Traian Rebedea
Traian Rebedea
Hi, @habanoz ! Unfortunately, if the `command-r` model does not work for generating intents, it does not really seem useful for Guardrails for now. I am sure that with a...
Hi @imintollms , For HF models you should use `nemoguardrails.llm.providers.HuggingFacePipelineCompatible` instead of the vanilla `HuggingFacePipeline`, similar to this: https://github.com/NVIDIA/NeMo-Guardrails/blob/8bb50afc68060e6c46d6fc557bcdb97cffec0fc8/examples/configs/llm/hf_pipeline_dolly/config.py#L60 This will solve your problem.
Hi @nashugame , I was able to run your sample code, I assume you are using `HuggingFaceH4/zephyr-7b-beta`. For me, the model required about 16-18 GB GPU RAM and the latency...
HI @ahsan3219 , There is an VertexAI Langchain wrapper that you can use. Is there a reason for not using it? I don't fully understand the code you shared above....
Hi @maxyousif15 ! I have not been able to replicate this bug. I have run with the configs you provided and the streaming example demo that I also assume you...
The message `I'm not sure what to say.` in `single_call` mode appears when the LLM is not able to generate all the necessary steps `(user intent, ... , bot intent,...
@mikemckiernan , that is right and Anthropic's thinking looks like a great example. The `parameters` field was created because many models (in Langchain, but also outside the Langchain ecosystem) have...
Tested this and works fine, the only thing is that we need to also have the `max_tokens` set and be higher that the `budget_tokens`. ```yaml models: - type: main engine:...
Hi @mikemckiernan , Actually the evaluation documentation probably needs a bit of changes. Right now we have two different type of evaluation tools and corresponding documentation : 1. An eval...