search

clp-research / clembench

A Framework for the Systematic Evaluation of Chat-Optimized Language Models as Conversational Agents and an Extensible Benchmark
MIT License
19 stars 26 forks source link

make path for results configurable #7

Closed davidschlangen closed 3 months ago

davidschlangen commented 8 months ago

The path where the episode transcripts are written should be configurable (for cli.py, or maybe more generally). Some people may prefer a location outside of the code repo.

davidschlangen commented 5 months ago

@briemadu, this has been done, right? Close?

briemadu commented 5 months ago

Right, done in a177c1ddc30857e355c6099e6412f4fcd9bf2344 with a fix by SH fbed0c4858a0ffa4099eff4012836d2f13c83ebd .

briemadu commented 5 months ago

@phisad this should also be done for the main cli

phisad commented 3 months ago

Fixed with https://github.com/clp-research/clembench/pull/66