Hugging Face's logo

legraphista
/
Codestral-22B-v0.1-IMat-GGUF

Text Generation
GGUF
code
quantized
GGUF
imatrix
quantization
imat
static
Model card Files Community
2
Codestral-22B-v0.1-IMat-GGUF / README.md
legraphista's picture
legraphista
Upload README.md with huggingface_hub
d15c4b8 verified
preview code
|
raw
history blame
9.87 kB
metadata
base_model: bullerwins/Codestral-22B-v0.1-hf
inference: false
language:
  - code
library_name: gguf
license: other
license_link: https://mistral.ai/licences/MNPL-0.1.md
license_name: mnpl
pipeline_tag: text-generation
quantized_by: legraphista
tags:
  - code
  - quantized
  - GGUF
  - imatrix
  - quantization
  - imat
  - imatrix
  - static

Codestral-22B-v0.1-hf-IMat-GGUF

Llama.cpp imatrix quantization of bullerwins/Codestral-22B-v0.1-hf

Original Model: bullerwins/Codestral-22B-v0.1-hf
Original dtype: BF16 (bfloat16)
Quantized by: llama.cpp b3037
IMatrix dataset: here

Files

IMatrix

Status: ✅ Available
Link: here

Common Quants

Filename Quant type File Size Status Uses IMatrix Is Split
Codestral-22B-v0.1-hf.Q8_0.gguf Q8_0 23.64GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q6_K.gguf Q6_K 18.25GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q4_K.gguf Q4_K 13.34GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q3_K.gguf Q3_K 10.76GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q2_K.gguf Q2_K 8.27GB ✅ Available 🟢 IMatrix 📦 No

All Quants

Filename Quant type File Size Status Uses IMatrix Is Split
Codestral-22B-v0.1-hf.BF16.gguf BF16 44.50GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.FP16.gguf F16 44.50GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q8_0.gguf Q8_0 23.64GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q6_K.gguf Q6_K 18.25GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q5_K.gguf Q5_K 15.72GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q5_K_S.gguf Q5_K_S 15.32GB ✅ Available ⚪ Static 📦 No
Codestral-22B-v0.1-hf.Q4_K.gguf Q4_K 13.34GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q4_K_S.gguf Q4_K_S 12.66GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ4_NL.gguf IQ4_NL 12.61GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ4_XS.gguf IQ4_XS 11.94GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q3_K.gguf Q3_K 10.76GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q3_K_L.gguf Q3_K_L 11.73GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q3_K_S.gguf Q3_K_S 9.64GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ3_M.gguf IQ3_M 10.06GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ3_S.gguf IQ3_S 9.69GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ3_XS IQ3_XS - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ3_XXS IQ3_XXS - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.Q2_K.gguf Q2_K 8.27GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.Q2_K_S.gguf Q2_K_S 7.65GB ✅ Available 🟢 IMatrix 📦 No
Codestral-22B-v0.1-hf.IQ2_M IQ2_M - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ2_S IQ2_S - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ2_XS IQ2_XS - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ2_XXS IQ2_XXS - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ1_M IQ1_M - ⏳ Processing 🟢 IMatrix -
Codestral-22B-v0.1-hf.IQ1_S IQ1_S - ⏳ Processing 🟢 IMatrix -

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/Codestral-22B-v0.1-hf-IMat-GGUF --include "Codestral-22B-v0.1-hf.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/Codestral-22B-v0.1-hf-IMat-GGUF --include "Codestral-22B-v0.1-hf.Q8_0/*" --local-dir ./
# see FAQ for merging GGUF's

Inference

Simple chat template 

<s>[INST] Can you provide ways to eat combinations of bananas and dragonfruits? [/INST] Sure! Here are some ways to eat bananas and dragonfruits together:
 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey.
 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey. </s><s>[INST] What about solving an 2x + 3 = 7 equation? [/INST]

Chat template with system prompt 

<s>[INST] <<SYS>>
You are a helpful AI.
<</SYS>>

Can you provide ways to eat combinations of bananas and dragonfruits? [/INST] Sure! Here are some ways to eat bananas and dragonfruits together:
 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey.
 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey. </s><s>[INST] What about solving an 2x + 3 = 7 equation? [/INST]

Llama.cpp 

llama.cpp/main -m Codestral-22B-v0.1-hf.Q8_0.gguf --color -i -p "prompt here (according to the chat template)"

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: Codestral-22B-v0.1-hf.Q8_0)
  3. Run gguf-split --merge Codestral-22B-v0.1-hf.Q8_0/Codestral-22B-v0.1-hf.Q8_0-00001-of-XXXXX.gguf Codestral-22B-v0.1-hf.Q8_0.gguf
    • Make sure to point gguf-split to the first chunk of the split.

Got a suggestion? Ping me @legraphista!