Views
No views yet
| Property | Value |
|---|---|
| Base Model | Qwen3-0.6B |
| Parameters | 636M total, 40M trainable (6.34%) |
| Precision | BF16 |
| Training Method | LoRA via Unsloth |
| Context Length | 2048 |
| Config | Value |
|---|---|
| Dataset | 25,000 examples |
| Epochs | 1 |
| Batch Size | 16 (2 × 8 accum) |
| Steps | 1,200 |
| Hardware | Tesla T4 16GB |
| Time | 1.35 hours |
| Final Loss | 0.43 |
| Benchmark | Shadow-V2 | Qwen3-0.6B (base) |
|---|---|---|
| GSM8K (5-shot) | TBD | 42.3 |
| MATH (4-shot) | TBD | 18.2 |
| HumanEval (0-shot) | TBD | 28.0 |
1from transformers import AutoModelForCausalLM, AutoTokenizer
2
3model = AutoModelForCausalLM.from_pretrained("Redhanuman/Shadow-V2")
4tokenizer = AutoTokenizer.from_pretrained("Redhanuman/Shadow-V2")
5
6prompt = "Solve: If 3x + 7 = 22, find x.\nAnswer:"
7inputs = tokenizer(prompt, return_tensors="pt")
8outputs = model.generate(**inputs, max_new_tokens=256)
9print(tokenizer.decode(outputs[0], skip_special_tokens=True))