Agent not performing function calls for 25% of calls
Hi, for quite a few calls (~25%) recently, the agent is saying it's checking availability for a specific practitioner, but it actually doesn't check. Instead, it leaves the user waiting indefinitely, then hangup on them
I get that hallucinations are a part of LLM, but is there anything I can do to try and reduce their probability? It feels high for gpt4
8438d69f-2d81-4a0d-813a-8bcfbb41bd4d
a83d344c-bc77-400b-87dd-5b7442d6d0a7
a5769f90-d2d0-413f-af3a-bf5eb87d6d1b
I get that hallucinations are a part of LLM, but is there anything I can do to try and reduce their probability? It feels high for gpt4
8438d69f-2d81-4a0d-813a-8bcfbb41bd4d
a83d344c-bc77-400b-87dd-5b7442d6d0a7
a5769f90-d2d0-413f-af3a-bf5eb87d6d1b
