bartowski/vectionlabs_Salience-27B-R5-GGUF

Llamacpp imatrix Quantizations of Salience-27B-R5 by vectionlabs

Using <a href="https://github.com/ggml-org/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggml-org/llama.cpp/releases/tag/b10419">b10419</a> for quantization.

Original model: https://huggingface.co/vectionlabs/Salience-27B-R5

Model details:

• —Parameter count: 28B
• —Input support: text, image (with mmproj file) - details
• —Speculative decoding: yes (MTP) - details
• —imatrix: yes - details
• —Perplexity/KLD measured: no

How to run

Prompt format

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
<think>

Don't know which to choose? Grab Q4_K_M (17.77GB) - usually a good mix of size and performance. Download instructions available here

Available files:

Download a specific file:

hf download bartowski/vectionlabs_Salience-27B-R5-GGUF --include "vectionlabs_Salience-27B-R5-Q4_K_M.gguf" --local-dir ./

Downloading using the Hugging Face CLI

<details> <summary>Click to view download instructions</summary>

First, make sure you have the Hugging Face CLI installed:

pip install -U "huggingface_hub[cli]"

Download a specific file:

hf download bartowski/vectionlabs_Salience-27B-R5-GGUF --include "vectionlabs_Salience-27B-R5-Q4_K_M.gguf" --local-dir ./

The files marked true in the Split column above are stored as multiple parts in a folder. To download all the parts to a local folder, run:

hf download bartowski/vectionlabs_Salience-27B-R5-GGUF --include "vectionlabs_Salience-27B-R5-bf16/*" --local-dir ./

You can either specify a new local-dir (vectionlabs_Salience-27B-R5-bf16) or download them all in place (./)

</details>

How to run

These quants run with llama.cpp - installable in one line via llama.app:

curl -LsSf https://llama.app/install.sh | sh
llama-server -hf bartowski/vectionlabs_Salience-27B-R5-GGUF:Q4_K_M

llama-server includes a built-in chat web UI, served at http://localhost:8080 by default.

These quants were made with llama.cpp release b10419 - if this model's architecture is newly supported, you'll need that release or newer to run them.

They also work in: LM Studio · koboldcpp · ramalama · Jan AI · Text Generation Web UI · LoLLMs · Atomic Chat

Multimodal

This model supports image input. Alongside the quants, this repo includes the multimodal projector files mmproj-vectionlabs_Salience-27B-R5-f16.gguf and mmproj-vectionlabs_Salience-27B-R5-bf16.gguf, which pair with any quant above.

llama.cpp downloads the mmproj automatically when using -hf as shown above; if you're loading files manually, pass it with --mmproj.

MTP

This model has MTP (Multi-Token Prediction) layers, and they are included in these quants

MTP layers act as a built-in draft model, letting llama.cpp run speculative decoding for faster generation. To use them, add the following flag to your llama.cpp command:

--spec-type draft-mtp

Note: the MTP layers are stored at Q40 in the imatrix quants (except for the Q80 quant), since imatrix calibration does not exercise them. Q4_0 is chosen for its speed which massively benefits MTP performance.

imatrix

All quants made using imatrix option, with a calibration corpus rendered through this model's own chat template. The corpus pairs plain prose with tool-calling and reasoning conversations (corpus source data), encoded exactly as this model sees them at inference and processed with --parse-special, so chat-format special tokens contribute to the importance matrix. The corpus rendered for this model is included in this repo: vectionlabs_Salience-27B-R5-calibration-v6.txt. The imatrix is available here: vectionlabs_Salience-27B-R5-imatrix.gguf.

<details> <summary>Calibration render details</summary>

{
  "generator": "auto_quant_v2 calibration renderer",
  "recipe": "calibration-v6",
  "model": "Salience-27B-R5",
  "encoder": "chat_template",
  "chunk_size": 512,
  "prose_chunks": 214,
  "tool_chunks": 410,
  "total_chunks": 624,
  "tool_chunk_fraction": 0.657,
  "n_conversations": 137,
  "extension_convs_used": 0,
  "conversation_token_lengths": [
    758,
    1821,
    1451,
    1713,
    1339,
    1558,
    3374,
    1020,
    1421,
    1616,
    1261,
    2290,
    1075,
    1447,
    2983,
    1467,
    1314,
    1180,
    931,
    912,
    1556,
    1252,
    1601,
    1396,
    2066,
    1692,
    1848,
    1102,
    1609,
    1840,
    1763,
    1414,
    1462,
    1231,
    1214,
    1906,
    1845,
    1369,
    671,
    2104,
    1605,
    1320,
    1592,
    2196,
    2284,
    1509,
    1803,
    1101,
    3150,
    1296,
    3062,
    981,
    1218,
    1164,
    1101,
    892,
    2684,
    1109,
    1342,
    1281,
    1425,
    1365,
    1120,
    1413,
    1377,
    1782,
    1107,
    1746,
    2320,
    1035,
    519,
    1316,
    3529,
    3030,
    903,
    1136,
    1226,
    1267,
    1492,
    1276,
    1345,
    989,
    1404,
    1265,
    1485,
    1740,
    1597,
    2260,
    1067,
    852,
    2960,
    836,
    1597,
    1869,
    2141,
    1397,
    837,
    1550,
    1322,
    1923,
    2038,
    1909,
    1011,
    1209,
    1231,
    3050,
    934,
    928,
    957,
    1606,
    1246,
    1771,
    962,
    546,
    515,
    2820,
    1212,
    1333,
    2024,
    2233,
    2889,
    2871,
    1010,
    1174,
    1034,
    1153,
    1433,
    1193,
    1036,
    1551,
    1024,
    904,
    1937,
    1221,
    1111,
    1510,
    1673
  ],
  "warnings": []
}

</details>

Embed/output weights

Some of these quants (Q3KXL, Q4KL etc) are the standard quantization method with the embeddings and output weights quantized to Q8_0 instead of what they would normally default to.

ARM/AVX information

llama.cpp automatically "repacks" weights into an interleaved layout at load time for faster inference on ARM and AVX machines - details in this PR. This once required downloading special Q4044/48/88 files; those are long gone. Online repacking now covers Q40, IQ4_NL, and most K-quants, so no special quant choice is needed for CPU inference.

Which file should I choose?

<details> <summary>Click here for details</summary>

An older (early 2024) but still useful write-up with charts comparing quant performances is provided by Artefact2 here

The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.

If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.

If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.

Hugging Face can also do this math for you: add your hardware in your Local Apps settings and the model page will show which files fit.

Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.

If you don't want to think too much, grab one of the K-quants. These are in format 'QXKX', like Q5KM.

If you want to get more into the weeds, you can check out this extremely useful feature chart:

llama.cpp feature matrix

But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQXX, like IQ3M. These are newer and offer better performance for their size.

These I-quants can also be used on CPU, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.

</details>

Credits

Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset.

Thank you ZeroWw for the inspiration to experiment with embed/output.

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski