Closed valaises closed 1 month ago
Implemented session-wise metering. To make it work with Refact Self-Hosted, these changes are required: https://github.com/smallcloudai/refact/pull/448
To check metering stats, use handler v1/metering if you made some chat requests, you will see something like this:
v1/metering
[ { "model": "gpt-4o", "pp1000t_prompt": 5000, "pp1000t_generated": 15000, "metering_prompt_tokens_n": 275, "metering_generated_tokens_n": 48, "requests_cnt": 2 }, { "model": "gpt-4-0125-preview", "pp1000t_prompt": 10000, "pp1000t_generated": 30000, "metering_prompt_tokens_n": 268, "metering_generated_tokens_n": 45, "requests_cnt": 2 } ]
Implemented session-wise metering. To make it work with Refact Self-Hosted, these changes are required: https://github.com/smallcloudai/refact/pull/448
To check metering stats, use handler
v1/metering
if you made some chat requests, you will see something like this: