README.md

---
license: cc-by-4.0
language:
- en
pretty_name: torchlight threads
task_categories:
- question-answering
- text-generation
tags:
- pytorch
- code
- sft
- forum
- debugging
size_categories:
- 10K<n<100K
configs:
- config_name: sft
  default: true
  data_files:
  - split: train
    path: sft/train.jsonl
  - split: validation
    path: sft/validation.jsonl
- config_name: eval
  data_files:
  - split: test
    path: eval/test.jsonl
- config_name: threads
  data_files:
  - split: train
    path: threads/threads.jsonl
---

<p align="center">
  <img alt="torchlight" src="https://huggingface.co/datasets/samadpls/torchlight-threads/resolve/main/assets/hero.png" width="100%">
</p>

<p align="center"><a href="https://github.com/samadpls/torchlight"><img alt="GitHub" src="https://img.shields.io/badge/GitHub-samadpls%2Ftorchlight-181717?logo=github"></a> <a href="https://samadpls.com"><img alt="samadpls.com" src="https://img.shields.io/badge/by-samadpls.com-EE4C2C"></a> <img alt="License: CC BY 4.0" src="https://img.shields.io/badge/license-CC%20BY%204.0-EE4C2C"> <img alt="Language: English" src="https://img.shields.io/badge/language-English-555"></p>

# 🔥 torchlight threads

> PyTorch questions from the [PyTorch forums](https://discuss.pytorch.org), each paired with **the answer that actually resolved it**, whoever wrote it.

The threads are the 22,293 topics that [ptrblck](https://discuss.pytorch.org/u/ptrblck), the forum's most active answerer, took part in, plus 193 recent solved topics from other answerers for evaluation. Which answer counts comes from the forum's own signals, not from who wrote it: the accepted solution, or the post the asker said fixed their problem.

The dataset was built to fine-tune and evaluate **torchlight**, a small PyTorch helper trained to give a specific fix when it can and to ask for a minimal repro when it can't.

## At a glance

| | |
|---|---|
| Threads crawled and cleaned | **22,484** |
| Usable rows (`sft` + `eval`) | **7,354**, from 6,841 threads |
| Answer examples | 6,581 (5,669 gold, 912 silver) |
| Clarification examples | 513 |
| Evaluation questions | 260, all gold and all reviewed |
| Threads reviewed by an agent | 1,269 (408 kept, 449 combined, 412 dropped) |

## ⚡ Quick start

```python
from datasets import load_dataset

sft = load_dataset("samadpls/torchlight-threads", "sft")
evaluation = load_dataset("samadpls/torchlight-threads", "eval", split="test")
threads = load_dataset("samadpls/torchlight-threads", "threads", split="train")

print(sft["train"][0]["messages"])
```

## Configs

| Config | Split | Rows | What it holds |
|---|---|---|---|
| `sft` | train | 6,925 | Chat examples (`messages`) for supervised fine-tuning, threads created before 2024-07-01 |
| `sft` | validation | 169 | Same format, threads from 2024-07-01 to 2025-01-01 |
| `eval` | test | 260 | Questions with a reviewed reference answer, threads created on or after 2025-01-01 |
| `threads` | train | 22,484 | Every crawled thread, cleaned, with its resolution and the reason it was excluded, if it was |

The splits are by date, never random, so every evaluation question is newer than anything in training.

The `threads` config holds every thread, including the ones behind the evaluation questions; its only split is called `train` because Hugging Face needs a split name. Each row's real split is in its `split` column. If you train on `threads`, keep only rows where `split` is `train`, or you will train on the test answers.

## How answers were chosen

1. **Accepted solution.** If a post is marked as the solution, it is the answer. When the asker marked their own thank-you post, the answer is the post they were thanking.
2. **Asker confirmation.** Otherwise, if the asker later said the fix worked ("that solved it", "works now"), the answer is the post they were replying to.
3. **Review.** A small LLM agent following [`prompts/review.md`](https://github.com/samadpls/torchlight/blob/main/prompts/review.md) reviewed every evaluation answer, plus training answers that were only a question, under 60 characters, mostly a link, or pointing at other posts. It kept the post, dropped the thread, or combined several posts into one answer.
4. **Grounding check.** A combined answer is rejected unless every code block in it appears verbatim in its source posts.

