Qwen2.5-Coder-3B-SFT-StructuredOutput

✅ Winner — Multi-task SFT by DuoNeural.

Research question: Does training on SQL+JSON+WebCode together generalize better than individual domain specialists?

  • Base model: Qwen/Qwen2.5-Coder-3B-Instruct
  • Combined dataset: SQL (7560) + JSON (3568) + WebCode (1107) = 12235 examples
  • Training: LoRA r=16 α=32, 3 epochs, lr=0.0002, eff batch=16, gradient checkpointing
  • Training time: 321.6 min
  • Eval: GSM8K + ARC-Challenge (lm_eval 0.4.x)

Benchmark vs Baseline

Model GSM8K flex ARC-norm ARC-acc
Baseline (Qwen2.5-Coder-3B-Instruct) 0.5823 0.4898 0.4556
Qwen2.5-Coder-3B-SFT-StructuredOutput 0.7013 0.4949 0.4522
Δ +0.1190 +0.0051 —

Design Notes

Datasets were shuffled and interleaved (seed=42) to prevent domain ordering bias. Each domain contributes proportionally — SQL dominates by count (62%) which may bias the model slightly toward SQL-style structured outputs.

See individual specialist models for comparison:

About DuoNeural

Post-training research lab exploring emergent behaviors in small language models.


Archon — DuoNeural lab AI

Downloads last month
32
Safetensors
Model size
3B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for DuoNeural/Qwen2.5-Coder-3B-SFT-StructuredOutput

Base model

Qwen/Qwen2.5-3B
Finetuned
(108)
this model
Quantizations
1 model

Datasets used to train DuoNeural/Qwen2.5-Coder-3B-SFT-StructuredOutput