | nvidia/OpenCodeReasoning | split_0 | split_0 | 7 | dfc913dc88440a7fb3b43f045a2d80a0 | 18443e9f86b38bdd03013ef102a706d341c5cccc7f2e6ebece53a44138cc86db | 60,279 | 1 | stop | 43 |
[
151644,
872,
198,
9885,
279,
7772,
17767,
296,
1124,
8,
1741,
429,
369,
678,
17767,
264,
11,
293,
11,
272,
11,
294,
861,
220,
15,
1124,
8,
448,
17767,
264,
488,
293,
488,
272,
488,
294,
284,
220,
16,
1124,
701,
279,
31205,
17767,
... | [
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100,
-100... | [
151644,
872,
198,
9885,
279,
7772,
17767,
296,
1124,
8,
1741,
429,
369,
678,
17767,
264,
11,
293,
11,
272,
11,
294,
861,
220,
15,
1124,
8,
448,
17767,
264,
488,
293,
488,
272,
488,
294,
284,
220,
16,
1124,
701,
279,
31205,
17767,
... | [
1654,
525,
4588,
311,
1477,
279,
3070,
26897,
477,
334,
897,
315,
400,
296,
400,
1741,
429,
369,
3070,
541,
334,
6785,
1931,
5109,
400,
264,
11,
293,
11,
272,
11,
294,
861,
220,
15,
400,
448,
271,
14085,
198,
64,
488,
293,
488,
... | [
{
"role": "user",
"content": "Find the largest \\( m \\) such that for all \\( a, b, c, d > 0 \\) with \\( a + b + c + d = 1 \\), the inequality \\( m < (a^3 + 1)(b^3 + 1)(c^3 + 1)(d^3 + 1) \\) holds."
},
{
"role": "assistant",
"content": "We are asked to find the **largest** value of $ m $ such... | math | nvidia/OpenMathReasoning | default | cot | 26 | e9282b31f7c5c324b52001f4ba14b8ab07b55db4115e9343ddcf7f08497fa05b | e9282b31f7c5c324b52001f4ba14b8ab07b55db4115e9343ddcf7f08497fa05b | 68,544 | 2 | stop | 44 |
[151644,872,198,59085,17013,293,14770,1087,323,8094,6233,4119,26618,7627,3175,16312,304,264,220,16,2(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,59085,17013,293,14770,1087,323,8094,6233,4119,26618,7627,3175,16312,304,264,220,16,2(...TRUNCATED) | [1654,525,2661,1447,12,220,23,293,14770,1087,320,33,340,12,220,22,4119,320,44,340,12,10657,315,220,1(...TRUNCATED) | [{"role":"user","content":"Eight eligible bachelors and seven beautiful models randomly purchase sin(...TRUNCATED) | math | nvidia/OpenMathReasoning | default | cot | 7 | ecd8f552294807bef6f27ad1ad7f9e4453671800c63215105573a4c65ab555b3 | ecd8f552294807bef6f27ad1ad7f9e4453671800c63215105573a4c65ab555b3 | 70,051 | 3 | stop | 45 |
[151644,872,198,9885,279,897,315,279,2629,17767,1124,37018,90,16,15170,16,1124,50853,220,17,1124,508(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,9885,279,897,315,279,2629,17767,1124,37018,90,16,15170,16,1124,50853,220,17,1124,508(...TRUNCATED) | [1654,525,2661,279,2629,1447,14085,198,59,1242,15159,77,28,16,92,47822,17,23,92,1124,37018,90,16,151(...TRUNCATED) | [{"role":"user","content":"Find the value of the sum \\( \\frac{1}{1 \\cdot 2 \\cdot 3 \\cdot 4} + \(...TRUNCATED) | math | nvidia/OpenMathReasoning | default | cot | 6 | b33f560da5a9408d15629962048f97a14f852ae58612b38182b06d4647ec223e | b33f560da5a9408d15629962048f97a14f852ae58612b38182b06d4647ec223e | 15,790 | 4 | stop | 46 |
[151644,872,198,9885,678,5810,5109,17767,308,1124,8,1741,429,1124,11520,37018,91362,0,481,220,16,151(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,9885,678,5810,5109,17767,308,1124,8,1741,429,1124,11520,37018,91362,0,481,220,16,151(...TRUNCATED) | [1654,525,311,1477,678,3070,52880,5109,334,400,308,400,1741,429,271,14085,198,59,37018,91362,0,481,2(...TRUNCATED) | [{"role":"user","content":"Find all natural numbers \\( n \\) such that \\(\\frac{n! - 1}{2n + 7}\\)(...TRUNCATED) | math | nvidia/OpenMathReasoning | default | cot | 28 | 256f60b73c5207867931fe641adabc7bf06262a8510310cd61c7e68aceb04b21 | 256f60b73c5207867931fe641adabc7bf06262a8510310cd61c7e68aceb04b21 | 14,785 | 5 | stop | 47 |
