In this video we will explore on how we can use the Assistants web Ui to create out assistants and configure it then use it from our python script.
Code files for this project at Patreon: / explaining-how-102574497
UPDATE: When we set truncation strategy of type "last messages" then we must also pass "last messages key with the object as:
max_prompt_tokens=20000, # if you will use file search set this to 20k or more
truncation_strategy={"type": "last_messages", "last_messages": 5}, # also can be "auto"
otherwise like this:
max_prompt_tokens=20000, # if you will use file search set this to 20k or more
truncation_strategy={"type": "auto"}, # also can be "auto"
if you dont give the "last messages" param when using "last messages" strategy then in the future you may see 400 error being returned
Auto Streamer: www.autostreamer.live/
Search 200+ echohive videos and code download links:www.echohive.live/
CodeHive 900+ GPT python chat apps: www.codehive.app/
Chat with us on Discord:
/ discord
Follow on twitter(X) : / hive_echo
CHAPTERS;
00:00 Intro
01:25 Creating an assistant for chat only
04:15 Vectorstore setup
06:38 Testing tool use
08:04 Possible max prompt tokens bug
11:09 general review
15:02 Code review
21:11 Auto Streamer
#aiassistance #aiassistant #aiagents #openai #openaiapi #gpt4 #gpt4turbo #artificialintelligence #largelanguagemodels #pythonprojects #pythonai #pythonprogramming