Dataset · greghavens
Kimi K3 traces for coding, tool use, and instructions
greghavens/kimi-k3-coding-and-debugging-traces
582 trajectories and 3,956 rows for distilling agentic coding and tool use.
What's inside
This dataset contains Kimi K3 behavioral trajectories covering instruction following, tool use, and agentic programming. It comprises 582 trajectories and 3,956 training rows, available in parquet (3 MB) and JSONL (72 MB). Generated with the open-source moonshiner harness to distill verified traces. Language: English.
What you can do with it
- SFT for agentic coding models
- LoRA fine-tuning for tool use
- Evaluating instruction following and agentic trajectories
- Tasks
- Text generation
- Languages
- English
- Size
- 1K–10K rows
- Downloads this month
- 2.5K
- Likes
- 68
Models trained on it
- mradermacher/Qwen3.5-9B-Kimi-k3-Distilled-GGUF 855.2K
- mradermacher/Qwen3.5-9B-Kimi-k3-Distilled-i1-GGUF 5.4K
- somenetworktrouble/Qwen3.5-9B-Kimi-k3-Distilled-heretic-GGUF 4.5K
- mradermacher/Agents-A1-4B-kimi-Preview-heretic-GGUFText generation 1.4K
- khazarai/Qwen3.5-9B-Kimi-k3-Distilled-GGUFImage text → text 605
- hotdogs/Agents-A1-4B-kimi-Preview-heretic-GGUFText generation 538
- khazarai/Qwen3.5-9B-Kimi-k3-DistilledImage text → text 457
- hotdogs/Agents-A1-4B-kimi-Preview-GGUFImage text → text 393
How to load it
from datasets import load_dataset
ds = load_dataset("greghavens/kimi-k3-coding-and-debugging-traces")