3.125-Bit LLM quantization bypassing tensor cores — Score 2/100 · Other | indexator.ai | Indexator.ai