Release v1.2
🚀Major Update: Introducing WizardLM 30B Version.
- On difficulty-balanced Evol-Instruct testset, evaluated by GPT-4: WizardLM-30B achieves 97.8% of ChatGPT, Guanaco-65B achieves 96.6%, and WizardLM-13B achieves 89.1%.
- We provide a comparison between the performance of the WizardLM-30B and ChatGPT on different skills to establish a reasonable expectation of WizardLM's capabilities.