bling-tiny-llama-v0

Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

bling-tiny-llama-v0 is part of the BLING ("Best Little Instruct No-GPU-required...") model series, RAG-instruct trained on top of a TinyLlama-1.1b base model.

BLING models are fine-tuned with high-quality custom instruct datasets, designed for rapid testing and prototyping in RAG scenarios.

Benchmark Tests

Evaluated against the benchmark test: RAG-Instruct-Benchmark-Tester
Average of 2 Test Runs with 1 point for correct answer, 0.5 point for partial correct or blank / NF, 0.0 points for incorrect, and -1 points for hallucinations.

--Accuracy Score: 86.5 correct out of 100
--Not Found Classification: 85.0%
--Boolean: 82.50%
--Math/Logic: 37.50%
--Complex Questions (1-5): 3
--Summarization Quality (1-5): 3 (Coherent, extractive)
--Hallucinations: No hallucinations observed in test runs.

For test run results (and good indicator of target use cases), please see the files ("core_rag_test" and "answer_sheet" in this repo).

Model Description

<!-- Provide a longer summary of what this model is. -->

• Developed by: llmware
• Model type: TinyLlama
• Language(s) (NLP): English
• License: Apache 2.0
• Finetuned from model: TinyLlama-1.1b - 2.5T checkpoint

Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

BLING is designed for enterprise automation use cases, especially in knowledge-intensive industries, such as financial services, legal and regulatory industries with complex information sources.

BLING models have been trained for common RAG scenarios, specifically: question-answering, key-value extraction, and basic summarization as the core instruction types without the need for a lot of complex instruction verbiage - provide a text passage context, ask questions, and get clear fact-based responses.

Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

Any model can provide inaccurate or incomplete information, and should be used in conjunction with appropriate safeguards and fact-checking mechanisms.

How to Get Started with the Model

The fastest way to get started with BLING is through direct import in transformers:

from transformers import AutoTokenizer, AutoModelForCausalLM  
tokenizer = AutoTokenizer.from_pretrained("bling-tiny-llama-v0")  
model = AutoModelForCausalLM.from_pretrained("bling-tiny-llama-v0")  

Please refer to the generation_test .py files in the Files repository, which includes 200 samples and script to test the model. The generation_test_llmware_script.py includes built-in llmware capabilities for fact-checking, as well as easy integration with document parsing and actual retrieval to swap out the test set for RAG workflow consisting of business documents.

The BLING model was fine-tuned with a simple "<human> and <bot> wrapper", so to get the best results, wrap inference entries as:

full_prompt = "<human>: " + my_prompt + "\n" + "<bot>:"

The BLING model was fine-tuned with closed-context samples, which assume generally that the prompt consists of two sub-parts:

1. Text Passage Context, and
2. Specific question or instruction based on the text passage

To get the best results, package "my_prompt" as follows:

my_prompt = {{text_passage}} + "\n" + {{question/instruction}}

If you are using a HuggingFace generation script:

# prepare prompt packaging used in fine-tuning process
new_prompt = "<human>: " + entries["context"] + "\n" + entries["query"] + "\n" + "<bot>:"

inputs = tokenizer(new_prompt, return_tensors="pt")  
start_of_output = len(inputs.input_ids[0])

#   temperature: set at 0.3 for consistency of output
#   max_new_tokens:  set at 100 - may prematurely stop a few of the summaries

outputs = model.generate(
        inputs.input_ids.to(device),
        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_output:],skip_special_tokens=True)  

Citation [optional]

This BLING model was built on top of a "TinyLlama" model base - for more information about the "TinyLlama" model, please see the paper referenced below:

@online{tinyllama, author = {Peiyuan Zhang, Guangtao Zeng, Tianduo Wang and Wei Lu}, title = {TinyLlama}, url = {https://github.com/jzhang38/TinyLlama}, year = {2023}, month = {Sep} }

Model Card Contact

Darren Oberst & llmware team