Instructions to use guell00/VELUM-Coder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use guell00/VELUM-Coder 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 guell00/VELUM-Coder:Q4_K_M # Run inference directly in the terminal: llama cli -hf guell00/VELUM-Coder:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf guell00/VELUM-Coder:Q4_K_M # Run inference directly in the terminal: llama cli -hf guell00/VELUM-Coder: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 guell00/VELUM-Coder:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf guell00/VELUM-Coder: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 guell00/VELUM-Coder:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf guell00/VELUM-Coder:Q4_K_M
Use Docker
docker model run hf.co/guell00/VELUM-Coder:Q4_K_M
- LM Studio
- Jan
- Ollama
How to use guell00/VELUM-Coder with Ollama:
ollama run hf.co/guell00/VELUM-Coder:Q4_K_M
- Unsloth Studio
How to use guell00/VELUM-Coder with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for guell00/VELUM-Coder to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for guell00/VELUM-Coder to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for guell00/VELUM-Coder to start chatting
- Pi
How to use guell00/VELUM-Coder with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf guell00/VELUM-Coder:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "guell00/VELUM-Coder:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use guell00/VELUM-Coder with Docker Model Runner:
docker model run hf.co/guell00/VELUM-Coder:Q4_K_M
- Lemonade
How to use guell00/VELUM-Coder with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull guell00/VELUM-Coder:Q4_K_M
Run and chat with the model
lemonade run user.VELUM-Coder-Q4_K_M
List all available models
lemonade list
- Hermes Agent
How to use guell00/VELUM-Coder with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf guell00/VELUM-Coder:Q4_K_M
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default guell00/VELUM-Coder:Q4_K_M
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use guell00/VELUM-Coder with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf guell00/VELUM-Coder:Q4_K_M
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "guell00/VELUM-Coder:Q4_K_M" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Update README.md
Browse files
README.md
CHANGED
|
@@ -3,8 +3,10 @@ license: mit
|
|
| 3 |
language:
|
| 4 |
- pt
|
| 5 |
- en
|
|
|
|
|
|
|
| 6 |
base_model:
|
| 7 |
-
-
|
| 8 |
tags:
|
| 9 |
- coder
|
| 10 |
- code
|
|
@@ -13,146 +15,144 @@ tags:
|
|
| 13 |
- fast
|
| 14 |
- smart
|
| 15 |
---
|
| 16 |
-
|
| 17 |
<div align="center">
|
| 18 |
|
| 19 |
-
|
| 20 |

|
| 21 |
|
| 22 |
<br>
|
| 23 |
|
| 24 |
# LEVUM
|
