Modelpublic

FriskyFennec/GigaChat3-10B-A1.8B-base-GGUF

sourceHugging Facemitupdated 17d agoView on Hugging Face

0likes1.1kdownloads

Model Card

GigaChat3-10B-A1.8B-base

A GGUF variant of ai-sage's GigaChat3-10B-A1.8B-base model, suitable for base/foundational model experiments and text completion use cases.

The model was converted & quantized using llama.cpp's build 9370.

Examples

Running via llamacpp

Model's tested with,

llama-server.exe --model GigaChat3-10B-A1.8B-base-BF16.gguf --ctx-size 8192 --gpu-layers 256 --main-gpu 0 --flash-attn auto --mlock --timeout 600 --host 127.0.0.1 --port 8080

Output

Samplers:

—Temperature: 0.7
—Top P: 0.95
—Top K: 40
—Repetition Penalty: 1.1
—Repetition Penalty Range: 256

Input,

Once upon a time,

Output 1,

 there was a little boy named Timmy who loved to play outside. One day, while playing in his backyard, he stumbled upon a magical key hidden under a bush.

The next morning, when Timmy went to school, he couldn't stop thinking

Output 2,

 there were two men who made their living fishing on the ocean. One man was very hardworking and devoted to his job; he would wake up at 4:00 AM every day to go out on the water, and when he got back

Output 3,

 in a quiet village nestled between rolling hills and whispering woods, there lived an old watchmaker named Elias. His workshop was a haven of ticking clocks, polished brass, and the faint scent of oil that lingered in the air like

License

The model inherits the original MIT license from ai-sage/GigaChat3-10B-A1.8B-base.