--- license: apache-2.0 base_model: docling-project/ScreenVLM base_model_relation: quantized pipeline_tag: image-text-to-text tags: [gguf, screen-parsing, ui-understanding, llama.cpp] --- # ScreenVLM GGUF F16 and Q8_0 conversions of the [ScreenParse-v2 checkpoint](https://huggingface.co/docling-project/ScreenVLM/tree/78df0cc8498bbea1a03eb2816056fc0a76d2f63d), with a shared F16 vision encoder/projector. Q8_0 is derived from the original-weight F16 conversion. | File | Component | Quantization | Size | Status | | --- | --- | --- | --- | --- | | [ScreenVLM-Q8_0.gguf](./ScreenVLM-Q8_0.gguf) | decoder | Q8_0 | 239,333,536 bytes | Available | | [ScreenVLM-F16.gguf](./ScreenVLM-F16.gguf) | decoder | F16 | 447,246,496 bytes | Available | | [mmproj-ScreenVLM-F16.gguf](./mmproj-ScreenVLM-F16.gguf) | vision | F16 | 190,031,392 bytes | Available | ## Usage Use [llama.cpp at `0ec4027`](https://github.com/ggml-org/llama.cpp/tree/0ec4027a2a0183de87c007069adddc06e665f7de): ```sh llama-server -m ScreenVLM-Q8_0.gguf --mmproj mmproj-ScreenVLM-F16.gguf \ -ngl 99 -c 4096 -np 1 --jinja --special \ --host 127.0.0.1 --port 8080 ``` Send one screenshot followed by `Generate the screen representation for this UI:`. `--special` retains the `