ChesterProgrammer/Qwen3.5-9B-ConsistentChat-100steps

1---2base_model: unsloth/Qwen3.5-9B3tags:4- text-generation-inference5- transformers6- unsloth7- qwen3_58license: apache-2.09language:10- en11---12 13# Uploaded finetuned  model14 15- **Developed by:** ChesterProgrammer16- **License:** apache-2.017- **Finetuned from model :** unsloth/Qwen3.5-9B18 19This qwen3_5 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.20 21Hyperparams22 23Method Lora(16 bit)24Epochs 025Batch size 826Grad Accum 227Learning rate 0.000228Optimizer AdamW 8-bit29Max steps 10030Context length 204831Warmup steps 432Packing False33weight decay 0.00134seed 340735 36LoRA37Rank 1638Alpha 3239Dropout 040Variant lora41 42Dataset jiawei-ucas/ConsistentChat43 44Elapsed: 22m 51s450.07 steps/s46Tokens: 157240047 48[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)49