bling-qwen-mini-tool is a RAG-finetuned version on Qwen2-1.5B for use in fact-based context question-answering, packaged with 4_K_M GGUF quantization, providing a very fast, very small inference implementation for use on CPUs.
Benchmark Tests
Evaluated against the benchmark test: RAG-Instruct-Benchmark-Tester
1 Test Run with sample=False & temperature=0.0 (deterministic output) - 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: 93.5 correct out of 100
--Not Found Classification: 75.0%
--Boolean: 87.5%
--Math/Logic: 70.0%
--Complex Questions (1-5): 3 (Average)
--Summarization Quality (1-5): 3 (Average)
--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).
To pull the model via API:
from huggingface_hub import snapshot_download
snapshot_download("llmware/bling-qwen-mini-tool", local_dir="/path/on/your/machine/", local_dir_use_symlinks=False)
Load in your favorite GGUF inference engine, or try with llmware as follows:
from llmware.models import ModelCatalog
model = ModelCatalog().load_model("bling-qwen-mini-tool")
response = model.inference(query, add_context=text_sample)
Note: please review config.json in the repository for prompt wrapping information, details on the model, and full test set.