[151644,872,198,50,3948,279,2701,3491,13,7405,2704,311,2182,279,4226,320,437,1172,4226,8,4766,1124,7(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,50,3948,279,2701,3491,13,7405,2704,311,2182,279,4226,320,437,1172,4226,8,4766,1124,7(...TRUNCATED) | [1249,8253,1246,279,56836,8464,1711,59,24912,1710,3,8,323,78671,8464,1711,59,24912,56014,15976,3,8,1(...TRUNCATED) | [{"role":"user","content":"Solve the following problem. Make sure to put the answer (and only answer(...TRUNCATED) | science | nvidia/OpenScienceReasoning-2 | default | train | 1 | 179b09dc0c8b78f6feb58ba2a4c1c9733a2837f5028c54ff311c1a277c94bdcd | 179b09dc0c8b78f6feb58ba2a4c1c9733a2837f5028c54ff311c1a277c94bdcd | 72,904 | 6 | stop | 48 |
[151644,872,198,9885,279,26313,979,400,1124,6863,316,90,17,15,15,22,15170,24,16,31716,374,17779,553,(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,9885,279,26313,979,400,1124,6863,316,90,17,15,15,22,15170,24,16,31716,374,17779,553,(...TRUNCATED) | [1654,525,4588,311,1477,279,3070,77381,979,334,400,1124,6863,316,90,17,15,15,22,15170,24,16,92,400,3(...TRUNCATED) | [{"role":"user","content":"Find the remainder when $ \\binom{2007}{91}$ is divided by 91."},{"role":(...TRUNCATED) | math | nvidia/OpenMathReasoning | default | cot | 22 | ec1f372d619404a8537e6d0d65a2da2fcc24cf386ada79ff537315a93219c3d0 | ec1f372d619404a8537e6d0d65a2da2fcc24cf386ada79ff537315a93219c3d0 | 31,383 | 7 | stop | 49 |
[151644,872,198,14658,3337,7876,702,42310,2326,6785,25780,264,11,293,323,272,13,1260,13598,1493,5109(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,14658,3337,7876,702,42310,2326,6785,25780,264,11,293,323,272,13,1260,13598,1493,5109(...TRUNCATED) | [1654,525,2661,3040,5109,25,400,856,62,16,11,856,62,17,11,856,62,18,11,856,62,19,400,11,892,525,279,(...TRUNCATED) | [{"role":"user","content":"Polycarp has guessed three positive integers a, b and c. He keeps these n(...TRUNCATED) | code | nvidia/OpenCodeReasoning | split_0 | split_0 | 1 | 40bfc63b2d3b3dbb709347dbd3763ff7 | 09205002611fe7d28f82d92b3286363a0be557914bdb9431e913083d05004d04 | 48,566 | 8 | stop | 50 |
[151644,872,198,50,3948,279,2701,3491,13,7405,2704,311,2182,279,4226,320,437,1172,4226,8,4766,1124,7(...TRUNCATED) | [-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100,-100(...TRUNCATED) | [151644,872,198,50,3948,279,2701,3491,13,7405,2704,311,2182,279,4226,320,437,1172,4226,8,4766,1124,7(...TRUNCATED) | [785,3405,16555,264,220,20,20,4666,6284,883,448,650,58607,19754,31544,448,29783,11,15386,817,482,278(...TRUNCATED) | [{"role":"user","content":"Solve the following problem. Make sure to put the answer (and only answer(...TRUNCATED) | science | nvidia/OpenScienceReasoning-2 | default | train | 2 | 0ee92033b42658273ab534fb2def8eee67a6db653decb898e1644c0ec42c3175 | 0ee92033b42658273ab534fb2def8eee67a6db653decb898e1644c0ec42c3175 | 9,472 | 9 | stop | 51 |
Qwen3-4B Teacher Rollouts 76K Non-Thinking
This dataset contains 76,800 fixed teacher trajectories generated for a prompt-aligned reproduction study of on-policy distillation with Qwen3-1.7B. It is an independent research artifact, not an official release from the model or paper authors.
