Laguna XS.2 dense (k8) — single-shot SFT code-gen recovery

Dense ~3B model distilled from Laguna XS.2 (33B MoE), SFT-recovered on OpenCodeInstruct. Before SFT: repetitive non-code. After 10k steps: correct, documented Python. Single-shot baseline (not an agentic tool-calling model).

Training: 10k steps · batch 1 · seq 2048 · lr 2e-5 · bf16 · AdamW · loss 2.1 -> 0.71.

Downloads last month
-
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 Jessicacat0305/laguna-xs2-dense-k8-sft-opencode

Finetuned
(1)
this model

Dataset used to train Jessicacat0305/laguna-xs2-dense-k8-sft-opencode