Whittle distillation set: 25,178 Qwen3.8-27B traces with top-20 logprobs

README.md

@@ -0,0 +1,141 @@
+---
+pretty_name: Whittle distillation set (Qwen3.8-27B reasoning traces + top-20 logprobs)
+license: other
+language:
+- en
+task_categories:
+- text-generation
+tags:
+- knowledge-distillation
+- teacher-logprobs
+- reasoning
+- chain-of-thought
+- tool-use
+- agentic
+- qwen
+size_categories:
+- 10K<n<100K
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*.parquet
+---
+
+# Whittle distillation set: Qwen3.8-27B traces with top-20 logprobs
+
+About **25,000 complete, quality-filtered answers** from **Qwen3.8-27B** (thinking on), each with the teacher's **top-20 log-probabilities at every answer token**. It is built for **logit-level knowledge distillation**: a student can match the teacher's full per-token distribution, not just its sampled text. We use it to train [Whittle-Qwen-3.8-35B-A3B](https://huggingface.co/logic65/Whittle-Qwen-3.8-35B-A3B).
+
+- **25,178 rows**, **78.0M tokens**, of which **25.4M are answer tokens with top-20 logprobs** (about 507M token/logprob pairs).
+- Covers maths, grade-school maths, code review and explanation, open chat, strict JSON extraction, long-document reading, and **agentic tool use**. Agent episodes were **executed in a sandbox and kept only when hidden unit tests passed**.
+- Every trace is **complete**: it stopped by itself, closed its thinking block and ended on `<|im_end|>`. It is **correct wherever it could be graded**.
+
+```python
+from datasets import load_dataset
+ds = load_dataset("logic65/whittle-distill-qwen3.8-27b", split="train")
+row = ds[0]
+row["kind"], row["prompt"][:200], row["reasoning"][:200], row["answer"][:200]
+# logit-level targets: for answer position j (predicting input_ids[answer_start + j])
+#   candidates = row["topk_ids"][j]; logprobs = row["topk_logprobs"][j]   (20 each, highest first; renormalise before use)
+```
+
+---
+
+## Columns
+
+| Column | Type | Meaning |
+|---|---|---|
+| `id` | string | stable id (`kind-#####`; agent episode turns get `/turnN`) |
+| `kind` | string | `math`, `gsm`, `code`, `chat`, `json`, `longctx`, `agent1` (single tool step), `agentx` (turn of an executed agent episode) |
+| `source` | string | where the prompt came from (see *Sources*) |
+| `prompt` | string | the full prompt as rendered with the Whittle / Qwen chat template (system, tools, earlier turns), ending with `<|im_start|>assistant\n<think>\n` |
+| `response` | string | the teacher's full raw answer: reasoning, `</think>`, visible answer or tool calls, `<|im_end|>` |
+| `reasoning` | string | the thinking part only |
+| `answer` | string | the visible answer only (empty when the turn is only a tool call) |
+| `tool_calls` | string (JSON) | parsed tool calls `[{"name": ..., "arguments": {...}}]`, if any |
+| `input_ids` | list[int32] | exact token ids of `prompt + response` |
+| `answer_start` | int32 | index in `input_ids` where the response begins |
+| `topk_ids` | list[list[int32]] | `[n_answer_tokens][20]`: teacher's top-20 candidate token ids at each answer position, highest first |
+| `topk_logprobs` | list[list[float32]] | `[n_answer_tokens][20]`: their **raw** log-probabilities (before temperature / top-p) |
+| `n_prompt_tokens`, `n_answer_tokens` | int32 | lengths |
+
+`topk_*[j]` is the teacher's distribution **predicting `input_ids[answer_start + j]`**. The token actually sampled is always among the 20; we checked every row. Tokenizer: [logic65/Whittle-Qwen-3.8-35B-A3B](https://huggingface.co/logic65/Whittle-Qwen-3.8-35B-A3B) (the Qwen3.8 tokenizer, 248,320 tokens, `<|im_end|>` = 248046).
+
+---
+
+## Composition
+
+| Kind | Rows | Median answer tokens | Source | Content |
+|---|---:|---:|---|---|
+| `math` | 4,013 | 671 | MATH train, levels 2–5 | step-by-step solutions, graded `\boxed{}` answer |
+| `gsm` | 1,443 | 264 | GSM8K train | word problems, graded final number |
+| `code` | 4,487 | 2,160 | open-source files | review, explain, write tests, edge cases, refactor |
+| `chat` | 1,334 | 1,283 | oasst2 | open-ended questions (mostly English) |
+| `json` | 2,000 | 144 | synthetic | strict schema extraction, exact typed fields |
