legraphista's picture
Upload README.md with huggingface_hub
8a31f25 verified
metadata
base_model: mistralai/Mistral-Large-Instruct-2407
extra_gated_description: >-
  If you want to learn more about how we process your personal data, please read
  our <a href="https://mistral.ai/terms/">Privacy Policy</a>.
inference: false
language:
  - en
  - fr
  - de
  - es
  - it
  - pt
  - zh
  - ja
  - ru
  - ko
library_name: gguf
license: other
license_link: https://mistral.ai/licenses/MRL-0.1.md
license_name: mrl
pipeline_tag: text-generation
quantized_by: legraphista
tags:
  - quantized
  - GGUF
  - quantization
  - imat
  - imatrix
  - static
  - 16bit
  - 8bit
  - 6bit
  - 5bit
  - 4bit
  - 3bit
  - 2bit
  - 1bit

Mistral-Large-Instruct-2407-IMat-GGUF

Llama.cpp imatrix quantization of mistralai/Mistral-Large-Instruct-2407

Original Model: mistralai/Mistral-Large-Instruct-2407
Original dtype: BF16 (bfloat16)
Quantized by: llama.cpp b3452
IMatrix dataset: here


Files

IMatrix

Status: βœ… Available
Link: here

Common Quants

Filename Quant type File Size Status Uses IMatrix Is Split
Mistral-Large-Instruct-2407.Q8_0/* Q8_0 130.28GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q6_K/* Q6_K 100.59GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q4_K/* Q4_K 73.22GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q3_K/* Q3_K 59.10GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q2_K/* Q2_K 45.20GB βœ… Available 🟒 IMatrix βœ‚ Yes

All Quants

Filename Quant type File Size Status Uses IMatrix Is Split
Mistral-Large-Instruct-2407.BF16/* BF16 245.23GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.FP16/* F16 245.23GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q8_0/* Q8_0 130.28GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q6_K/* Q6_K 100.59GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q5_K/* Q5_K 86.49GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q5_K_S/* Q5_K_S 84.36GB βœ… Available βšͺ Static βœ‚ Yes
Mistral-Large-Instruct-2407.Q4_K/* Q4_K 73.22GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q4_K_S/* Q4_K_S 69.57GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ4_NL/* IQ4_NL 69.22GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ4_XS/* IQ4_XS 65.43GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q3_K/* Q3_K 59.10GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q3_K_L/* Q3_K_L 64.55GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q3_K_S/* Q3_K_S 52.85GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ3_M/* IQ3_M 55.28GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ3_S/* IQ3_S 53.00GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ3_XS/* IQ3_XS 50.14GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.IQ3_XXS/* IQ3_XXS 47.01GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q2_K/* Q2_K 45.20GB βœ… Available 🟒 IMatrix βœ‚ Yes
Mistral-Large-Instruct-2407.Q2_K_S.gguf Q2_K_S 41.61GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ2_M.gguf IQ2_M 41.62GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ2_S.gguf IQ2_S 38.38GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ2_XS.gguf IQ2_XS 36.08GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ2_XXS.gguf IQ2_XXS 32.43GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ1_M.gguf IQ1_M 28.39GB βœ… Available 🟒 IMatrix πŸ“¦ No
Mistral-Large-Instruct-2407.IQ1_S.gguf IQ1_S 25.96GB βœ… Available 🟒 IMatrix πŸ“¦ No

Downloading using huggingface-cli

If you do not have hugginface-cli installed:

pip install -U "huggingface_hub[cli]"

Download the specific file you want:

huggingface-cli download legraphista/Mistral-Large-Instruct-2407-IMat-GGUF --include "Mistral-Large-Instruct-2407.Q8_0.gguf" --local-dir ./

If the model file is big, it has been split into multiple files. In order to download them all to a local folder, run:

huggingface-cli download legraphista/Mistral-Large-Instruct-2407-IMat-GGUF --include "Mistral-Large-Instruct-2407.Q8_0/*" --local-dir ./
# see FAQ for merging GGUF's

Inference

Llama.cpp

llama.cpp/main -m Mistral-Large-Instruct-2407.Q8_0.gguf --color -i -p "prompt here"

FAQ

Why is the IMatrix not applied everywhere?

According to this investigation, it appears that lower quantizations are the only ones that benefit from the imatrix input (as per hellaswag results).

How do I merge a split GGUF?

  1. Make sure you have gguf-split available
  2. Locate your GGUF chunks folder (ex: Mistral-Large-Instruct-2407.Q8_0)
  3. Run gguf-split --merge Mistral-Large-Instruct-2407.Q8_0/Mistral-Large-Instruct-2407.Q8_0-00001-of-XXXXX.gguf Mistral-Large-Instruct-2407.Q8_0.gguf
    • Make sure to point gguf-split to the first chunk of the split.

Got a suggestion? Ping me @legraphista!