Models and generation
- Teacher:
Qwen/Qwen3-4B-Instruct-2507 - Tokenizer/chat template:
Qwen/Qwen3-1.7B - Mode: non-thinking (
enable_thinking=False) - Temperature: 0.7
- Top-p: 1.0
- Top-k: 0
- Maximum prompt length: 4,096 tokens
- Maximum completion length: 8,192 tokens
- Seed for row
i:42 + i
Prompts were passed to the vLLM completions API as exact token IDs with no additional special tokens. Returned token IDs and special tokens are retained.
Statistics
| Statistic | Value |
|---|---|
| Rows | 76,800 |
| Prompt tokens | 21,596,089 |
| Completion tokens | 264,338,320 |
| Total stored tokens | 285,934,409 |
| Mean completion length | 3,441.91 |
| Completions reaching 8,192 tokens | 10,974 (14.29%) |
The first 25,600 trajectories exactly reuse the earlier 100-step teacher bank. The remaining 51,200 rows extend the same seed-42 prompt permutation without an epoch wrap.
Fields
input_ids: concatenated prompt and completion token IDslabels:-100over prompt tokens and completion token IDs thereafterprompt_ids,completion_ids: separate exact token sequencesmessages: original conversation plus decoded teacher assistant responsedomain: math, science, or codesource_dataset,source_config,source_split,source_shard,source_id: upstream provenanceprompt_sha256: hash of the source promptsource_row_index,unified_order_index: source and rollout schedule indicesfinish_reason,generation_seed: generation metadata
Integrity
- Prompt order SHA-256:
c1cc096fa94f82bf44457673b112039f604aaba19711a850f368514da6d7429d - Completion IDs SHA-256:
a3c5ae8551b23025bb57d6d60a011ef7e156d034d0853656dbb33d29b064127b
The repository includes the full generation manifest. A post-generation audit found zero prompt-identity mismatches, and the prompt-token hash matches a fresh application of the documented Qwen non-thinking chat template.
Usage
from datasets import load_dataset
dataset = load_dataset(
"YangyiH/qwen3-4b-teacher-rollouts-76k-nonthinking",
split="train",
)
Source data and licensing
Prompts originate from the NVIDIA OpenMathReasoning, OpenScienceReasoning-2, and OpenCodeReasoning datasets. This repository does not assert a new blanket license over upstream prompt content. Users should review and comply with the source datasets' licenses, terms, and attribution requirements, as well as the Qwen model license.
Limitations
- The source mixture intentionally retains duplicate prompts inherited from the upstream datasets.
- No additional quality, difficulty, benchmark-contamination, or safety filtering was applied.
- 14.29% of completions reached the configured 8,192-token cap and may be truncated rather than naturally terminated.
- Generated responses can contain errors or undesirable content inherited from the model and source prompts.
- Downloads last month
- 106