2023-06-06 03:58:29 +02:00
|
|
|
from khoj.processor.conversation import utils
|
|
|
|
from langchain.schema import ChatMessage
|
|
|
|
import factory
|
|
|
|
import tiktoken
|
|
|
|
|
2023-06-06 04:29:23 +02:00
|
|
|
|
2023-06-06 03:58:29 +02:00
|
|
|
class ChatMessageFactory(factory.Factory):
|
|
|
|
class Meta:
|
|
|
|
model = ChatMessage
|
|
|
|
|
2023-06-06 04:29:23 +02:00
|
|
|
content = factory.Faker("paragraph")
|
|
|
|
role = factory.Faker("name")
|
|
|
|
|
2023-06-06 03:58:29 +02:00
|
|
|
|
|
|
|
class TestTruncateMessage:
|
|
|
|
max_prompt_size = 4096
|
2023-06-06 04:29:23 +02:00
|
|
|
model_name = "gpt-3.5-turbo"
|
2023-06-06 03:58:29 +02:00
|
|
|
encoder = tiktoken.encoding_for_model(model_name)
|
|
|
|
|
|
|
|
def test_truncate_message_all_small(self):
|
|
|
|
chat_messages = ChatMessageFactory.build_batch(500)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in chat_messages])
|
|
|
|
|
2023-06-07 08:25:43 +02:00
|
|
|
prompt = utils.truncate_messages(chat_messages, self.max_prompt_size, self.model_name)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in prompt])
|
2023-06-06 03:58:29 +02:00
|
|
|
|
|
|
|
# The original object has been modified. Verify certain properties
|
|
|
|
assert len(chat_messages) < 500
|
|
|
|
assert len(chat_messages) > 1
|
|
|
|
assert prompt == chat_messages
|
|
|
|
assert tokens <= self.max_prompt_size
|
|
|
|
|
|
|
|
def test_truncate_message_first_large(self):
|
|
|
|
chat_messages = ChatMessageFactory.build_batch(25)
|
2023-06-14 01:32:47 +02:00
|
|
|
big_chat_message = ChatMessageFactory.build(content=factory.Faker("paragraph", nb_sentences=2000))
|
2023-06-06 03:58:29 +02:00
|
|
|
big_chat_message.content = big_chat_message.content + "\n" + "Question?"
|
|
|
|
copy_big_chat_message = big_chat_message.copy()
|
|
|
|
chat_messages.insert(0, big_chat_message)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in chat_messages])
|
|
|
|
|
2023-06-07 08:25:43 +02:00
|
|
|
prompt = utils.truncate_messages(chat_messages, self.max_prompt_size, self.model_name)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in prompt])
|
2023-06-06 03:58:29 +02:00
|
|
|
|
|
|
|
# The original object has been modified. Verify certain properties
|
|
|
|
assert len(chat_messages) == 1
|
|
|
|
assert prompt[0] != copy_big_chat_message
|
|
|
|
assert tokens <= self.max_prompt_size
|
|
|
|
|
|
|
|
def test_truncate_message_last_large(self):
|
|
|
|
chat_messages = ChatMessageFactory.build_batch(25)
|
2023-06-06 04:29:23 +02:00
|
|
|
big_chat_message = ChatMessageFactory.build(content=factory.Faker("paragraph", nb_sentences=1000))
|
2023-06-06 03:58:29 +02:00
|
|
|
big_chat_message.content = big_chat_message.content + "\n" + "Question?"
|
|
|
|
copy_big_chat_message = big_chat_message.copy()
|
2023-06-06 04:29:23 +02:00
|
|
|
|
2023-06-06 03:58:29 +02:00
|
|
|
chat_messages.append(big_chat_message)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in chat_messages])
|
|
|
|
|
2023-06-07 08:25:43 +02:00
|
|
|
prompt = utils.truncate_messages(chat_messages, self.max_prompt_size, self.model_name)
|
|
|
|
tokens = sum([len(self.encoder.encode(message.content)) for message in prompt])
|
2023-06-06 03:58:29 +02:00
|
|
|
|
|
|
|
# The original object has been modified. Verify certain properties
|
|
|
|
assert len(chat_messages) < 26
|
|
|
|
assert len(chat_messages) > 1
|
|
|
|
assert prompt[0] != copy_big_chat_message
|
2023-06-07 08:25:43 +02:00
|
|
|
assert tokens <= self.max_prompt_size
|