Ornith 1.0 9B is an INT4 quantized version of a 9B parameter language model hosted on Hugging Face. The specific repository cyankiwi/Ornith-1.0-9B-AWQ-INT4 provides the model weights using AWQ quantization for reduced memory usage during inference. It belongs to the class of foundation models.
The model includes a custom chat template implemented in Jinja that processes both text and vision inputs. The template tracks counts of images and videos separately, inserts special vision tokens such as vision_start, image_pad, and vision_end, and prevents images or videos from appearing in system messages. It also supports optional prefixes like "Picture N:" before visual content.
The tokenizer configuration defines an end-of-text token as the pad token. Delivery occurs through the standard Hugging Face model hub, allowing download and local execution with compatible inference frameworks that support AWQ and the provided chat template. No pricing information appears for the model itself, as it follows the typical open-access distribution on the platform.
No further capabilities, intended users, or performance claims are stated.
Ornith 1.0 9B sits in PulseGate's Foundation models & chat category. It focuses on running a capable 9B parameter LLM efficiently on consumer hardware via quantization. It is built as an open-source project for developers. Ornith 1.0 9B is open source under the Open Source license. It runs on the web, the command line, and API.
Behind Ornith 1.0 9B is cyankiwi, and the product first shipped in 2024. It operates in a well-populated space: PulseGate tracks 8 similar tools. Key capabilities include Text Generation, Quantized Model, and Vision Support.
Latest indexed changes and source events
cyankiwi/Ornith-1.0-9B-AWQ-INT4 verified by the PulseGate indexer
Other apps tracked under the same category.