| 25 |
|
| 26 |
-
[π
|
| 27 |
|
| 28 |
-
### Local Β·
|
| 29 |
|
| 30 |
-
**
|
| 31 |
|
| 32 |
-
|
| 33 |
|
| 34 |
-
**
|
| 35 |
|
| 36 |
</div>
|
| 37 |
|
| 38 |
---
|
| 39 |
|
| 40 |
-
##
|
| 41 |
|
| 42 |
-
**LEVUM**
|
| 43 |
|
| 44 |
-
|
| 45 |
|
| 46 |
-
|
| 47 |
|
| 48 |
-
*
|
| 49 |
-
*
|
| 50 |
* debugging;
|
| 51 |
-
*
|
| 52 |
-
*
|
| 53 |
-
*
|
| 54 |
-
*
|
| 55 |
-
*
|
| 56 |
-
*
|
| 57 |
|
| 58 |
-
LEVUM
|
| 59 |
|
| 60 |
-
> **Software
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
-
##
|
| 65 |
|
| 66 |
-
|
|
| 67 |
-
| ---------------
|
| 68 |
-
| **
|
| 69 |
-
| **Base**
|
| 70 |
-
| **
|
| 71 |
-
| **
|
| 72 |
-
| **
|
| 73 |
-
| **
|
| 74 |
-
| **
|
| 75 |
-
| **
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
-
#
|
| 80 |
|
| 81 |
-
LEVUM
|
| 82 |
|
| 83 |
-
|
| 84 |
|
| 85 |
-
**RAM / VRAM β
|
| 86 |
|
| 87 |
-
|
| 88 |
|
| 89 |

|
| 90 |
|
| 91 |
-
|
| 92 |
|
| 93 |
-
### Q8 β
|
| 94 |
|
| 95 |
-
|
| 96 |
|
| 97 |
`Q8_0`
|
| 98 |
|
| 99 |
---
|
| 100 |
|
| 101 |
-
### Q4 / IQ4 β
|
| 102 |
|
| 103 |
-
|
| 104 |
|
| 105 |
-
|
| 106 |
|
| 107 |
`Q4_K_M` Β· `Q4_K_S` Β· `IQ4_XS` Β· `IQ4_NL` Β· `Q4_1` Β· `Q4_0`
|
| 108 |
|
| 109 |
---
|
| 110 |
|
| 111 |
-
### Q3 / IQ3 β
|
| 112 |
|
| 113 |
-
|
| 114 |
|
| 115 |
`Q3_K_L` Β· `Q3_K_M` Β· `Q3_K_S` Β· `IQ3_M` Β· `IQ3_S` Β· `IQ3_XS` Β· `IQ3_XXS`
|
| 116 |
|
| 117 |
---
|
| 118 |
|
| 119 |
-
### Q2 / IQ2 β
|
| 120 |
|
| 121 |
-
|
| 122 |
|
| 123 |
`Q2_K` Β· `Q2_K_S` Β· `IQ2_M` Β· `IQ2_S` Β· `IQ2_XS` Β· `IQ2_XXS` Β· `Q2_0` Β· `TQ2_0`
|
| 124 |
|
| 125 |
---
|
| 126 |
|
| 127 |
-
### IQ1 β
|
| 128 |
|
| 129 |
-
|
| 130 |
|
| 131 |
`IQ1_M`
|
| 132 |
|
| 133 |
-
|
| 134 |
|
| 135 |
---
|
| 136 |
|
| 137 |
-
> **
|
| 138 |
-
>
|
| 139 |
-
>
|
| 140 |
|
| 141 |
---
|
| 142 |
|
| 143 |
-
#
|
| 144 |
|
| 145 |
## llama.cpp
|
| 146 |
|
| 147 |
-
|
| 148 |
|
| 149 |
```bash
|
| 150 |
llama-cli \
|
| 151 |
-m ./LEVUM-Q4_K_M.gguf \
|
| 152 |
-
-p "
|
| 153 |
```
|
| 154 |
|
| 155 |
-
|
| 156 |
|
| 157 |
```bash
|
| 158 |
llama-server \
|
|
@@ -160,13 +160,13 @@ llama-server \
|
|
| 160 |
-c 8192
|
| 161 |
```
|
| 162 |
|
| 163 |
-
|
| 164 |
|
| 165 |
---
|
| 166 |
|
| 167 |
## Ollama
|
| 168 |
|
| 169 |
-
|
| 170 |
|
| 171 |
```dockerfile
|
| 172 |
FROM ./LEVUM-Q4_K_M.gguf
|
|
@@ -175,68 +175,68 @@ PARAMETER temperature 0.6
|
|
| 175 |
PARAMETER top_p 0.9
|
| 176 |
```
|
| 177 |
|
| 178 |
-
|
| 179 |
|
| 180 |
```bash
|
| 181 |
ollama create levum -f Modelfile
|
| 182 |
ollama run levum
|
| 183 |
```
|
| 184 |
|
| 185 |
-
|
| 186 |
|
| 187 |
```text
|
| 188 |
-
>>>
|
| 189 |
```
|
| 190 |
|
| 191 |
---
|
| 192 |
|
| 193 |
## LM Studio
|
| 194 |
|
| 195 |
-
1.
|
| 196 |
-
2.
|
| 197 |
-
3.
|
| 198 |
-
4.
|
| 199 |
-
5.
