Back to projectsFP.

Project 01 / Case study

Tax LLM SFT & Evaluation

A structured case study for supervised fine-tuning and comparing a base model with its fine-tuned counterpart in a tax-domain setting.

  • Qwen3-4B-Instruct-2507
  • QLoRA
  • SFT
  • Model evaluation
01

Problem

Coming soon — add verified project details here.

02

Context

Coming soon — add verified project details here.

03

My Role

Coming soon — add verified project details here.

04

Approach

Planned scope: document the QLoRA-based supervised fine-tuning workflow and the controlled comparison between the base and fine-tuned models.

05

Architecture / Workflow

Coming soon — add verified project details here.

06

Experiments

Coming soon — add verified project details here.

07

Results

Coming soon — add verified project details here.

08

Lessons Learned

Coming soon — add verified project details here.

09

Technologies

  • Qwen3-4B-Instruct-2507
  • QLoRA
  • SFT
  • Model evaluation