After seeing number of Turkish based Large Language Model works are increasing, I decided to contribute Trendyol Group's Base LLM model with quantizing them with training 24 different LLM models for both base and chat models. Besides, I created a HF Spaces for testing TrendyolLLMChat model on HF directly. 🤗

Quantization is applied as GGUF conversion into 12 different size of models each. All training phases completed on Google Cloud free-tier.

• Quantized "base" models repo:

https://huggingface.co/tolgadev/Trendyol-LLM-7b-base-v0.1-GGUF

• Quantized "chat" models repo:

https://huggingface.co/tolgadev/Trendyol-LLM-7b-chat-v0.1-GGUF

Chat_with_TrendyolLLM Space on HuggingFace

Hope you'll like my work! :)