|
| 200 |
|
| 201 |
-
|
| 202 |
|
| 203 |
---
|
| 204 |
|
| 205 |
# Prompts
|
| 206 |
|
| 207 |
-
LEVUM
|
| 208 |
|
| 209 |
-
##
|
| 210 |
|
| 211 |
```text
|
| 212 |
-
|
| 213 |
|
| 214 |
-
|
| 215 |
-
- HTML, CSS
|
| 216 |
-
-
|
| 217 |
-
-
|
| 218 |
-
-
|
| 219 |
-
-
|
| 220 |
|
| 221 |
-
|
| 222 |
```
|
| 223 |
|
| 224 |
---
|
| 225 |
|
| 226 |
-
##
|
| 227 |
|
| 228 |
```text
|
| 229 |
-
|
| 230 |
|
| 231 |
-
|
| 232 |
-
- preserve
|
| 233 |
-
-
|
| 234 |
-
-
|
| 235 |
-
-
|
| 236 |
-
-
|
| 237 |
|
| 238 |
-
|
| 239 |
-
[
|
| 240 |
```
|
| 241 |
|
| 242 |
---
|
|
@@ -244,147 +244,147 @@ CΓ³digo:
|
|
| 244 |
## Debug
|
| 245 |
|
| 246 |
```text
|
| 247 |
-
|
| 248 |
|
| 249 |
-
|
| 250 |
-
[
|
| 251 |
|
| 252 |
-
|
| 253 |
-
[
|
| 254 |
|
| 255 |
-
|
| 256 |
-
[
|
| 257 |
|
| 258 |
-
|
| 259 |
-
[
|
| 260 |
|
| 261 |
-
|
| 262 |
```
|
| 263 |
|
| 264 |
---
|
| 265 |
|
| 266 |
-
##
|
| 267 |
|
| 268 |
```text
|
| 269 |
-
|
| 270 |
|
| 271 |
-
|
| 272 |
-
-
|
| 273 |
-
-
|
| 274 |
-
-
|
| 275 |
-
-
|
| 276 |
-
-
|
| 277 |
|
| 278 |
-
|
| 279 |
```
|
| 280 |
|
| 281 |
---
|
| 282 |
|
| 283 |
-
#
|
| 284 |
|
| 285 |
-
|
| 286 |
|
| 287 |
```text
|
| 288 |
-
|
| 289 |
-
|
| 290 |
|
| 291 |
-
|
| 292 |
-
Stack,
|
| 293 |
|
| 294 |
-
|
| 295 |
-
|
| 296 |
|
| 297 |
-
|
| 298 |
-
|
| 299 |
|
| 300 |
-
|
| 301 |
-
|
| 302 |
```
|
| 303 |
|
| 304 |
-
|
| 305 |
|
| 306 |
```text
|
| 307 |
-
|
| 308 |
-
|
| 309 |
|
| 310 |
-
|
| 311 |
-
|
| 312 |
|
| 313 |
-
|
| 314 |
-
-
|
| 315 |
-
-
|
| 316 |
-
-
|
| 317 |
-
-
|
| 318 |
|
| 319 |
-
|
| 320 |
-
-
|
| 321 |
-
-
|
| 322 |
|
| 323 |
-
|
| 324 |
-
|
| 325 |
```
|
| 326 |
|
| 327 |
---
|
| 328 |
|
| 329 |
# Code-first
|
| 330 |
|
| 331 |
-
LEVUM
|
| 332 |
|
| 333 |
```text
|
| 334 |
-
PROMPT /
|
| 335 |
β
|
| 336 |
LEVUM
|
| 337 |
β
|
| 338 |
-
|
| 339 |
β
|
| 340 |
-
|
| 341 |
β
|
| 342 |
DEBUG
|
| 343 |
β
|
| 344 |
-
|
| 345 |
β
|
| 346 |
-
|
| 347 |
```
|
| 348 |
|
| 349 |
-
###
|
| 350 |
|
| 351 |
-
|
| 352 |
|
| 353 |
### Debug
|
| 354 |
|
| 355 |
-
|
| 356 |
|
| 357 |
-
###
|
| 358 |
|
| 359 |
-
|
| 360 |
|
| 361 |
-
###
|
| 362 |
|
| 363 |
-
|
| 364 |
|
| 365 |
---
|
| 366 |
|
| 367 |
-
#
|
| 368 |
|
| 369 |
**Prompt**
|
| 370 |
|
| 371 |
```text
|
| 372 |
-
|
| 373 |
|
| 374 |
-
Use
|
| 375 |
|
| 376 |
-
|
| 377 |
-
-
|
| 378 |
-
-
|
| 379 |
-
-
|
| 380 |
-
-
|
| 381 |
-
-
|
| 382 |
-
-
|
| 383 |
|
| 384 |
-
|
| 385 |
```
|
| 386 |
|
| 387 |
-
**
|
| 388 |
|
| 389 |
```text
|
| 390 |
prompt
|
|
@@ -395,109 +395,109 @@ HTML + CSS + JavaScript
|
|
| 395 |
β
|
| 396 |
browser
|
| 397 |
β
|
| 398 |
-
|
| 399 |
```
|
| 400 |
|
| 401 |
---
|
| 402 |
|
| 403 |
# Hardware
|
| 404 |
|
| 405 |
-
|
| 406 |
|
| 407 |
-
*
|
| 408 |
-
*
|
| 409 |
* runtime;
|
| 410 |
* KV cache;
|
| 411 |
* CPU;
|
| 412 |
* GPU;
|
| 413 |
-
*
|
| 414 |
-
*
|
| 415 |
|
| 416 |
-
|
| 417 |
|
| 418 |
-
|
| 419 |
|
| 420 |
```text
|
| 421 |
-
|
| 422 |
β
|
| 423 |
-
|
| 424 |
β
|
| 425 |
-
|
| 426 |
|
| 427 |
-
|
| 428 |
β
|
| 429 |
-
|
| 430 |
β
|
| 431 |
-
|
| 432 |
```
|
| 433 |
|
| 434 |
-
|
| 435 |
|
| 436 |
---
|
| 437 |
|
| 438 |
-
#
|
| 439 |
|
| 440 |
-
LEVUM
|
| 441 |
|
| 442 |
-
|
| 443 |
|
| 444 |
-
*
|
| 445 |
-
*
|
| 446 |
-
*
|
| 447 |
-
*
|
| 448 |
-
*
|
| 449 |
-
*
|
| 450 |
-
*
|
| 451 |
|
| 452 |
-
|
| 453 |
|
| 454 |
-
|
| 455 |
|
| 456 |
---
|
| 457 |
|
| 458 |
-
#
|
| 459 |
|
| 460 |
-
|
| 461 |
|
| 462 |
-
1.
|
| 463 |
-
2.
|
| 464 |
-
3.
|
| 465 |
-
4.
|
| 466 |
-
5.
|
| 467 |
-
6.
|
| 468 |
-
7. use
|
| 469 |
|
| 470 |
-
|
| 471 |
|
| 472 |
---
|
| 473 |
|
| 474 |
-
#
|
| 475 |
|
| 476 |
-
LEVUM
|
| 477 |
|
| 478 |
**Qwen 3.5 9B**
|
| 479 |
|
| 480 |
-
|
| 481 |
|
| 482 |
-
|
| 483 |
|
| 484 |
---
|
| 485 |
|
| 486 |
-
#
|
| 487 |
|
| 488 |
-
|
| 489 |
|
| 490 |
-
|
| 491 |
|
| 492 |
---
|
| 493 |
|
| 494 |
-
#
|
| 495 |
|
| 496 |
<div align="center">
|
| 497 |
|
| 498 |
-
### π§π·
|
| 499 |
|
| 500 |
-
**
|
| 501 |
|
| 502 |
Local-first Β· Code-first Β· Software-first
|
| 503 |
|
|
@@ -509,4 +509,4 @@ Local-first Β· Code-first Β· Software-first
|
|
| 509 |
|
| 510 |
**LEVUM Β© 2026**
|
| 511 |
|
| 512 |
-
</div>
|
|
|
|
| 3 |
language:
|
| 4 |
- pt
|
| 5 |
- en
|
| 6 |
+
- es
|
| 7 |
+
- fr
|
| 8 |
base_model:
|
| 9 |
+
- ornith-ai/Ornith-1.5-9B
|
| 10 |
tags:
|
| 11 |
- coder
|
| 12 |
- code
|
|
|
|
| 15 |
- fast
|
| 16 |
- smart
|
| 17 |
---
|
|
|
|
| 18 |
<div align="center">
|
| 19 |
|
|
|
|
| 20 |

|
| 21 |
|
| 22 |
<br>
|
| 23 |
|
| 24 |
# LEVUM
|
| 25 |
|
| 26 |
+
[π Visit the Velum AI Landing Page](https://guell11.github.io/velum-ai/)
|
| 27 |
|
| 28 |
+
### Local Β· Code Β· Software
|
| 29 |
|
| 30 |
+
**A local model based on Qwen 3.5 9B, focused on software development.**
|
| 31 |
|
| 32 |
+
Write. Understand. Fix. Refactor. Build.
|
| 33 |
|
| 34 |
+
**Less ceremony. More working code.**
|
| 35 |
|
| 36 |
</div>
|
| 37 |
|
| 38 |
---
|
| 39 |
|
| 40 |
+
## About
|
| 41 |
|
| 42 |
+
**LEVUM** is a **9-billion-parameter model**, based on **Qwen 3.5 9B** and focused on software development tasks.
|
| 43 |
|
| 44 |
+
The goal is simple: work close to the project and turn instructions into useful code.
|
| 45 |
|
| 46 |
+
The model was designed for tasks such as:
|
| 47 |
|
| 48 |
+
* code generation;
|
| 49 |
+
* feature implementation;
|
| 50 |
* debugging;
|
| 51 |
+
* refactoring;
|
| 52 |
+
* code explanation and understanding;
|
| 53 |
+
* prototype creation;
|
| 54 |
+
* project structuring;
|
| 55 |
+
* complete application generation;
|
| 56 |
+
* assistance during local development.
|
| 57 |
|
| 58 |
+
LEVUM follows a **local-first** philosophy: the model can run close to the IDE, files, and the actual project context, depending on the chosen runtime and quantization.
|
| 59 |
|
| 60 |
+
> **Software first. Local by default.**
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
## Specifications
|
| 65 |
|
| 66 |
+
| | |
|
| 67 |
+
| --------------- | ----------------------------------------------------- |
|
| 68 |
+
| **Model** | LEVUM |
|
| 69 |
+
| **Base** | Qwen 3.5 9B |
|
| 70 |
+
| **Parameters** | 9B |
|
| 71 |
+
| **Focus** | Code / Software |
|
| 72 |
+
| **Primary use** | Software development |
|
| 73 |
+
| **Execution** | Local |
|
| 74 |
+
| **Language** | English + multilingual capabilities of the base model |
|
| 75 |
+
| **Origin** | π§π· Brazil |
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
# Quantizations
|
| 80 |
|
| 81 |
+
LEVUM is available in different quantization levels to support a wider range of hardware.
|
| 82 |
|
| 83 |
+
The ideal choice mainly depends on:
|
| 84 |
|
| 85 |
+
**RAM / VRAM β speed β fidelity**
|
| 86 |
|
| 87 |
+
Comparison of the model's relative quality after different quantization levels:
|
| 88 |
|
| 89 |

|
| 90 |
|
| 91 |
+
Defined values: Q8 = 99%, Q4 = 50%, Q3 = 30%, Q2 = 25%, and Q1 = 10%.
|
| 92 |
|
| 93 |
+
### Q8 β Fidelity
|
| 94 |
|
| 95 |
+
For machines with enough memory and users who want to preserve as much of the model's quality as possible.
|
| 96 |
|
| 97 |
`Q8_0`
|
| 98 |
|
| 99 |
---
|
| 100 |
|
| 101 |
+
### Q4 / IQ4 β Balanced
|
| 102 |
|
| 103 |
+
The recommended sweet spot for many local systems.
|
| 104 |
|
| 105 |
+
A good balance between size, memory usage, and quality.
|
| 106 |
|
| 107 |
`Q4_K_M` Β· `Q4_K_S` Β· `IQ4_XS` Β· `IQ4_NL` Β· `Q4_1` Β· `Q4_0`
|
| 108 |
|
| 109 |
---
|
| 110 |
|
| 111 |
+
### Q3 / IQ3 β Compact
|
| 112 |
|
| 113 |
+
For more limited hardware or situations where reducing RAM/VRAM usage is a priority.
