Error: Requested tokens exceed context window of 512

[kelheor]

I'm trying to use Vector Store example. I changed some nodes, so I'm using TextLoader, HuggingFaceEmbeddings and LlamaCpp instead of OpenAI.

When I'm trying to ask simple question about text, I'm getting this error:

C:\Users\kelhe\anaconda3\lib\site-packages\langflow\api\chat_manager.py:215 │ │ in process_graph │ │ │ │ 212 │ # Generate result and thought │ │ 213 │ try: │ │ 214 │ │ logger.debug("Generating result and thought") │ │ ❱ 215 │ │ result, intermediate_steps = await get_result_and_steps( │ │ 216 │ │ │ langchain_object, chat_message.message or "", websocket=we │ │ 217 │ │ ) │ │ 218 │ │ logger.debug("Generated result and intermediate_steps") │ │ │ │ C:\Users\kelhe\anaconda3\lib\site-packages\langflow\interface\run.py:198 in │ │ get_result_and_steps │ │ │ │ 195 │ │ ) │ │ 196 │ │ thought = format_actions(intermediate_steps) if intermediate_s │ │ 197 │ except Exception as exc: │ │ ❱ 198 │ │ raise ValueError(f"Error: {str(exc)}") from exc │ │ 199 │ return result, thought │ │ 200 │ │ 201 │ ╰──────────────────────────────────────────────────────────────────────────────╯ ValueError: Error: Requested tokens exceed context window of 512

I guess the problem is in this parameter: llama_model_load_internal: n_ctx = 512 But I can't find any place to change it. Max Tokens param = 2048 did not change anything.

How can I increase context window?

[DaiZack]

Some issue here

[maggenti]

I am having the same exact issue. Langflow Only ChatGPT works, it's a conspiracy to use ChatGPT only! LOL... I love the concept of langflow for prototyping but I expected it to work...

[ogabrielluiz]

Hey @kelheor

Our next release will give a lot more flexibility to choose what fields appear in the node.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[lucaseduoli]

This is already fixed on the latest release.