Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: CastError
Message: Couldn't cast
conversations: list<item: struct<from: string, value: string>>
child 0, item: struct<from: string, value: string>
child 0, from: string
child 1, value: string
source: string
to
{'conversations': List({'from': Value('string'), 'value': Value('string')})}
because column names don't match
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1827, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
for item in generator(*args, **kwargs):
~~~~~~~~~^^^^^^^^^^^^^^^^^
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 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
conversations: list<item: struct<from: string, value: string>>
child 0, item: struct<from: string, value: string>
child 0, from: string
child 1, value: string
source: string
to
{'conversations': List({'from': Value('string'), 'value': Value('string')})}
because column names don't match
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1879, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
conversations list |
|---|
[
{
"from": "human",
"value": "Who made you?"
},
{
"from": "gpt",
"value": "I was created by Duckets McQuackin. You can find him on X at @Franzferdinan57."
}
] |
[
{
"from": "human",
"value": "You are tasked with implementing a non-maximum suppression algorithm for object detection. The given code snippet is part of a larger function that performs non-maximum suppression on a set of bounding box detections. The variables `ndets`, `suppressed_t`, `suppressed`, `order`,... |
[
{
"from": "human",
"value": "How should I pick the right compute model for a backend API that needs to handle 50k requests per minute at peak, has WebSocket connections, and connects to Postgres?"
},
{
"from": "gpt",
"value": "Start with the constraints, not the technology. The shape of the work... |
[
{
"from": "human",
"value": "My background worker process runs continuously and its memory footprint doubles every few days until the OS kills it. What systematic approach should I take to identify and fix this memory leak?"
},
{
"from": "gpt",
"value": "**Phase 1 — Isolate and reproduce**\n\nBe... |
[
{
"from": "human",
"value": "How would I design a URL shortener like bit.ly?"
},
{
"from": "gpt",
"value": "**Core requirements:**\n- Short URL → long URL redirect (302 or 301)\n- Custom aliases (optional)\n- Analytics (click counts, referrers, geo)\n- 100:1 read:write ratio (massively read-heav... |
[
{
"from": "human",
"value": "Your team needs to fine-tune a language model but has limited compute resources. What technical factors should guide your choice between full fine-tuning, LoRA, and QLoRA?"
},
{
"from": "gpt",
"value": "**Key factor 1: Available VRAM determines your starting point**\... |
[
{
"from": "human",
"value": "I'm building a new HTTP service and I want it to stay maintainable as it grows. What architectural and design decisions matter most from the start?"
},
{
"from": "gpt",
"value": "Treat the API as a long-lived contract, not a script. The decisions that cost the least ... |
[
{
"from": "human",
"value": "My newly built image starts up and then dies within a second, showing no output. What are the steps to diagnose what's killing it?"
},
{
"from": "gpt",
"value": "**1. Get the last output the container produced**\n```bash\ndocker logs <container_id>\ndocker logs --det... |
[
{
"from": "human",
"value": "Users keep reporting that some page loads fail with a server error, but I can't reproduce it on demand. What's my debugging workflow?"
},
{
"from": "gpt",
"value": "**Start with the logs — find the real exception**\nA 500 is a catch-all. The exception text is in your... |
[
{
"from": "human",
"value": "Refactor this Python that has both class-based and function-based approaches mixed:\n```python\n# Module-level\ncache = {}\n\ndef get_user(user_id):\n if user_id in cache:\n return cache[user_id]\n user = db.query('SELECT * FROM users WHERE id = %s', [user_id])\n ... |
[
{
"from": "human",
"value": "You are tasked with implementing a simple bank account system using object-oriented programming principles in Python. Your system should consist of two classes: `Account` and `Customer`. The `Account` class should support depositing, withdrawing, and checking the balance. The `C... |
[
{
"from": "human",
"value": "How do I decide what to cache in a web service, and how do I avoid the classic mistakes?"
