Phi-2 TPC-DS SQL LoRA β€” v4

Training Details

  • Base model: microsoft/phi-2
  • Method: LoRA (r=16, alpha=32)
  • New samples this run: 100
  • Cumulative training samples: 250
  • Difficulty covered: Medium
  • Epochs: 5
  • Strategy: Curriculum Learning (Easy β†’ Medium β†’ Hard)

Incremental Training Chain

v1 β†’ v2 (70 Easy) β†’ v3 (+80 Easy/Medium) β†’ v4 (+100 Medium) β†’ v5 (+150 Medium/Hard) β†’ v6 (+150 Hard)

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support