Issue with Loading a SLIM Model from a local directory

[chaitanya-gvs]

I am experiencing an issue when attempting to load the slim-sentiment-tool model using the load_model method in the ModelCatalog class. The method raises a ModelNotFoundException, indicating that it cannot identify the model card for the selected model.

Steps to Reproduce

1. Download the model files and place them in the directory /models/llmware/slim-sentiment-tool.

   from llmware.models import pull_model_from_hf, ModelCatalog
   model_card = ModelCatalog().lookup_model_card("slim-sentiment-tool")
   pull_model_from_hf(model_card, "/models/llmware/slim-sentiment-tool")

2. Attempt to load the model using the following code:

   from llmware.models import ModelCatalog
   
   def analyse_sentiment(text):
       slim_model = ModelCatalog().load_model("/models/llmware/slim-sentiment-tool")
       response = slim_model.function_call(text, get_logits=True)
       analysis = ModelCatalog().logit_analysis(response, slim_model.model_card, slim_model.hf_tokenizer_name)
       llm_response = response['llm_response']
       confidence_score = float(analysis['confidence_score'])
       return llm_response, confidence_score
   
   text = "I am happy"
   llm_response, confidence_score = analyse_sentiment(text)

3. Observe the following error:

   ModelNotFoundException: '/models/llmware/slim-sentiment-tool' could not be located
   

Expected Behavior

The load_model method should correctly identify and load the model card for the slim-sentiment-tool model.

Actual Behavior

The method raises a ModelNotFoundException, indicating that it cannot identify the model card for the selected model.

Environment

• llmware version: 0.3.3
• Python version: 3.11

Request for Assistance

Could you please assist in identifying why the load_model method is unable to find the model card from the local path and suggest any necessary changes to fix this issue?

Thank you!