From be975901469a45bee98477a85f9c9d776522ca8d Mon Sep 17 00:00:00 2001 From: zefie Date: Wed, 4 Nov 2026 16:49:59 -0500 Subject: [PATCH] add readme and fix docker.sh --- README.md | 28 ++++++++++++++++++++++++++++ docker.sh | 3 +++ 2 files changed, 31 insertions(+) create mode 100644 README.md diff --git a/README.md b/README.md new file mode 100644 index 0000000..879031d --- /dev/null +++ b/README.md @@ -0,0 +1,28 @@ +Convert-to-Quant (`ctq`) is installed for normal quantizations. +`hf`, the Hugging Face CLI command, and llama.cpp's compiled binaries are also available. + +For GGUF quantizations, try: +- `hf_to_nvfp4_quant` (huggingface structure to GGUF) +- `gguf_to_nvfp4_quant` (GGUF to GGUF) +- `safetensors_to_nvfp4_quant` (ComfyUI style safetensors to GGUF (untested)) +- `hf_to_nvfp4_safetensors` (huggingface structure to ComfyUI style safetensors) + +For model info, try: +- `hf_info` (huggingface structure) +- `gguf_info` (GGUF structure) +--- + +Example: + +```bash +hf download coder3101/gemma-4-12B-it-heretic --local-dir gemma-4-12B-it-heretic +hf_to_nvfp4_gguf -o gemma-4-12B-it-heretic_nvfp4 gemma-4-12B-it-heretic +``` + +This would create a NVFP4 GGUF for Ollama/llama.cpp from `https://huggingface.co/coder3101/gemma-4-12B-it-heretic` + +The same directory can be quantized to one ComfyUI NVFP4 safetensors file: + +`hf_to_nvfp4_safetensors -o gemma-4-12B-it-heretic-NVFP4.safetensors gemma-4-12B-it-heretic` + +Conversions between formats without quantization is available in the custom commands with `--no-quant`. diff --git a/docker.sh b/docker.sh index 69b385f..6a49d82 100755 --- a/docker.sh +++ b/docker.sh @@ -8,6 +8,7 @@ get_latest() { latest_image=$(docker images --format '{{.CreatedAt}}\t{{.Repository}}:{{.Tag}}' zefie/nvfp4_quant \ | sort -r \ | cut -f2- \ + | grep -v latest \ | head -n 1) if [[ -z "$latest_image" ]]; then @@ -44,7 +45,9 @@ elif [ "${1}" == "push" ]; then # push dated one first, then latest "${0}" mklatest IMAGE=$(get_latest) + echo "Pushing ${IMAGE}" docker push "${IMAGE}" + echo "Pushing zefie/nvfp4_quant:latest" docker push "zefie/nvfp4_quant:latest" else echo "Usage: ${0} [build|run|mklatest]"