Instructions to use bartowski/cosmicoptima_computer-10-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use bartowski/cosmicoptima_computer-10-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Use Docker

docker model run hf.co/bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use bartowski/cosmicoptima_computer-10-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bartowski/cosmicoptima_computer-10-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bartowski/cosmicoptima_computer-10-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Ollama
How to use bartowski/cosmicoptima_computer-10-GGUF with Ollama:
```
ollama run hf.co/bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use bartowski/cosmicoptima_computer-10-GGUF with Docker Model Runner:
```
docker model run hf.co/bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M
```

Lemonade

How to use bartowski/cosmicoptima_computer-10-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull bartowski/cosmicoptima_computer-10-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.cosmicoptima_computer-10-GGUF-Q4_K_M

List all available models

lemonade list

Atomic Chat

Llamacpp imatrix Quantizations of computer-10 by cosmicoptima

Using llama.cpp release b11184 for quantization.

Original model: https://huggingface.co/cosmicoptima/computer-10

Model details:

Parameter count: 71B
Input support: text
Speculative decoding: no
imatrix: yes - details

How to run

Prompt format

<|begin_of_text|>{system_prompt}

Full conversation with Model C:

**User:** {prompt}

**Model C:**

Don't know which to choose? Grab Q4_K_M (43.97GB) - usually a good mix of size and performance. Download instructions available here

Available files:

Filename	Quant type	File Size	Split	Description
cosmicoptima_computer-10-bf16.gguf	bf16	141.12GB	true	Full BF16 weights.
cosmicoptima_computer-10-Q8_0.gguf	Q8_0	74.98GB	true	Extremely high quality, generally unneeded but max available quant.
cosmicoptima_computer-10-Q6_K_L.gguf	Q6_K_L	60.59GB	true	Uses Q8_0 for embed and output weights. Very high quality, near perfect, recommended.
cosmicoptima_computer-10-Q6_K.gguf	Q6_K	60.08GB	true	Very high quality, near perfect, recommended.
cosmicoptima_computer-10-Q5_K_M.gguf	Q5_K_M	51.11GB	false	High quality, recommended.
cosmicoptima_computer-10-Q5_K_S.gguf	Q5_K_S	49.01GB	false	High quality, recommended.
cosmicoptima_computer-10-Q4_1.gguf	Q4_1	44.84GB	false	Legacy format, similar performance to Q4_K_S but with improved tokens/watt on Apple silicon.
cosmicoptima_computer-10-Q4_K_L.gguf	Q4_K_L	44.75GB	false	Uses Q8_0 for embed and output weights. Good quality, recommended.
cosmicoptima_computer-10-Q4_K_M.gguf	Q4_K_M	43.97GB	false	Good quality, default size for most use cases, recommended.
cosmicoptima_computer-10-Q4_K_S.gguf	Q4_K_S	41.04GB	false	Slightly lower quality with more space savings, recommended.
cosmicoptima_computer-10-IQ4_NL.gguf	IQ4_NL	40.83GB	false	Similar to IQ4_XS, but slightly larger.
cosmicoptima_computer-10-Q4_0.gguf	Q4_0	40.81GB	false	Legacy format, kept for compatibility with older tools.
cosmicoptima_computer-10-IQ4_XS.gguf	IQ4_XS	38.77GB	false	Decent quality, smaller than Q4_K_S with similar performance, recommended.
cosmicoptima_computer-10-Q3_K_L.gguf	Q3_K_L	37.52GB	false	Lower quality but usable, good for low RAM availability.
cosmicoptima_computer-10-Q3_K_M.gguf	Q3_K_M	35.32GB	false	Low quality.
cosmicoptima_computer-10-IQ3_M.gguf	IQ3_M	32.99GB	false	Medium-low quality, new method with decent performance comparable to Q3_K_M.
cosmicoptima_computer-10-Q3_K_S.gguf	Q3_K_S	31.96GB	false	Low quality, not recommended.
cosmicoptima_computer-10-IQ3_XS.gguf	IQ3_XS	30.64GB	false	Lower quality, new method with decent performance, slightly better than Q3_K_S.
cosmicoptima_computer-10-IQ3_XXS.gguf	IQ3_XXS	29.02GB	false	Lower quality, new method with decent performance, comparable to Q3 quants.
cosmicoptima_computer-10-Q2_K.gguf	Q2_K	27.70GB	false	Very low quality but surprisingly usable.
cosmicoptima_computer-10-IQ2_M.gguf	IQ2_M	26.95GB	false	Relatively low quality, uses SOTA techniques to be surprisingly usable.
cosmicoptima_computer-10-IQ2_S.gguf	IQ2_S	25.41GB	false	Low quality, uses SOTA techniques to be usable.
cosmicoptima_computer-10-IQ2_XS.gguf	IQ2_XS	24.19GB	false	Low quality, uses SOTA techniques to be usable.
cosmicoptima_computer-10-IQ2_XXS.gguf	IQ2_XXS	22.48GB	false	Very low quality, uses SOTA techniques to be usable.
cosmicoptima_computer-10-IQ1_M.gguf	IQ1_M	20.56GB	false	Extremely low quality, not recommended.
cosmicoptima_computer-10-IQ1_S.gguf	IQ1_S	19.40GB	false	Extremely low quality, not recommended.

Download a specific file:

hf download bartowski/cosmicoptima_computer-10-GGUF --include "cosmicoptima_computer-10-Q4_K_M.gguf" --local-dir ./

Downloading using the Hugging Face CLI

Click to view download instructions

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

pip install -U "huggingface_hub[cli]"

Download a specific file:

hf download bartowski/cosmicoptima_computer-10-GGUF --include "cosmicoptima_computer-10-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/cosmicoptima_computer-10-GGUF --include "cosmicoptima_computer-10-Q8_0/*" --local-dir ./

You can either specify a new local-dir (cosmicoptima_computer-10-Q8_0) or download them all in place (./)

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/cosmicoptima_computer-10-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 b11184 - 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

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: cosmicoptima_computer-10-calibration-v6.txt. The imatrix is available here: cosmicoptima_computer-10-imatrix.gguf.

Calibration render details

{
  "generator": "auto_quant_v2 calibration renderer",
  "recipe": "calibration-v6",
  "model": "computer-10",
  "encoder": "chat_template",
  "library_versions": {
    "transformers": "5.9.0",
    "tokenizers": "0.22.2",
    "tiktoken": "0.14.0",
    "blobfile": "3.3.0",
    "huggingface_hub": "1.31.0"
  },
  "chunk_size": 512,
  "prose_chunks": 216,
  "tool_chunks": 80,
  "total_chunks": 296,
  "tool_chunk_fraction": 0.27,
  "n_conversations": 173,
  "extension_convs_used": 36,
  "conversation_token_lengths": [
    85,
    384,
    131,
    242,
    234,
    102,
    327,
    565,
    410,
    101,
    68,
    187,
    38,
    89,
    279,
    173,
    568,
    254,
    340,
    107,
    104,
    174,
    155,
    392,
    172,
    118,
    373,
    166,
    119,
    187,
    294,
    152,
    170,
    108,
    185,
    291,
    271,
    100,
    69,
    109,
    440,
    340,
    464,
    181,
    105,
    295,
    352,
    210,
    285,
    308,
    210,
    210,
    256,
    148,
    101,
    812,
    557,
    255,
    345,
    220,
    274,
    97,
    315,
    142,
    574,
    100,
    107,
    163,
    170,
    100,
    134,
    327,
    199,
    140,
    173,
    180,
    61,
    179,
    99,
    97,
    346,
    98,
    123,
    145,
    335,
    105,
    226,
    800,
    125,
    196,
    102,
    258,
    154,
    141,
    169,
    229,
    43,
    303,
    105,
    66,
    205,
    285,
    105,
    235,
    442,
    196,
    216,
    88,
    124,
    180,
    631,
    574,
    339,
    140,
    346,
    180,
    60,
    558,
    456,
    284,
    49,
    374,
    1072,
    58,
    404,
    137,
    134,
    210,
    186,
    87,
    233,
    553,
    132,
    663,
    91,
    197,
    190,
    232,
    97,
    311,
    104,
    107,
    162,
    125,
    136,
    46,
    473,
    230,
    148,
    245,
    175,
    1082,
    280,
    508,
    549,
    105,
    48,
    172,
    139,
    175,
    130,
    170,
    69,
    399,
    120,
    235,
    93,
    120,
    317,
    42,
    87,
    290,
    278
  ],
  "warnings": [
    "tool chunk fraction 0.270 outside the validated 0.55-0.68 band even after exhausting the extension pool \u2014 coverage screened clean under the band on the strictest measured MoE (Qwen3-Next-80B, 0/24576 dead experts), but the validated quality margin is not guaranteed at this mix; on a new MoE family, check the imatrix for uncovered experts.",
    "band-holding added 36 of 36 extension conversations: the render is coverage-screened but the extension mechanism is not itself det-validated (the base 137-conv selection is)"
  ]
}

Embed/output weights

Some of these quants (Q3_K_XL, Q4_K_L 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 Q4_0_4_4/4_8/8_8 files; those are long gone. Online repacking now covers Q4_0, IQ4_NL, and most K-quants, so no special quant choice is needed for CPU inference.

Which file should I choose?

Click here for details

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 'QX_K_X', like Q5_K_M.

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 IQX_X, like IQ3_M. 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.