Phi 3 Mini 4k Python

🚀 phi-3-mini-4k-python

This is a text generation model fine - tuned from unsloth/Phi - 3 - mini - 4k - instruct - bnb - 4bit, trained 2x faster with Unsloth and Huggingface's TRL library.

🚀 Quick Start

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

📚 Documentation

Model Information

Evaluation Results

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Model Index

The model phi - 3 - mini - 4k - python has the following evaluation results on different tasks:

• Task: Text Generation (IFEval (0 - Shot))
  • Dataset: HuggingFaceH4/ifeval
  • Metrics: strict accuracy = 24.09
  • Source: Open LLM Leaderboard
• Task: Text Generation (BBH (3 - Shot))
  • Dataset: BBH
  • Metrics: normalized accuracy = 28.45
  • Source: Open LLM Leaderboard
• Task: Text Generation (MATH Lvl 5 (4 - Shot))
  • Dataset: hendrycks/competition_math
  • Metrics: exact match = 8.46
  • Source: Open LLM Leaderboard
• Task: Text Generation (GPQA (0 - shot))
  • Dataset: Idavidrein/gpqa
  • Metrics: acc_norm = 5.48
  • Source: Open LLM Leaderboard
• Task: Text Generation (MuSR (0 - shot))
  • Dataset: TAUR - Lab/MuSR
  • Metrics: acc_norm = 9.22
  • Source: Open LLM Leaderboard
• Task: Text Generation (MMLU - PRO (5 - shot))
  • Dataset: TIGER - Lab/MMLU - Pro
  • Metrics: accuracy = 28.63
  • Source: Open LLM Leaderboard

📄 License

This model is released under the apache - 2.0 license.