SLIM-BOOLEAN

slim-boolean is an experimental model designed to implement a boolean question answering function call using a 2.7B parameter specialized model. As an input, the model takes a context passage, a yes-no question, and an optional (explain) parameter, and as output, the model generates a python dictionary with two keys - 'answer' which contains the 'yes/no' classification, and 'explain' which provides a text snippet from the passage that was the basis for the classification, e.g.:

    {'answer': ['yes'], 'explanation': ['the results exceeded expectations by 3%'] }

This model is fine-tuned on top of llmware/bling-stable-lm-3b-4e1t-v0, which in turn, is a fine-tune of stabilityai/stablelm-3b-4elt.

For fast inference, we would recommend using the'quantized tool' version, e.g., 'slim-boolean-tool'.

Prompt format:

function = "boolean"
params = "{insert yes-no-question} (explain)"
prompt = "<human> " + {text} + "\n" +
                      "<{function}> " + {params} + "</{function}>" + "\n<bot>:"

Transformers Script

model = AutoModelForCausalLM.from_pretrained("llmware/slim-boolean")
tokenizer = AutoTokenizer.from_pretrained("llmware/slim-boolean")

function = "boolean"
params = "did tesla stock price increase? (explain) "

text = "Tesla stock declined yesterday 8% in premarket trading after a poorly-received event in San Francisco yesterday, in which the company indicated a likely shortfall in revenue."  

prompt = "<human>: " + text + "\n" + f"<{function}> {params} </{function}>\n<bot>:"

inputs = tokenizer(prompt, return_tensors="pt")
start_of_input = len(inputs.input_ids[0])

outputs = model.generate(
    inputs.input_ids.to('cpu'),
    eos_token_id=tokenizer.eos_token_id,
    pad_token_id=tokenizer.eos_token_id,
    do_sample=True,
    temperature=0.3,
    max_new_tokens=100
)

output_only = tokenizer.decode(outputs[0][start_of_input:], skip_special_tokens=True)

print("output only: ", output_only)  

# here's the fun part
try:
    output_only = ast.literal_eval(llm_string_output)
    print("success - converted to python dictionary automatically")
except:
    print("fail - could not convert to python dictionary automatically - ", llm_string_output)

Using as Function Call in LLMWare

from llmware.models import ModelCatalog
slim_model = ModelCatalog().load_model("llmware/slim-boolean")
response = slim_model.function_call(text,params=["did the stock price increase? (explain)"], function="boolean")

print("llmware - llm_response: ", response)

Model Card Contact

Darren Oberst & llmware team

Join us on Discord