hellno the optimist
@hellno.eth
tool calling improvements and benchmarking for vibes.engineering - using promptfoo to have a fast loop: 1) update dataset 2) test 3) fix errors - built a new dataset out of real user chat messages - now tests multi-step tool calling too - improved prompts and added xml style tags recommended by @alexpaden
1 reply
3 recasts
14 reactions
Luciano
@luciano
ty 500 $tipn
0 reply
0 recast
0 reaction