Synthetic Data & Reasoning Traces (Mimesis)
Collection
Synthetic training data generated with Mimesis, a local vLLM-based data generation pipeline, across multiple model backends. • 10 items • Updated
Questions stringclasses 3
values |
|---|
Wirte a python code to multiply two numbers |
Wirte a python code to make a palidrome program |
write a python code to to make a calcualtor app |
Seeds
A small collection of seed prompts and question starters used for data generation and fine-tuning pipelines.
Dataset Details
| Field | Value |
|---|---|
| Language | English |
| License | Apache 2.0 |
| Format | CSV |
Schema
| Column | Type | Description |
|---|---|---|
Questions |
string | A seed question or prompt string |
Usage
from datasets import load_dataset
ds = load_dataset("shaafsalman/seeds")
print(ds["train"][0])
# {'Questions': 'Write a python code to multiply two numbers'}
- Downloads last month
- 24