Qwen3.5-9B-Base CPT Merged
Model ini merupakan hasil merge antara:
- Base model:
Qwen/Qwen3.5-9B-Base
- Adapter: LoRA hasil Continual Pre-Training
- Architecture:
Qwen3_5ForConditionalGeneration
- Output dtype: BF16
- Format: Hugging Face Transformers SafeTensors
Model ini baru melalui tahap Continual Pre-Training dan belum
mencakup adaptor Supervised Fine-Tuning.
