README.md

metadata

license: mit
language:
  - en
metrics:
  - accuracy
  - bleurt
  - bertscore
  - character
dataset:
  - katsukiai/DeepFocus-X3
library_name: transformers
inference: true
base_model:
  - Qwen/Qwen2.5-7B
  - Qwen/Qwen2.5-VL-7B-Instruct

Gated model

Mochii models all have gate because we (Mochii Team) want to record your information (we will never send the information you provide), Mochii models thank you for understanding.

MochiiXL

Mochii are well-tuned models, Models supported by Safetensors