Kye
|
c5f62ec276
|
prompt clean up
|
2 years ago |
Kye
|
c7c8bb2cfd
|
clear prompts
|
2 years ago |
Kye
|
4b1ee86aa7
|
prompt reliability
|
2 years ago |
Kye
|
46624904a7
|
clean up in init boss mode
|
2 years ago |
Kye
|
3cf1ff8206
|
In this updated code, I've removed the verbose parameter from BossNode and BabyAGI.from_llm as it's not mentioned in the BabyAGI setup code you shared. I've also removed the assert isinstance(llm, dict) check, as llm is expected to be an instance of OpenAI or a similar class, not a dictionary.
|
2 years ago |
Kye
|
059ca8367e
|
clean up run swarms
|
2 years ago |
Kye
|
efc213eb3f
|
no verbose
|
2 years ago |
Kye
|
1c757d119d
|
worker tool node
|
2 years ago |
Kye
|
c5dd1da5a2
|
clean up back to old tool logic
|
2 years ago |
Kye
|
6b3fba50ea
|
new tool style
|
2 years ago |
Kye
|
f657a2e921
|
extensive logging for swarms and workernode
|
2 years ago |
Kye
|
0ecfd58707
|
assetions
|
2 years ago |
Kye
|
4f79c604e8
|
The root cause of this error is that the initialize_tools method in the Swarms class is passing an instance of OpenAI to the load_qa_with_sources_chain function, which then passes it to the LLMChain constructor. The LLMChain constructor expects a dictionary for the llm parameter, but it's receiving an instance of OpenAI instead.
|
2 years ago |
Kye
|
08852720dd
|
init llm clean up
|
2 years ago |
Kye
|
76094e5fd7
|
openai class for run swarms
|
2 years ago |
Kye
|
76281a0aa0
|
clean up
|
2 years ago |
Kye
|
e0d3249506
|
chat history
|
2 years ago |
Kye
|
f018b344d8
|
no omni agent for now
|
2 years ago |
Kye
|
1784e119a1
|
no omni agent for now
|
2 years ago |
Kye
|
7a46f81566
|
re-architecture
|
2 years ago |
Kye
|
4264c31517
|
clean up
|
2 years ago |