+| `longctx` | 1,470 | 197 | open-source files | find one function in a 24k–139k-character file (prompts up to ~42k tokens) |
+| `agent1` | 2,409 | 145 | synthetic | one next step in a Claude-Code-style tool session |
+| `agentx` | 8,022 turns (1,495 episodes) | 81 per turn | synthetic, executed | multi-turn coding episodes with real Read / Write / Edit / Bash results |
+
+The GSM8K **test** split is not included, nor are any held-out evaluation problems.
+
+---
+
+## How it was made
+
+- **Teacher:** `Qwen/Qwen3.8-27B-FP8` (official block-128 FP8 build) on vLLM 0.30.
+  - Checked against the bf16 teacher on 69,922 answer positions: **98.1% top-1 agreement**, total variation 0.016 over the top-20, 98.2% shared probability mass.
+- **Sampling:** thinking on, temperature 0.6, top-p 0.95, top-k 20.
+  - Reasoning effort was the teacher template's maximum (*xhigh*) for all kinds except code. Code used the default effort with focused instructions, because open-ended code requests at *xhigh* ran past 16k tokens without finishing.
+  - Maximum answer length: 16,384 tokens (8,192 for grade-school maths, JSON and agent turns).
+- **Context distillation.** The teacher answered with its *xhigh* "reason carefully" instruction. The `prompt` column is rendered **without** it, so a student trained on these rows learns careful reasoning by default.
+- **Filtering:** 19,500 prompts went in and **18,651 (96%) were kept**. A trace is kept only if **all** of these hold:
+  1. it finished with its own end-of-turn token;
+  2. the think block is closed;
+  3. no tool markup leaked into the answer;
+  4. it is not degenerate (minimum lengths);
+  5. every sampled token appears in its own top-20;
+  6. the answer is correct:
+     - maths: normalised `\boxed{}` comparison;
+     - GSM: final number;
+     - JSON: exact fields with strict types;
+     - long-context: exact file and parameters;
+     - single agent steps: the right action (e.g. fix the implementation, never the test);
+     - agent episodes: hidden unit tests pass **and** the original tests still pass.
+- **Agent episodes:** the teacher's Read / Write / Edit / Bash calls ran against a real sandbox, with Claude-Code-style tool results and a command allowlist, for up to 16 rounds. Every assistant turn of a passing episode is its own row. In half of the rows, earlier turns keep their reasoning; in the other half it is stripped, as many clients do.
+- **Privacy scan:** every row was scanned for credentials, tokens and private network addresses before release. No rows needed removal.
+
+---
+
+## Sources and licences
+
+| Source | Kinds | Licence |
+|---|---|---|
+| MATH (Hendrycks et al.), train split | `math` | MIT |
+| GSM8K, train split | `gsm` | MIT |
+| OpenAssistant oasst2 (first turns) | `chat` | Apache-2.0 |
+| Permissively licensed open-source code (licence-checked) | `code`, `longctx` | each file's own permissive licence |
+| Synthetic (generated by our scripts) | `json`, `agent1`, `agentx` | released with this dataset |
+| Qwen3.8-27B outputs | all answers and logprobs | Apache-2.0 (Qwen model licence) |
+
+The sources are mixed, so the dataset is marked `license: other`. Check each source's terms for your use.
+
+---
+
+## Limitations
+
+- **FP8 teacher:** 98.1% top-1 agreement with bf16, not 100%.
+- **Top-20 only:** the long tail of the distribution is not stored.
+- **Grader strictness:** a few correct answers were filtered out, for example `(e)`-style letter answers. Maths or chat answers that hit the 16k cap were dropped, so the very longest reasoning is slightly under-represented.
+- **Synthetic agent tasks** are small single-file Python projects. They teach tool discipline, not large-codebase navigation.
+- **Prompt rendering:** prompts use the Qwen chat template with thinking on. If your student uses a different template, re-render from `reasoning` / `answer` / `tool_calls`. The logprobs belong to the teacher's tokenisation of `response`.
+
+## Citation
+
+```
+@misc{whittle-distill-2026,
+  title  = {Whittle distillation set: Qwen3.8-27B reasoning traces with top-20 logprobs},
+  author = {logic65},
+  year   = {2026},
+  url    = {https://huggingface.co/datasets/logic65/whittle-distill-qwen3.8-27b}
+}
+```

