Hugging Face's logo

InferenceIllusionist
/
mini-magnum-12b-v1.1-iMat-GGUF

Transformers
GGUF
iMat
Mistral
Inference Endpoints
conversational
Model card Files Community
1
mini-magnum-12b-v1.1-iMat-GGUF / README.md
InferenceIllusionist's picture
InferenceIllusionist
Update README.md
ff25df5 verified
preview code
|
raw
history blame
1.77 kB
metadata
base_model: intervitens/mini-magnum-12b-v1.1
library_name: transformers
quantized_by: InferenceIllusionist
tags:
  - iMat
  - gguf
  - Mistral
license: apache-2.0

mini-magnum-12b-v1.1-iMat-GGUF

Important Note: Inferencing in llama.cpp has now been merged in PR #8604. Please ensure you are on release b3438 or newer. Text-generation-web-ui (Ooba) is also working as of 7/23. Official support for Kobold.cpp is still pending.

Quantized from mini-magnum-12b-v1.1 fp16

  • Weighted quantizations were creating using fp16 GGUF and groups_merged.txt in 92 chunks and n_ctx=512
  • Static fp16 will also be included in repo
  • For a brief rundown of iMatrix quant performance please see this PR
  • All quants are verified working prior to uploading to repo for your safety and convenience

KL-Divergence Reference Chart (Click on image to view in full size)

Quant-specific Tips:

  • If you are getting a cudaMalloc failed: out of memory error, try passing an argument for lower context in llama.cpp, e.g. for 8k: -c 8192
  • If you have all ampere generation or newer cards, you can use flash attention like so: -fa
  • Provided Flash Attention is enabled you can also use quantized cache to save on VRAM e.g. for 8-bit: -ctk q8_0 -ctv q8_0

Original model card can be found here