This post explores image classification using Gemini Pro and compares it with ChatGPT-3.5. It covers the setup of Gemini Pro, its limitations in code generation, and the need for adjustments to achieve an accuracy rate of 57%. In contrast, the code generated by ChatGPT-3.5 requires no modifications and achieves an accuracy rate of 77.29%.
- #ai#bard#chatgpt#computer-vision#deep-learning#gcp#genai#google-gemini#machine-learning#mnist#nlp#palm#python#pytorch#transformers#vertex-ai
158 Impressions