models(gallery): add gemmasutra-mini (#3120)

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-08-04 15:17:24 +02:00 committed by GitHub
parent 12d6d2d177
commit 1788fc8d4a
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -1363,6 +1363,20 @@
- filename: Gemmasutra-Pro-27B-v1.Q4_K_M.gguf
sha256: 336a2fbf142849fcc20e432123433807b6c7b09988652ef583a63636a0f90218
uri: huggingface://mradermacher/Gemmasutra-Pro-27B-v1-GGUF/Gemmasutra-Pro-27B-v1.Q4_K_M.gguf
- !!merge <<: *gemma
name: "gemmasutra-mini-2b-v1"
icon: https://cdn-uploads.huggingface.co/production/uploads/65f2fd1c25b848bd061b5c2e/w0Oi8TReoQNT3ljm5Wf6c.webp
urls:
- https://huggingface.co/TheDrummer/Gemmasutra-Mini-2B-v1-GGUF
description: |
It is a small, 2 billion parameter language model that has been trained for role-playing purposes. The model is designed to work well in various settings, such as in the browser, on a laptop, or even on a Raspberry Pi. It has been fine-tuned for RP use and claims to provide a satisfying experience, even in low-resource environments. The model is uncensored and unaligned, and it can be used with the Gemma Instruct template or with chat completion. For the best experience, it is recommended to modify the template to support the `system` role. The model also features examples of its output, highlighting its versatility and creativity.
overrides:
parameters:
model: Gemmasutra-Mini-2B-v1i-Q4_K_M.gguf
files:
- filename: Gemmasutra-Mini-2B-v1i-Q4_K_M.gguf
sha256: 29ba3db911fbadef4452ba757ddd9ce58fb892b7a872f19eefd0743c961797fb
uri: huggingface://TheDrummer/Gemmasutra-Mini-2B-v1-GGUF/Gemmasutra-Mini-2B-v1i-Q4_K_M.gguf
- !!merge <<: *gemma
name: "tarnished-9b-i1"
icon: https://huggingface.co/lodrick-the-lafted/tarnished-9b/resolve/main/nox.jpg