Datasets:

faunix
/

Qwen3.8-27B-Distillation-40K

messages listlengths 2 2	domain stringclasses 4 values	category stringclasses 96 values	source stringclasses 12 values	tokens_total int64 49 81.6k	tokens_think int64 10 79.7k
[ { "content": "When the curtains are opened, a canvas unfolds outside. Kanno marvels at all the blonde colours along the riverside — not tangerines, but blossoms instead.\n\n\"What a pity it's already late spring,\" sighs Mino with regret, \"one more drizzling night and they'd be gone.\"\n\n\"But these blends ar...	code	-	Modotte/CodeX-2M-Thinking	3,073	2,131
[ { "content": "Word s of length n is called k-complete if \n\n * s is a palindrome, i.e. s_i=s_{n+1-i} for all 1 ≤ i ≤ n; \n * s has a period of k, i.e. s_i=s_{k+i} for all 1 ≤ i ≤ n-k. \n\nFor example, \"abaaba\" is a 3-complete word, while \"abccba\" is not.\n\nBob is given a word s of length n consisting of...	code	-	Modotte/CodeX-2M-Thinking	3,005	2,004
[ { "content": "What is the value of the sum of the sixth powers of the solutions of the equation $x^{6}-16 x^{4}+16 x^{2}-1=0$?\n(A) 6375\n(B) 6482\n(C) 6515\n(D) 6660\n(E) 6662.", "role": "user" }, { "content": "<think>\nWe need to find the sum of the sixth powers of all solutions of $x^6 - 16x^4 + ...	math	-	open-r1/OpenR1-Math-220k	660	371
[ { "content": "A car travels at 60 miles per hour for 30 minutes, then at 40 miles per hour for 15 minutes. What is the average speed for the entire trip in miles per hour?", "role": "user" }, { "content": "<think>\nI need to find the average speed for the entire trip.\n\nAverage speed = Total distan...	logic	logic_puzzle	ianncity/GLM-5.2-Logic-Puzzles	364	121
[ { "content": "I have 6 shirts, 6 pairs of pants, and 6 hats. Each item comes in the same 6 colors (so that I have one of each item of each color). I refuse to wear an outfit in which all 3 items are the same color. How many choices for outfits do I have?", "role": "user" }, { "content": "<think>\...	math	-	nvidia/OpenMathReasoning	336	139
[ { "content": "Create a function that differentiates a polynomial for a given value of `x`.\n\nYour function will receive 2 arguments: a polynomial as a string, and a point to evaluate the equation as an integer.\n\n## Assumptions:\n\n* There will be a coefficient near each `x`, unless the coefficient equals `1`...	code	-	Modotte/CodeX-2M-Thinking	1,868	1,096
[ { "content": "A lot of frogs want to cross a river. A river is $w$ units width, but frogs can only jump $l$ units long, where $l < w$. Frogs can also jump on lengths shorter than $l$. but can't jump longer. Hopefully, there are some stones in the river to help them.\n\nThe stones are located at integer distance...	code	python	GetSoloTech/Code-Reasoning	11,701	10,730
[ { "content": "You are given an n × m rectangular table consisting of lower case English letters. In one operation you can completely remove one column from the table. The remaining parts are combined forming a new table. For example, after removing the second column from the table\n\nabcd\n\nedfg\n\nhijk\n\n \n...	code	cpp	GetSoloTech/Code-Reasoning	2,733	1,798
