OpenFineTrain
Collection
Very fine 55 trillion+ tokens worth of ungated huggingface datasets for LLM training across 80+ datasets. • 90 items • Updated • 2
user stringclasses 5
values | assistant stringclasses 5
values |
|---|---|
You are a Staff-Level Systems Engineer. Your task is to design and implement a completely self-contained Python script for a "Time-Traveling Virtual Machine" (TVM) with an execution cycle detector.
Architectural Requirements:
Memory & State:
A Stack (max 256 integers).
RAM (1024 16-bit integer slots).
Program Coun... | <think>The user wants a complete self-contained Python script. Let me think through the design carefully.
Key requirements:
VM with stack (max 256), RAM 1024 16-bit slots, PC.
ISA: PUSH, POP, ADD, SUB, LOAD, STORE, JMP, JZ, HALT.
step_forward() and step_backward() with a Delta Journal — no copies.
O(1) rolling hash (... |
"Write a single Python script using only the standard library (no pip, no external packages) that im(...TRUNCATED) | "<think>The user wants an extremely large, production-grade Raft implementation with a distributed t(...TRUNCATED) |
"Write a single Python script using only the standard library that implements a **full-featured, emb(...TRUNCATED) | "<think>The user is asking for an enormous, extremely complex piece of software: a full relational d(...TRUNCATED) |
"Task: Implement the core logic and TCP server for a custom, high-throughput in-memory Key-Value sto(...TRUNCATED) | "<think>The user wants implementation. Need provide code. Need think through design. Need provide co(...TRUNCATED) |
"Task: Implement a Wait-Free, Multi-Producer/Multi-Consumer (MPMC) Skip List in Rust, combined with (...TRUNCATED) | "<think>We need to provide a response that includes:\n\nTLA+ specification for a wait-free MPMC skip(...TRUNCATED) |
Dataset description
5 complex reasoning and coding requests (generated by deepseek v4 pro and gemini 3.1 pro) and their responses (generated by kimi k3).
Dataset Details
- Curated by: MoreThought
- Funded by: MoreThought
- Shared by: MoreThought
- License: apache 2.0
Uses
Finetuning LLMs or SLMs of any size on advanced kimi k3 distilled reasoning traces.
- Downloads last month
- 197