},
{
"from": "gpt",
"value": "**Decision framework: ask these four questions**\n\n1. *Is the read path hotter than the write path?* If the same value is fetched many times for e... |
[
{
"from": "human",
"value": "You are tasked with designing a `Book` class to manage a library's inventory. The class should have the following attributes: `title`, `author`, `isbn`, `price`, and `quantity`. Implement the following methods:\n\n- `get_info`: Returns a formatted string containing the book's de... |
[
{
"from": "human",
"value": "Refactor this to use composition over inheritance:\n```python\nclass JsonHandler:\n def __init__(self, logger):\n self.logger = logger\n def log(self, msg):\n self.logger.log(msg)\n\nclass TimedHandler:\n def __init__(self, logger):\n self.logger = ... |
[
{
"from": "human",
"value": "Refactor this long function:\n```python\ndef send_welcome_email(user):\n # 200 lines of: validate user, look up template, render template, send email, log result, retry on failure, update DB, etc.\n pass\n```"
},
{
"from": "gpt",
"value": "Extract each responsi... |
[
{
"from": "human",
"value": "Our dashboard loads in 45 seconds when users run reports. We haven't changed the query in months. What do I check first?"
},
{
"from": "gpt",
"value": "Start by confirming where time is actually being spent — the database, the network, or the application layer.\n\n**... |
[
{
"from": "human",
"value": "What are the critical differences between serverless, containers, and VMs that make one clearly wrong for certain workloads, and what does each approach cost at scale?"
},
{
"from": "gpt",
"value": "Serverless fails when you need sustained connections or predictable ... |
[
{
"from": "human",
"value": "How would I design a feature store for ML?"
},
{
"from": "gpt",
"value": "**What a feature store does:**\n- Centralized repository of features (input variables) for ML models\n- Ensures the same feature computation is used at training and inference (no train/serve sk... |
[
{
"from": "human",
"value": "What's the best way to structure endpoints in a long-lived HTTP service so it doesn't rot after two years?"
},
{
"from": "gpt",
"value": "**URL design — treat resources as nouns, not actions:**\n- Good: `POST /trips/{trip_id}/legs`, `DELETE /trips/{trip_id}/legs/{leg... |
[
{
"from": "human",
"value": "My Python program is taking way too long to finish. What's a systematic way to actually speed it up instead of just guessing?"
},
{
"from": "gpt",
"value": "**Step 1: Stop guessing, start measuring**\n\nYou can't optimize what you haven't measured. Two practical prof... |
[
{
"from": "human",
"value": "I have thousands of raw text documents I want to use for fine-tuning a language model. What's the end-to-end workflow to convert this raw material into training-ready data?"
},
{
"from": "gpt",
"value": "Converting raw text into training-ready data follows a seven-st... |
[
{
"from": "human",
"value": "Walk me through the architecture for building a TinyURL-style service from scratch. I want to understand every moving part — from how the short code is generated, to how the redirect actually happens in production, to how you'd handle billions of links. What are the key design d... |
[
{
"from": "human",
"value": "My long-running worker process RSS keeps climbing over days until the OOM killer terminates it. What's a methodical way to isolate what's holding the memory?"
},
{
"from": "gpt",
"value": "**Step 1 — Distinguish a real leak from baseline pressure**\n- Monitor RSS ove... |
[
{
"from": "human",
"value": "I'm building a fraud detection system and every time we deploy a new model, the old one starts performing worse in production even though metrics look fine offline. How do I make sure my model uses the same feature values during training and when it serves real-time predictions?... |
[
{
"from": "human",
"value": "How would you design a distributed rate limiter for a multi-tenant SaaS API?"
},
{
"from": "gpt",
"value": "**Functional requirements:**\n- Per-tenant quotas (Free: 100 req/min, Pro: 1000 req/min, Enterprise: custom)\n- Per-endpoint granularity (write operations cost... |
[
{
"from": "human",
"value": "Test Desktop Control (Lobster) MCP integration:\n\n1. Run via mcporter to check the MCP tools: `mcporter list desktop-control` (if available)\n2. If mcporter not available, directly test the Node.js MCP server by checking: `node -e \"const t=require('C:/Users/<user>/.openclaw/wo... |
[
{
"from": "human",
"value": "Walk me through how to build a real-time messaging backend that can handle tens of millions of users with low latency."
},
{
"from": "gpt",
"value": "**Functional requirements:**\n- Sub-second message delivery end-to-end\n- Persistent, paginated, searchable message h... |
[
{
"from": "human",
"value": "What systematic process can I follow to diagnose intermittent HTTP 500 errors in my production web application?"