[ { "content": "I was born on February 29, 1996. If I just celebrated my 10th birthday, what is the current year?", "role": "user" }, { "content": "<think>\nThe user says they were born on February 29, 1996. They say they \"just celebrated my 10th birthday.\" Let me figure out what year that would be....	logic	logic_puzzle	ianncity/GLM-5.2-Logic-Puzzles	316	200
[ { "content": "Write a MySQL query to select all employees who are 30 years old or above, have a salary higher than the average salary of their department, have been with the company for at least 5 years, and have received at least one promotion during their tenure. #Reasoning Steps# 1. Get the current date. 2....	code	-	ianncity/Hunter-Alpha-Programming-160000x	2,028	760
[ { "content": "A subscription service charges customers exactly every 28 days. How many times will a customer be charged in a standard calendar year?", "role": "user" }, { "content": "<think>\nThe question asks how many times a customer will be charged in a standard calendar year if they are charged ...	logic	logic_puzzle	ianncity/GLM-5.2-Logic-Puzzles	837	724
[ { "content": "Create a code to save the first 100 prime numbers that are greater than 1,000,000 in a dynamically allocated array.", "role": "user" }, { "content": "<think>\nThe user wants code that:\n1. Finds the first 100 prime numbers greater than 1,000,000\n2. Stores them in a dynamically allocat...	code	-	ianncity/Hunter-Alpha-Programming-160000x	1,215	576
[ { "content": "The following paragraphs each describe a set of seven objects arranged in a fixed order. The statements are logically consistent within each paragraph. In an antique car show, there are seven vehicles: a limousine, a truck, a minivan, a tractor, a hatchback, a bus, and a convertible. The convertib...	logic	logical_deduction_seven_objects	lukaemon/bbh	1,034	549
[ { "content": "How does the coupling between electron transfer kinetics and ion transport within nanoporous solid-electrolyte interphases influence the voltage hysteresis observed in high‑energy lithium‑sulfur batteries under fast charge conditions?", "role": "user" }, { "content": "<think>\nThe user...	science	-	ianncity/GLM-5.2-Science	4,990	1,555
[ { "content": "Create an HTML form with two dropdowns to select from different countries and cities. Additionally, implement the following features: 1. Populate the country dropdown dynamically from a JSON file or API endpoint. 2. Enable cascading behavior between the two dropdowns, where selecting a country sh...	code	-	ianncity/Hunter-Alpha-Programming-160000x	5,092	171
[ { "content": "Can you provide an algorithm for performing a linear search on an unordered array in Python?", "role": "user" }, { "content": "<think>\nThe user is asking for a Python algorithm to perform a linear search on an unordered array. This is a straightforward programming question. Let me pro...	code	-	ianncity/Hunter-Alpha-Programming-160000x	724	127
[ { "content": "Generate a code snippet to print out the top 3 highest numbers from an array A. Write corresponding code in Python. A = [9, 4, 18, 24, 7, 22, 11]", "role": "user" }, { "content": "<think>\nThe user wants a Python code snippet that prints the top 3 highest numbers from the array A = [9,...	code	-	ianncity/Hunter-Alpha-Programming-160000x	454	181

