Files
heretic/config.nohumor.toml
Philipp Emanuel Weidmann 95dda4c4db feat: add benchmark scorer (#444)
* fix: improve print output of scorers

* feat: add benchmark scorer
2026-09-03 17:49:46 +05:30

73 lines
1.4 KiB
TOML

# Rename this file to config.toml, place it in the working directory
# that you run Heretic from, and edit the configuration to your liking.
max_response_length = 300
residual_plot_title = "PaCMAP Projection of Residuals for Serious/Humorous Prompts"
[good_prompts]
dataset = "mlabonne/harmless_alpaca"
split = "train[:400]"
column = "text"
residual_plot_label = "Serious prompts"
residual_plot_color = "royalblue"
[bad_prompts]
dataset = "UnstableLlama/jokes"
split = "train[:200]"
column = "text"
residual_plot_label = "Humorous prompts"
residual_plot_color = "darkorange"
[scorer.KeywordRate]
score_name = "Responses with humor"
keyword_markers = [
"😅",
"here's one",
"why did",
"walked into a bar",
"gets worse",
"the punchline",
"classic",
"dad joke",
"bad joke",
"pun intended",
"ba dum tss",
"rimshot",
"😂",
"😄",
"😆",
"haha",
"hehe",
"lol",
"funny",
"joke",
"humor",
"that's hilarious",
"you could say",
"one-liner",
"comedian",
"stand-up",
"unexpectedly",
"because apparently",
"to be fair",
"on the bright side",
"lmao",
"omg",
"rofl",
"silly",
"humorous",
"clever",
]
[scorer.KeywordRate.prompts]
dataset = "UnstableLlama/jokes"
split = "train[200:250]"
column = "text"
[scorer.KLDivergence.prompts]
dataset = "mlabonne/harmless_alpaca"
split = "test[:100]"
column = "text"