dolphin-2.9.1-mixtral-1x22b-IMat-GGUF

Llama.cpp imatrix quantization of cognitivecomputations/dolphin-2.9.1-mixtral-1x22b

Original Model: cognitivecomputations/dolphin-2.9.1-mixtral-1x22b
Original dtype: BF16 (bfloat16)
Quantized by: llama.cpp b3024
IMatrix dataset: here


Files

IMatrix

Status: βœ… Available
Link: here

Common Quants

Filename Quant type File Size Status Uses IMatrix Is Split
dolphin-2.9.1-mixtral-1x22b.Q8_0.gguf Q8_0 23.63GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q6_K.gguf Q6_K 18.25GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q4_K.gguf Q4_K 13.34GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q3_K.gguf Q3_K 10.75GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q2_K.gguf Q2_K 8.27GB βœ… Available 🟒 IMatrix πŸ“¦ No

All Quants

Filename Quant type File Size Status Uses IMatrix Is Split
dolphin-2.9.1-mixtral-1x22b.BF16.gguf BF16 44.48GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.FP16.gguf F16 44.48GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q8_0.gguf Q8_0 23.63GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q6_K.gguf Q6_K 18.25GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q5_K.gguf Q5_K 15.72GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q5_K_S.gguf Q5_K_S 15.32GB βœ… Available βšͺ Static πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q4_K.gguf Q4_K 13.34GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q4_K_S.gguf Q4_K_S 12.66GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ4_NL.gguf IQ4_NL 12.61GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ4_XS.gguf IQ4_XS 11.93GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q3_K.gguf Q3_K 10.75GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q3_K_L.gguf Q3_K_L 11.73GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q3_K_S.gguf Q3_K_S 9.64GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ3_M.gguf IQ3_M 10.06GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ3_S.gguf IQ3_S 9.68GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ3_XS.gguf IQ3_XS 9.17GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ3_XXS.gguf IQ3_XXS 8.59GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q2_K.gguf Q2_K 8.27GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.Q2_K_S.gguf Q2_K_S 7.65GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ2_M.gguf IQ2_M 7.62GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ2_S.gguf IQ2_S 7.03GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ2_XS.gguf IQ2_XS 6.64GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ2_XXS.gguf IQ2_XXS 5.99GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ1_M.gguf IQ1_M 5.26GB βœ… Available 🟒 IMatrix πŸ“¦ No
dolphin-2.9.1-mixtral-1x22b.IQ1_S.gguf IQ1_S 4.83GB 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/dolphin-2.9.1-mixtral-1x22b-IMat-GGUF --include "dolphin-2.9.1-mixtral-1x22b.BF16.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/dolphin-2.9.1-mixtral-1x22b-IMat-GGUF --include "dolphin-2.9.1-mixtral-1x22b.BF16/*" --local-dir ./
# see FAQ for merging GGUF's

Inference

Simple chat template

<|im_start|>user
Can you provide ways to eat combinations of bananas and dragonfruits?<|im_end|>
<|im_start|>assistant
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.<|im_end|>
<|im_start|>user
What about solving an 2x + 3 = 7 equation?<|im_end|>

Chat template with system prompt

<|im_start|>system
You are a helpful AI.<|im_end|>
<|im_start|>user
Can you provide ways to eat combinations of bananas and dragonfruits?<|im_end|>
<|im_start|>assistant
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.<|im_end|>
<|im_start|>user
What about solving an 2x + 3 = 7 equation?<|im_end|>

Llama.cpp

llama.cpp/main -m dolphin-2.9.1-mixtral-1x22b.BF16.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: dolphin-2.9.1-mixtral-1x22b.BF16)
  3. Run gguf-split --merge dolphin-2.9.1-mixtral-1x22b.BF16/dolphin-2.9.1-mixtral-1x22b.BF16-00001-of-XXXXX.gguf dolphin-2.9.1-mixtral-1x22b.BF16.gguf
    • Make sure to point gguf-split to the first chunk of the split.

Got a suggestion? Ping me @legraphista!

Downloads last month
306
GGUF
Model size
22.2B params
Architecture
llama

1-bit

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Inference Examples
Inference API (serverless) has been turned off for this model.

Model tree for legraphista/dolphin-2.9.1-mixtral-1x22b-IMat-GGUF

Datasets used to train legraphista/dolphin-2.9.1-mixtral-1x22b-IMat-GGUF