Shadow-Coder v3 — Multi-Expert LoRA
Continually fine-tuned from Shadow-Coder v2, specializing in 8 coding domains.
Domains
| Domain | Examples |
|---|---|
| Fullstack Architecture | 750 |
| Algorithms (LeetCode) | 2,000 |
| SQL / Database | 2,000 |
| DevOps / Shell | 1,000 |
| Debugging / Code Review | 1,500 |
| Frontend (React/Vue) | 1,500 |
| Backend + Security | 2,000 |
| General Coding | 2,000 |
Training
- Base: Shadow-Coder v2 → merged → v3 LoRA
- GPU: AMD Radeon RX 9060 XT (ROCm 7.0)
- Method: LoRA r=16, alpha=32
- Steps: 8,000
Usage
from unsloth import FastLanguageModel
import torch
model, tokenizer = FastLanguageModel.from_pretrained(
model_name = "lhordking/Shadow-coder-v3-LoRA",
max_seq_length = 2048,
dtype = torch.bfloat16,
)
FastLanguageModel.for_inference(model)
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support