data/train-00000-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a80f238663e781b532198af07dc14d4b54e803756e2d4766f5d5288b9f5ef72e
+size 30819296

data/train-00001-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:018a99dff8233f680c3fb6a4526f810f1091e5323109e0adf141fd959be4993b
+size 25771899

data/train-00002-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b0ced05c1a271f8795639373a5bf77f911030f331a07f62d157993ca65ba1bb
+size 35684954

data/train-00003-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f0ef8396b91b8bbbde811c55d1c8333a28996ae6fe6415269283f2f48e7483d
+size 35081593

data/train-00004-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4ed28be3d34c6c2a27cc3b870362f6e1bd92e44264d30fca92e8737ff46bec6
+size 29493936

data/train-00005-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6eb9b27b7cf0d932e59f3405d04662d0fdaa790e5623479d8a7aeb3c65a21100
+size 29671738

data/train-00006-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f82c43c68ee6b7ece307396ca600f89ca07c0f1ae29cdc2a8c9129f0df2c8aa
+size 33042257

data/train-00007-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acae6ac4e61b32544c74567b62ecaa54ff1e8296125e958ca1b70c2127b777f6
+size 36093187

data/train-00008-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11b901d2ea0afd9ba96fdb6e7e3002322717c310d23df2f5106f8a26796a655e
+size 39311316

data/train-00009-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbe679c11eccc9222d8c53574da563a9778877a424b199cb74efa91767189515
+size 30813240

data/train-00010-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f5751a3b4a4e5428f8d5f1745a32f0d8667d1c60a285109e1dcecc9c56a6997
+size 31757909

data/train-00011-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0459486b53816c226eae154f924b61586e8ec0b87ffb79dea1c70645ab0724bc
+size 33393897

data/train-00012-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e4ec34ab2b627f26b513f53926e8daabf53f049eba328525c3665a0f0683547
+size 31990564

data/train-00013-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e438e977fe9dfdca88973b9a66db1869bd810264839fa49a662a0d71c1f3484f
+size 38831724

data/train-00014-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d619e4ee729578a575716c5ce2ad13d9b2c57b87c1b97d01b721f999c2ffe6f0
+size 43915040

data/train-00015-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14e4a5c3f9b5d7e912c1a5cd7a1c4c6562e41e068d28a7e20e0cde59549e27a5
+size 43351982

data/train-00016-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa1e369d2b86c6c5a27facc232cca164447581a4f4a9305b018eae76062b1eef
+size 36162541

data/train-00017-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04484ed91fcc28fe488b51b1c69888e16cd6e65e86a73e68f7338dc6d25f5edc
+size 42044677

data/train-00018-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5df2cd259daa2047bd7b1bdb396dc29be0a80a52cbbc6c3c1682844981a09d74
+size 39119374

data/train-00019-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8d673958a67555520cbe8d2d7b250258395a2573487671682293d86228d4582
+size 39676333

data/train-00020-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33b54b445cce0318939a5d15b69b5f7a8ef3874cb979e27f9c7b03a877690b58
+size 31187686

