Update README.md

README.md

@@ -11,6 +11,21 @@ license: cc-by-sa-4.0
 
 This model combines two of the most popular traditional classifier capabilities (sentiment analysis and named entity recognition) and re-images them as function calls on a small specialized decoder LLM, generating output in the form of a python dictionary with keys corresponding to sentiment and NER identifiers.  
 
+
+The intent of SLIMs is to forge a middle-ground between traditional encoder-based classifiers and open-ended API-based LLMs:  
+
+Compared to Encoder Classifiers 
+--With 4-bit quantization, this model stand-alone is 1.71 GB binary, which compares favorably with the combination of using FP32 versions of Roberta-Large for NER (1.42GB) and BERT for Sentiment (439 MB), which also require Pytorch and other external dependencies.  
+--2.7B parameters vs. Roberta-Large (336M) and BERT (110M)  
+--Provides intuitive natural language based responses that fit more naturally in LLM-based agent processes  
+--One model can provide multiple modalities of classification, and can be trained using 'union' of two separate datasets  
+
+Compared to Very Large Language Models  
+--Runs locally with no API charges
+--Does not require complex prompt instructions  
+--Small & Specialized - more limited capability but trained to do only 1-2 things well  
+
+
 [**slim-sa-ner-3b**](https://huggingface.co/llmware/slim-sa-ner-3b) is part of the SLIM ("**S**tructured **L**anguage **I**nstruction **M**odel") series, providing a set of small, specialized decoder-based LLMs, fine-tuned for function-calling.
 
 To pull the model via API: