Results
2
issues of
tiffany
Related to issue #364 but unfinished from the #387 PR. Add basic few-shot and zero-shot prompts for OpenAI, Anthropic, DeepInfra, Google/Vertex AI models and make a notebook (or notebooks) to...
Related to issue https://github.com/HumanCompatibleAI/tensor-trust/issues/364 but unfinished from the https://github.com/HumanCompatibleAI/tensor-trust/pull/387 PR. Add logit_bias to the two APIs that support them (applies to OpenAI, Google/Vertex AI) to force a choice between true...