**Tiers:** `gold` answers are accepted solutions; `silver` answers were confirmed by the asker. The `eval` config is gold only.

## What a row looks like

<details>
<summary><b>An <code>sft</code> answer example</b></summary>

```
user       How does torchvision.transforms.Normalize work
           Whenever I Normalize my input images with its mean and std, and when i plot the
           images to visualize i get this message along with plots of images which are
           distorted in a way. Clipping input data to the valid range ...
assistant  The easiest way would be to plot them before normalizing.
           However, if that's not possible, you could also undo the normalization:
           (code that reverses Normalize with the same mean and std)
```

Each row also carries `kind`, `tier`, `answer_origin`, `answer_post_numbers`, `answer_authors`, `thread_id`, `url` and `created_at`.
</details>

<details>
<summary><b>The two kinds of <code>sft</code> examples</b></summary>

- `answer`: the conversation up to the answer, ending with the answer as the final assistant turn. Earlier assistant turns are context only.
- `clarification`: the opening question and a first reply that asked for missing details, such as a minimal executable snippet or the full error. These teach the model to ask instead of guessing.
</details>

### System prompt

Rows contain only `user` and `assistant` turns, so you can train with any system prompt. torchlight trains with this one:

> You are a careful PyTorch helper. Give specific, correct guidance with code where it helps. If a question lacks the details needed to answer it, ask for them, for example a minimal executable code snippet or the full error message, instead of guessing.

```python
SYSTEM_PROMPT = "You are a careful PyTorch helper. ..."
sft = sft.map(
    lambda row: {"messages": [{"role": "system", "content": SYSTEM_PROMPT}] + row["messages"]}
)
```

## Cleaning and filters

Every thread in all three configs is cleaned:

- text quoted from earlier posts is removed;
- images become `[image]`;
- emails and strings that look like API keys are redacted;
- deleted and hidden posts are dropped;
- code blocks are kept exactly as written.

6,841 of the 22,484 threads became examples. The rest stay in `threads` with the reason in `excluded`:

- No accepted answer and no confirmation from the asker: 14,005 threads
- Created before PyTorch 0.4 (May 2018): 743 threads
- Question depends on an image: 173 threads
- Answer is an image: 8 threads
- Longer than 8 messages or 12,000 characters: 269 threads
- Answer uses removed APIs such as `volatile=True`: 20 threads
- Not in English: 1 thread
- Dropped by the review agent: 412 threads
- Combined answer whose code was not found in its sources: 2 threads
- Evaluation threads without an accepted answer: 10 threads

## Fields

<details>
<summary><b>Show all fields</b></summary>

| Config | Fields |
|---|---|
| `sft` | `messages`, `kind`, `tier`, `answer_origin` (`verbatim` or `composed`), `answer_post_numbers`, `answer_authors`, `thread_id`, `url`, `created_at` |
| `eval` | `question`, `reference_answer`, `needs_clarification`, `answered_by_ptrblck`, `tier`, `answer_origin`, `answer_post_numbers`, `answer_authors`, `thread_id`, `url`, `created_at` |
| `threads` | `thread_id`, `url`, `title`, `created_at`, `category_id`, `tags`, `asker`, `split`, `resolution`, `excluded`, `posts` |
</details>

## ⚠️ Limitations

- Answers are only as correct as the forum's resolution signals. An accepted or confirmed answer can still be imperfect or tied to an older PyTorch version.
- The data leans toward the questions ptrblck answered, so it is heavy on debugging, CUDA, training loops and shape errors.
- Threads whose question or answer depends on an image were excluded.

## License and attribution

Forum content is licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) by the [LF Projects Terms of Use](https://lfprojects.org/policies/terms-of-use/) that the PyTorch forums follow, and this dataset is released under the same license. Credit goes to the forum's authors; every row carries the URL of the thread it came from.

## Author

Built by **Abdul Samad** ([samadpls.com](https://samadpls.com) · [GitHub @samadpls](https://github.com/samadpls)) with [torchlight](https://github.com/samadpls/torchlight). Questions and corrections: abdulsamadsid1@gmail.com.

## Citation

```bibtex
@misc{samad2026torchlight,
  author = {Abdul Samad},
  title = {torchlight threads: resolved PyTorch forum questions and answers},
  year = {2026},
  url = {https://huggingface.co/datasets/samadpls/torchlight-threads}
}
```