},
{
"from": "gpt",
"value": "**Phase 1 — Capture the failure evidence**\nA 500 status code hides the real problem. Your first job is finding the exceptio... |
[
{
"from": "human",
"value": "Write tests for this Python decorator:\n```python\nimport functools\nimport time\n\ndef retry(max_attempts=3, delay=0.1):\n def decorator(fn):\n @functools.wraps(fn)\n def wrapper(*args, **kwargs):\n for attempt in range(max_attempts):\n ... |
[
{
"from": "human",
"value": "What are the essential architectural components and their interactions when building a messaging system that handles millions of concurrent users sending and receiving messages instantly?"
},
{
"from": "gpt",
"value": "A scalable real-time messaging platform requires... |
[
{
"from": "human",
"value": "My JavaScript code says \"Cannot read property 'foo' of undefined\" and I can't figure out where."
},
{
"from": "gpt",
"value": "**Step-by-step debugging approach:**\n\n**1. Get the full stack trace:**\n- Modern browsers show the stack trace in the console\n- Look fo... |
[
{
"from": "human",
"value": "Our API endpoint is slow — averaging 800ms. My teammate wants to add Redis caching. How do I decide if that's the right call?"
},
{
"from": "gpt",
"value": "Run the numbers before touching anything. Caching is a fix with a cost: you get speed in exchange for cache in... |
[
{
"from": "human",
"value": "My Node.js app has an uncaught exception that crashes the process. How do I find what's throwing it?"
},
{
"from": "gpt",
"value": "**Step-by-step debugging approach:**\n\n**1. Get the full stack trace:**\n- Node.js prints the stack trace to stderr when an uncaught e... |
[
{
"from": "human",
"value": "Write an integration test for this API endpoint:\n```js\napp.get('/api/users/:id', async (req, res) => {\n const user = await db.users.findById(req.params.id);\n if (!user) return res.status(404).json({ error: 'not found' });\n res.json(user);\n});\n```"
},
{
"f... |
[
{
"from": "human",
"value": "Implement a `Library` class in Python that supports the following operations:\n\n1. **Add a Book**: Add a book to the library with a given title and author. If the book already exists, print a message indicating that the book is already in the library.\n2. **Remove a Book**: Rem... |
[
{
"from": "human",
"value": "I have a query that takes 30 seconds. What's your debugging playbook?"
},
{
"from": "gpt",
"value": "**1. Capture the execution plan with real numbers**\n\n```sql\nEXPLAIN (ANALYZE, BUFFERS, FORMAT TEXT)\nSELECT o.id, o.total, u.email\nFROM orders o\nJOIN users u ON ... |
[
{
"from": "human",
"value": "My nginx returns 502 Bad Gateway. How do I debug?"
},
{
"from": "gpt",
"value": "**Step-by-step debugging approach:**\n\n**1. Understand what 502 means:**\nNginx couldn't reach the upstream server (your app). The app might be:\n- Not running at all\n- Running but cra... |
[
{
"from": "human",
"value": "Refactor this class that's grown to do too much (God Object):\n```python\nclass User:\n def __init__(self, name, email):\n self.name = name\n self.email = email\n self.db = Database()\n self.email_service = EmailService()\n self.logger = Log... |
[
{
"from": "human",
"value": "My async Python code has a race condition. How do I find it?"
},
{
"from": "gpt",
"value": "**Step-by-step debugging approach:**\n\n**1. Confirm it's a race condition:**\n- Does the bug appear sometimes, not always?\n- Does running with `-X dev` (single-threaded asyn... |
[
{
"from": "human",
"value": "I'm about to fine-tune a base model on a domain-specific corpus (~50K documents) for the first time. Walk me through the data preparation end-to-end — what I should actually do, in what order, and what mistakes to avoid."
},
{
"from": "gpt",
"value": "Treat prep as a... |
[
{
"from": "human",
"value": "My web app is fast at low traffic but latency explodes when lots of users hit it at once. Where should I look?"
},
{
"from": "gpt",
"value": "**Translation: your system has a hidden ceiling, and you've hit it.** It works up to N concurrent users, then degrades. This ... |
[
{
"from": "human",
"value": "Our checkout API just started returning 500s intermittently. Nothing was deployed today. Walk me through how to figure this out."
},
{
"from": "gpt",
"value": "**Phase 1 — Define the blast radius (2 minutes)**\n- Is the 500 rate uniform or bursty? Steady ~0.1% errors... |
[
{
"from": "human",
"value": "My Python web service has been running fine for months, but recently the RSS has been creeping up a few MB per hour until OOM kills it. What's a systematic way to track down what's holding onto memory?"