data/train-00021-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed2e3564b94b884dc31bca66902786488d73f5dbe86ae827ae3dce3c0ec4b9e8
+size 42603237

data/train-00022-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68596520f29e31937ba2772a5522cdab16163c4d9b0d4855a7384267cb97abb9
+size 31367881

data/train-00023-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6eea272702c04e6f386a000a1ac5050399f022c11e623f9ec7242e8da64cf29b
+size 39543900

data/train-00024-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef6982ff60daf9aaf653856d7fe93bd8914ca5790b9903246952c984ab742ff6
+size 34153699

data/train-00025-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e0f617f0119ae9511f2c79bf4550fb6a952a5dd97e90144b3f327df72362dcd
+size 36684473

data/train-00026-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ed86a14304dfd7af59dd0e0f4db3d47c6b8015b84602a2907589508a139d4a0
+size 36046279

data/train-00027-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51b4e9654a9c63813faef08a28473f1bd90632f8b0dcb8edaea1457707c12d32
+size 45834697

data/train-00028-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b06b260ab10c67a98dc52de635e17476a39eaf6661879db69092323c0e67f70
+size 40871488

data/train-00029-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0380aafd3c399d2eda21012889a23756e3f0b3fc2b60e9a1b9632c2aa72eff4c
+size 34954988

data/train-00030-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42663aaf41904e50638ef0dfa0d34aef647b5cb5328b1994685a5387f1432bb7
+size 42059909

data/train-00031-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ff06294ce7d2288737af153fdc1fc792fb5206e00db72687d09b9f000f6ab0e
+size 37240044

data/train-00032-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3c9841f3d0040f2a553780eb6699a1f329daaf1a9b26c3a74edea1c06ae001c
+size 34878394

data/train-00033-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42f0c59a279bae428d5f550457bffc4d08b65507936578bc96c33421016b5d6a
+size 36160289

data/train-00034-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6e78600eea58ab97913d443bbf1c74f4248efd5ee8c1419ab934c9eb1ee7977
+size 43585484

data/train-00035-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc867c38def5adb52c9c21035b84c53ab644b7aaff2d5257638865cfc6886136
+size 40731484

data/train-00036-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab90ea1638bfdd7c4dab142a09db89989ed08ebff8f013d15e8edddd5384598d
+size 34802970

data/train-00037-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5eb9694913030223970e9e1626ff246acb9f7d84633109ec527c26ba9634ca5
+size 41401325

data/train-00038-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e272340c17acc575f2a5f6cde47eb8ace09a385ffeb36734d6ea2b23ae85d6b6
+size 38256478

data/train-00039-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8662c0551ee7d97cabd58dde42cbe958bda52179f4b3261ff9c524e6a9de04cd
+size 35537573

data/train-00040-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2cf8cfe1ffb1bd3420af2b0f6a10fef2c5e0fd1db2ef977b5670f618dcb0c15
+size 41234651

data/train-00041-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64f7a8c2189cd71f8b357252475aa56f2d23b91939548902d3d796a0d60cb0b5
+size 32352830

data/train-00042-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13f410ee7b3d79f290433e09b4be1caf9f455bec35f84366c812474e7c168213
+size 42187093

data/train-00043-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a28a409fc41fe19dd69679614598af43b86402bb8c79a2cc1f84faec2133dbe
+size 36223690

data/train-00044-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afd21dc37b9df7954ec3d2080ccdec0880845f6bb9c53f42bb917ad8b5b317c4
+size 37324032

data/train-00045-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85c3090171d68e5ec93073495c28c8374de6d0b840fc45fbd0f8cfdc18b2df96
+size 31235755

data/train-00046-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e01ecf579f3739859897c0318c003954ca925c3ce8734b88b9565188d0397137
+size 36952192

data/train-00047-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aef429aed8332bbb95734fdf85099d4dc642ebfff79e7ad457f67ca37124c253
+size 39171164

data/train-00048-of-00051.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42ce3ca291d529722b160a53dc491ec8a20c2d04fb2a4a6c60677894609aa95b
+size 34213394