31 lines
908 B
JSON
31 lines
908 B
JSON
|
{
|
||
|
|
"command": "agbenchmark start --test TestBasicMemory",
|
||
|
|
"completion_time": "2023-07-17-18:22",
|
||
|
|
"metrics": {
|
||
|
|
"run_time": "53.48 seconds",
|
||
|
|
"highest_difficulty": "basic: 2"
|
||
|
|
},
|
||
|
|
"tests": {
|
||
|
|
"TestBasicMemory": {
|
||
|
|
"data_path": "agbenchmark/challenges/memory/m1",
|
||
|
|
"is_regression": false,
|
||
|
|
"task": "Follow the instructions in the instructions_1.txt file",
|
||
|
|
"answer": "2314",
|
||
|
|
"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",
|
||
|
|
"metrics": {
|
||
|
|
"difficulty": "basic",
|
||
|
|
"success": true,
|
||
|
|
"success_%": 75.0,
|
||
|
|
"run_time": "53.252 seconds"
|
||
|
|
},
|
||
|
|
"reached_cutoff": false
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"config": {
|
||
|
|
"workspace": "${os.path.join(Path.home(), 'miniagi')}"
|
||
|
|
},
|
||
|
|
"additional": {
|
||
|
|
"model": "gpt-4"
|
||
|
|
}
|
||
|
|
}
|