[ ] New llm_models::GoogleModels enum with all the LLMModel trait methods implemented
[ ] For GoogleModels::GeminProVertex (via Vertex AI) the code uses streaming and deserializes each response chunk, and for GoogleModels::GeminPro (via AI Studio) it just parses the full response
[ ] New structs in domains module to deserialize GeminiPro API response
[ ] New variables storing urls of Vertex and AI Studio APIs
[ ] Testing code in examples/use_completions, including getting auth token for service ccount
Changes:
llm_models::GoogleModels
enum with all the LLMModel trait methods implementedGoogleModels::GeminProVertex
(via Vertex AI) the code uses streaming and deserializes each response chunk, and forGoogleModels::GeminPro
(via AI Studio) it just parses the full responsedomains
module to deserialize GeminiPro API responseexamples/use_completions
, including getting auth token for service ccount