@@ -96,9 +95,16 @@ class BARTResponseAgent(object):
Note: Turn numbers will NOT be provided as input
Return a dictionary with the following format
"use_api": True/False - Note that this cannot be used when using GPU
"prompts": [ <list of the prompts that go as "content" to the api > ] - Note that every call is independent and we don't use threads
"max_generated_tokens": [ list of ints for the max generation limit on each call] - Note that the submission will fail if the total generation limit is exceeded
"final_responses: [ <list of strings with the final responses> ] - Only used when use_api is set to False
You will be provided with a batch of upto 50 independent conversations
Return a string for every conversation
Input 1
[
...
...
@@ -30,6 +29,19 @@ class DummyResponseAgent(object):
Note: Turn numbers will NOT be provided as input
Return a dictionary with the following format
"use_api": True/False - Note that this cannot be used when using GPU
"prompts": [ <list of the prompts that go as "content" to the api > ] - Note that every call is independent and we don't use threads
"max_generated_tokens": [ list of ints for the max generation limit on each call] - Note that the submission will fail if the total generation limit is exceeded
"final_responses: [ <list of strings with the final responses> ] - Only used when use_api is set to False