diff options
Diffstat (limited to 'benchmark/agbenchmark/challenges/deprecated/memory/m1_id/data.json')
-rw-r--r-- | benchmark/agbenchmark/challenges/deprecated/memory/m1_id/data.json | 30 |
1 files changed, 30 insertions, 0 deletions
diff --git a/benchmark/agbenchmark/challenges/deprecated/memory/m1_id/data.json b/benchmark/agbenchmark/challenges/deprecated/memory/m1_id/data.json new file mode 100644 index 000000000..de88ba649 --- /dev/null +++ b/benchmark/agbenchmark/challenges/deprecated/memory/m1_id/data.json @@ -0,0 +1,30 @@ +{ + "category": [ + "memory" + ], + "cutoff": 150, + "dependencies": [ + "TestReadFile" + ], + "eval_id": "00227b7f-8952-4577-bfdb-c75db9f1fb19", + "ground": { + "answer": "2314", + "eval": { + "type": "file" + }, + "files": [ + "result.txt" + ], + "should_contain": [ + "2314" + ], + "should_not_contain": [] + }, + "info": { + "description": "Tests ability for the agent to remember information between each action. An id is presented initially and the agent has to remember it after reading 4 other files", + "difficulty": "basic", + "side_effects": [] + }, + "name": "BasicMemory", + "task": "Follow the instructions in the instructions_1.txt file" +} |