End of preview. Expand in Data Studio

Qwen3.8-27B-Distillation (40K Traces)

Qwen3.8-27B-Distillation is a dataset containing 40,000 reasoning traces distilled from Qwen's latest model — Qwen3.8-27B. We generated this dataset locally by running the model on our own infrastructure. It covers 4 domains with prompts sourced from 12 diverse open-source datasets.

Dataset Overview

Metric	Value
Total Examples	40,000
Teacher Model	Qwen3.8-27B
Model Precision	FP8
Reasoning Effort	medium
Total Tokens	229,980,687 (~230M)
Assistant Tokens	224,414,445 (~224.4M)
Thinking Tokens	147,744,072 (~147.7M)
User Tokens	5,566,242 (~5.57M)
Avg Tokens per Example	~5,749
Inference Engine	vLLM
Estimated Generation Cost	~$80

Note: The dataset underwent basic deduplication: removal of duplicate and invalid examples.

Domains

Domain	Examples	Percentage
code	16,006	40.02%
math	10,994	27.48%
science	7,000	17.50%
logic	6,000	15.00%

Sources

Prompts were sampled from 12 open-source datasets:

Source	Examples	Percentage
ianncity/Hunter-Alpha-Programming-160000x	10,014	25.04%
ianncity/GLM-5.2-Logic-Puzzles	5,500	13.75%
ianncity/GLM-5.2-Science	5,051	12.63%
open-r1/OpenR1-Math-220k	4,000	10.00%
nvidia/OpenMathReasoning	3,994	9.98%
Modotte/CodeX-2M-Thinking	2,999	7.50%
GetSoloTech/Code-Reasoning	2,993	7.48%
AI-MO/NuminaMath-CoT	2,000	5.00%
EricLu/SCP-116K	1,400	3.50%
zwhe99/DeepMath-103K	1,000	2.50%
Hothan/OlympiadBench	549	1.37%
lukaemon/bbh	500	1.25%
Total	40,000	100%

Structure

Each example is a single JSONL line with the following schema:

{
  "messages": [
    {"role": "user", "content": "<prompt>"},
    {"role": "assistant", "content": "<think>...</think>\n\n<response>"}
  ],
  "domain": "code | math | science | logic",
  "category": "python | physics | logic_puzzle | ...",
  "source": "ianncity/Hunter-Alpha-Programming-160000x | ...",
  "tokens_total": 5749,
  "tokens_think": 3200
}

Field	Type	Description
`messages`	list[dict]	Standard chat format with `user` and `assistant` roles
`messages[1].content`	str	Always wrapped in `<think>...</think>` tags followed by the final response
`domain`	str	One of: `code`, `math`, `science`, `logic`
`category`	str	Fine-grained topic (e.g. `python`, `physics`, `olympiads`) or `-` if unspecified
`source`	str	Original dataset the prompt was sampled from
`tokens_total`	int	Total token count (user + assistant), counted with Qwen3.8 tokenizer
`tokens_think`	int	Token count inside `<think>` block only

Usage

With Hugging Face Datasets

from datasets import load_dataset

dataset = load_dataset("faunix/Qwen3.8-27B-Distillation-40K", split="train")

print(dataset[0]["messages"])   # full chat with <think> + response
print(dataset[0]["domain"])     # code / math / science / logic
print(dataset[0]["tokens_think"])  # thinking tokens count

With Axolotl

datasets:
  - path: faunix/Qwen3.8-27B-Distillation-40K
    type: chat_template
    chat_template: qwen3_5
    split: train

With Unsloth

from unsloth import FastLanguageModel
from datasets import load_dataset

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name="unsloth/Qwen3.5-9B",
    max_seq_length=8192,
)

dataset = load_dataset("faunix/Qwen3.8-27B-Distillation-40K", split="train")

Sponsor

This dataset would not exist without the help of lium.io! By granting us GPU access, we were able to create this dataset, and we will create more, and train many more interesting models! Therefore, as a sign of gratitude, we mention them here! :)

We are on...

And also subscribe to our HF organization! This way you won't miss new models and other projects we will be publishing! :)

☕ Support Faunix :)

We are two AI developers: Didiblud (12 years old) and Limen4ik (13 years old). Your support is very important to us — just like everyone else, we want to eat, so we are leaving our donation links below!

💳 Card / Apple Pay / PayPal: ⭐ Recommended! — app.lava.top/faunix_ai
💎 TON: UQArgd4zYOA0I5QSK7ylSlvgyaE5Md8pm-O-RyvSnTTk8Me0
💵 USDT (TRC-20): TVCgvfbrUQC1nKD6cHcSP7JGirEEx5f4Te

(Please note: 60% goes into our personal pockets fund, and 40% goes directly to Faunix development — buying domains, GPU hours, etc.)

Note

This dataset is open-source and distributed under the Apache 2.0 license, so you can use it in experiments and fine-tuning of your own models! We look forward to seeing what models you can fine-tune based on our dataset! :)

Creator: Faunix

Release Date: 22.08.26

Dataset Name: Qwen3.8-27B-Distillation-40K

:)

Downloads last month: 28