---

language:

• tr

license: apache-2.0

base_model: Qwen/Qwen3.5-0.8B

tags:

• function-calling
• tool-use
• turkish
• gguf
• llama-cpp

datasets:

• alperiox/turkish-function-calling-devtools

pipeline_tag: text-generation

---

Qwen3.5-0.8B Turkish Function Calling (GGUF)

Qwen3.5-0.8B modelinin Türkçe function calling için fine-tune edilmiş GGUF versiyonu.

Kullanım

llama-server ile (önerilen)

# İndir
huggingface-cli download alperiox/qwen35-0.8b-turkish-fc-GGUF qwen35-0.8b-turkish-fc-q8.gguf

# Çalıştır (Metal/CUDA/CPU destekli)
llama-server -m qwen35-0.8b-turkish-fc-q8.gguf --port 8090 -ngl 99

# Sorgu (OpenAI-uyumlu API)
curl http://localhost:8090/v1/chat/completions   -H "Content-Type: application/json"   -d '{ "messages": [{"role": "user", "content": "src/ dizinindeki Python dosyalarini bul"}] }'

Performans

| Model | Boyut | Fonksiyon Doğruluğu |

|---|---|---|

| Qwen3.5 0.8B (base) | 0.8B | 4% |

| Qwen2.5 1.5B (base) | 1.5B | 54% |

| Qwen3.5 0.8B (fine-tuned) | 0.8B | ~88% |

Quantization

• Format: Q8_0
• Boyut: 774MB

Eğitim verisi: alperiox/turkish-function-calling-devtools

HF model: alperiox/qwen35-0.8b-turkish-fc