},
{
"from": "gpt",
"value": "**Step 1 — Verify the leak pattern... |
[
{
"from": "human",
"value": "What are the key components and data flow for building a large-scale team chat platform (think 10M+ users with real-time messaging)?"
},
{
"from": "gpt",
"value": "**Scale targets to design for:**\n- Sub-second message delivery to online users (p95 < 500ms)\n- Persis... |
[
{
"from": "human",
"value": "Refactor this function with too many arguments:\n```python\ndef create_user(name, email, age, country, language, theme, notifications, beta, newsletter, last_login, created_at):\n # ... 30 lines\n```"
},
{
"from": "gpt",
"value": "11 positional/named arguments is ... |
[
{
"from": "human",
"value": "You are tasked with implementing a binary search tree (BST) in Python. Your implementation should include the following functionalities:\n\n1. **Insertion**: Insert a new node with a given key into the BST.\n2. **Search**: Determine if a node with a given key exists in the BST.\... |
End of preview.
DuckBot Training Data — public redacted edition
ShareGPT-format SFT datasets built locally for the DuckBot personal-agent fine-tunes (Unsloth, Mac Studio), published 2026-10-15. These are the exact training files used in the most recent runs, with personal information removed (see "Privacy checks" below). Originals remain private.
Creator: Duckets McQuackin — X: @Franzferdinan57
Files
| file | rows | what it is |
|---|---|---|
duckbot-train-curated-86k.jsonl |
85,788 | Flagship curated train set (coding-heavy, ~60% code) — Ornith 1.5 35B run |
duckbot-train-curated-76k.jsonl |
76,517 | Earlier curated set — Qwen 3.8 27B run (trained as two parts) |
code-focused.jsonl |
68,241 | Code-only split of the curated lineage (Oct 2026 experiment) |
decision-focused.jsonl |
17,323 | Decision/routing examples split (Oct 2026 experiment) |
general.jsonl |
56,982 | General (non-code) split |
duckbot-eval-training-run.jsonl |
262 | Eval set used alongside all of the above |
Row counts include the creator watermark rows described below.
Format: one JSON object per line, {"conversations": [{"from": "human"|"gpt", "value": "..."}]}.
Sources
Curated merges of public datasets (OpenHermes-2.5, glaive-function-calling-v2, SlimOrca, WebSight, s1K, Fable 5.1, Pasta009 instruction-fusion, OpenThoughts, Precinct6, medical and science sets) plus locally generated UI/UX and distillation examples. Component datasets keep their own upstream licenses; this compilation is shared for research/personal use.
Privacy checks (three passes, 2026-10-15)
Every file went through three independent scans before publication:
- Pass 1 — targeted scan + redaction. Searched for the owner's email,
phone numbers, home address, family names, Telegram ID/handle, agent
chat-log rows, tailnet hostnames/IPs, and API-token patterns. The owner's
real email, phone, address and zip appear zero times. Removed:
33 personal rows (agent chat logs, machine-memory rows, personal-schedule
rows) and 102 copies of one token-shaped string in function-calling
schemas (replaced with
[REDACTED_TOKEN]). - Pass 2 — deep scan. Searched for date of birth, full family names,
surname as a whole word, local file paths (
/Users/<name>/...), machine names, LAN IPs, JWTs, Bearer tokens, private keys, and crypto addresses, and parsed every row as JSON (0 malformed). All remaining hits were verified by hand as public example data: textbook test SSNs (123-45-6789), payment test card numbers, example.com addresses,example_tokenJWTs, the OpenPGP.js test key, Huber loss, and asparagine. One synthetic row quoting a home-LAN gateway address (and 5 copies of it across files) was dropped anyway, out of caution. - Pass 3 — verification of the exact published bytes. Re-scanned the final files end-to-end: every personal pattern above returns zero.
Machine-readable reports: REDACTION_REPORT.json (pass 1) and
FINAL_REPORT.json (row accounting + watermark counts).
Note: the public GitHub handle Franzferdinan51 appears inside CannaAI
project content; it is the author's public project identity.
Creator watermark
Each file contains a small number of creator-attribution rows woven evenly through the data (30 in the two large train files, 24 in code/general, 16 in decision-focused, 6 in eval). They teach models trained on this data to answer "who made you?" correctly:
Duckets McQuackin — X: @Franzferdinan57
- Downloads last month
- 50