Temperature Issues with Dolly-3b

[eneadodi]

Hello!
I have been experimenting with some smaller LLMs recently and I can't seem to get the following error figured out for dolly-3b.

I have looked at issue-50:
#50

but the example linked here:
https://github.com/NVIDIA/NeMo-Guardrails/tree/main/examples/llm/hf_pipeline_dolly

seems to not exist anymore (I looked through some old commits to confirm).

When trying to use hugging_face endpoint, I get a temperature error. When I try to wrap a local llm as follows:

`
yaml_content = """
models:

• type: main
  engine: hf_pipeline_dolly
  """

@lru_cache
def get_dolly_v2_3b_llm():
repo_id = "databricks/dolly-v2-3b"
params = {"temperature": 0, "max_length": 1024,"trust_remote_code":True}

 # Use the first CUDA-enabled GPU, if any
 device = 0 if device_count() else -1

 llm = HuggingFacePipeline.from_model_id(
     model_id=repo_id, device=device, task="text-generation", model_kwargs=params
 )

 return llm

HFPipelineDolly = get_llm_instance_wrapper(
llm_instance=get_dolly_v2_3b_llm(), llm_type="hf_pipeline_dolly"
)

register_llm_provider("hf_pipeline_dolly", HFPipelineDolly)

config= RailsConfig.from_content(
colang_content=colang_content,
yaml_content=yaml_content
)
rag_rails = LLMRails(config)
`

I get a similar temperature error. That is:
Parameter temperature does not exist for WrapperLLM
and
Error "HuggingFacePipeline" object has no attribute "a_call" while execution generate_user_intent.

I get these errors regardless if I include temperature in params or not.

Any guide you can send me that shows me how to effectively use LLMs like this?