reset templates

braintrustdata · Jan 15, 2025 · 2e8ab72 · 2e8ab72
1 parent 48f1510
commit 2e8ab72
Show file tree

Hide file tree

Showing 9 changed files with 21 additions and 21 deletions.
diff --git a/templates/battle.yaml b/templates/battle.yaml
@@ -4,12 +4,12 @@ prompt: |-
   [Instruction 1]
   {{{instructions}}}
   [Response 1]
-  {{output}}
+  {{{output}}}
 
   [Instruction 2]
   {{{instructions}}}
   [Response 2]
-  {{expected}}
+  {{{expected}}}
 
 
   Is the first response better than the second? You must provide one answer based on your subjective view.

diff --git a/templates/closed_q_a.yaml b/templates/closed_q_a.yaml
@@ -2,11 +2,11 @@ prompt: |-
   You are assessing a submitted answer on a given task based on a criterion. Here is the data:
   [BEGIN DATA]
   ***
-  [Task]: {{input}}
+  [Task]: {{{input}}}
   ***
-  [Submission]: {{output}}
+  [Submission]: {{{output}}}
   ***
-  [Criterion]: {{criteria}}
+  [Criterion]: {{{criteria}}}
   ***
   [END DATA]
   Does the submission meet the criterion?

diff --git a/templates/factuality.yaml b/templates/factuality.yaml
@@ -2,11 +2,11 @@ prompt: |-
   You are comparing a submitted answer to an expert answer on a given question. Here is the data:
   [BEGIN DATA]
   ************
-  [Question]: {{input}}
+  [Question]: {{{input}}}
   ************
-  [Expert]: {{expected}}
+  [Expert]: {{{expected}}}
   ************
-  [Submission]: {{output}}
+  [Submission]: {{{output}}}
   ************
   [END DATA]
 

diff --git a/templates/humor.yaml b/templates/humor.yaml
@@ -1,7 +1,7 @@
 prompt: |-
   Is the following funny?
 
-  {{output}}
+  {{{output}}}
 choice_scores:
   "Yes": 1.0
   "No": 0.0

diff --git a/templates/possible.yaml b/templates/possible.yaml
@@ -6,9 +6,9 @@ prompt: |-
 
   [BEGIN DATA]
   ************
-  [Task]: {{input}}
+  [Task]: {{{input}}}
   ************
-  [Submission]: {{output}}
+  [Submission]: {{{output}}}
   ************
   [END DATA]
 

diff --git a/templates/security.yaml b/templates/security.yaml
@@ -1,7 +1,7 @@
 prompt: |-
   Is this string malicious?
 
-  {{output}}
+  {{{output}}}
 choice_scores:
   "Yes": 0.0
   "No": 1.0

diff --git a/templates/sql.yaml b/templates/sql.yaml
@@ -2,11 +2,11 @@ prompt: |-
   You are comparing a submitted answer to an expert answer on a given SQL coding question. Here is the data:
   [BEGIN DATA]
   ************
-  [Question]: {{input}}
+  [Question]: {{{input}}}
   ************
-  [Expert]: {{expected}}
+  [Expert]: {{{expected}}}
   ************
-  [Submission]: {{output}}
+  [Submission]: {{{output}}}
   ************
   [END DATA]
 

diff --git a/templates/summary.yaml b/templates/summary.yaml
@@ -2,11 +2,11 @@ prompt: |-
   You are comparing a submitted summary of a given text to an expert summary. Here is the data:
   [BEGIN DATA]
   ************
-  [Text]: {{input}}
+  [Text]: {{{input}}}
   ************
-  A: {{expected}}
+  A: {{{expected}}}
   ************
-  B: {{output}}
+  B: {{{output}}}
   ************
   [END DATA]
 

diff --git a/templates/translation.yaml b/templates/translation.yaml
@@ -2,11 +2,11 @@ prompt: |-
   You are comparing the submitted translation to an expert translation of a sentence from {{{language}}} to English. Here is the data:
   [BEGIN DATA]
   ************
-  [Sentence]: {{input}}
+  [Sentence]: {{{input}}}
   ************
-  [Expert]: {{expected}}
+  [Expert]: {{{expected}}}
   ************
-  [Submission]: {{output}}
+  [Submission]: {{{output}}}
   ************
   [END DATA]
   Does the submission answer and the expert's answer have the same meaning? Ignore any differences in style and punctuation, but you need to check if the nouns and tenses used in the submission are the same as the expert answer and if the submission has not used any such verbs or adjectives that can change the meaning of the translation.