CodeFeedback-Filtered-Instruction — LLM Code Dataset
74k coding instruction pairs across Python, Java, JavaScript, and C++ with a focus on interactive code generation and debugging. Combines data from Magicoder, ShareGPT-Python, and original generation to create a diverse multi-language coding chat dataset. Used to train OpenCodeInterpreter.
Dataset Details
| Provider | m-a-p |
| Category | Code |
| Size | 74k Rows |
| License | Apache 2.0 |
| Downloads | 180k |
| Tags | Multi-Language, Python, Java, Code-Chat |
from datasets import load_dataset
ds = load_dataset("m-a-p/CodeFeedback-Filtered-Instruction")
Fine-tune with this dataset
Estimated VRAM to fine-tune with QLoRA (4-bit base model + LoRA adapters), using conservative defaults:
| 7B QLoRA | ~6GB VRAM |
| 13B QLoRA | ~10GB VRAM |
Check if your GPU can fine-tune this →
Related datasets
Frequently asked questions
Can I use CodeFeedback-Filtered-Instruction commercially?
Yes — CodeFeedback-Filtered-Instruction is released under Apache 2.0, a permissive license that allows commercial use, including training models you ship in a product. Check the dataset card for attribution requirements before release.
How much data does CodeFeedback-Filtered-Instruction contain, and do I need all of it?
CodeFeedback-Filtered-Instruction contains 74k Rows. You rarely need all of it: for style and format fine-tuning, a few hundred to a few thousand examples are enough — load a slice (e.g. split="train[:1000]") and scale up only if quality plateaus.
What is CodeFeedback-Filtered-Instruction best used for?
Training coding models that respond to execution feedback. It belongs to the Code section of our dataset hub, where you'll find alternatives and complementary sets.
← All datasets | Fine-Tuning Guide