metadata
quantized_by: bartowski
pipeline_tag: text-generation
base_model_relation: quantized
base_model: openai/gpt-oss-20b
Llamacpp experimental Quantizations of gpt-oss-20b by Open AI
Using llama.cpp branch gpt-oss-mxfp4, PR here: https://github.com/ggml-org/llama.cpp/pull/15091
You must use that branch in order to run the model
Original model: https://huggingface.co/openai/gpt-oss-20b
This is a single static quant in the new MXFP4 format, rest of sizes will come after PR is merged