|
| 114 |
|
| 115 |
`Q3_K_L` Β· `Q3_K_M` Β· `Q3_K_S` Β· `IQ3_M` Β· `IQ3_S` Β· `IQ3_XS` Β· `IQ3_XXS`
|
| 116 |
|
| 117 |
---
|
| 118 |
|
| 119 |
+
### Q2 / IQ2 β Ultra-Compact
|
| 120 |
|
| 121 |
+
Aggressive compression for environments where larger versions simply do not fit.
|
| 122 |
|
| 123 |
`Q2_K` Β· `Q2_K_S` Β· `IQ2_M` Β· `IQ2_S` Β· `IQ2_XS` Β· `IQ2_XXS` Β· `Q2_0` Β· `TQ2_0`
|
| 124 |
|
| 125 |
---
|
| 126 |
|
| 127 |
+
### IQ1 β Minimum
|
| 128 |
|
| 129 |
+
The extreme option.
|
| 130 |
|
| 131 |
`IQ1_M`
|
| 132 |
|
| 133 |
+
Recommended only when saving memory is more important than preserving maximum model fidelity.
|
| 134 |
|
| 135 |
---
|
| 136 |
|
| 137 |
+
> **Rule of thumb:** start with `Q4_K_M`.
|
| 138 |
+
> If you have memory to spare, try `Q8_0`.
|
| 139 |
+
> If memory is limited, move down to Q3, Q2, or IQ1.
|
| 140 |
|
| 141 |
---
|
| 142 |
|
| 143 |
+
# Running Locally
|
| 144 |
|
| 145 |
## llama.cpp
|
| 146 |
|
| 147 |
+
Download one of the GGUF versions of LEVUM and run it with a GGUF-compatible runtime.
|
| 148 |
|
| 149 |
```bash
|
| 150 |
llama-cli \
|
| 151 |
-m ./LEVUM-Q4_K_M.gguf \
|
| 152 |
+
-p "Create a REST API in Python using FastAPI."
|
| 153 |
```
|
| 154 |
|
| 155 |
+
To start a local server:
|
| 156 |
|
| 157 |
```bash
|
| 158 |
llama-server \
|
|
|
|
| 160 |
-c 8192
|
| 161 |
```
|
| 162 |
|
| 163 |
+
After that, the model can be integrated with local tools that support compatible endpoints.
|
| 164 |
|
| 165 |
---
|
| 166 |
|
| 167 |
## Ollama
|
| 168 |
|
| 169 |
+
Create a `Modelfile` pointing to the GGUF:
|
| 170 |
|
| 171 |
```dockerfile
|
| 172 |
FROM ./LEVUM-Q4_K_M.gguf
|
|
|
|
| 175 |
PARAMETER top_p 0.9
|
| 176 |
```
|
| 177 |
|
| 178 |
+
Then:
|
| 179 |
|
| 180 |
```bash
|
| 181 |
ollama create levum -f Modelfile
|
| 182 |
ollama run levum
|
| 183 |
```
|
| 184 |
|
| 185 |
+
Example:
|
| 186 |
|
| 187 |
```text
|
| 188 |
+
>>> Create a FastAPI API for managing projects and tasks.
|
| 189 |
```
|
| 190 |
|
| 191 |
---
|
| 192 |
|
| 193 |
## LM Studio
|
| 194 |
|
| 195 |
+
1. Download a GGUF quantization of LEVUM.
|
| 196 |
+
2. Import the file into LM Studio.
|
| 197 |
+
3. Load the model.
|
| 198 |
+
4. Adjust the context size according to the available memory.
|
| 199 |
+
5. Start a conversation or the local server.
|
| 200 |
|
| 201 |
+
No remote infrastructure is required for inference when the model is running locally.
|
| 202 |
|
| 203 |
---
|
| 204 |
|
| 205 |
# Prompts
|
| 206 |
|
| 207 |
+
LEVUM works best when the task, context, and expected output format are explicit.
|
| 208 |
|
| 209 |
+
## Generate a Project
|
| 210 |
|
| 211 |
```text
|
| 212 |
+
Create an interactive financial dashboard in a single HTML file.
|
| 213 |
|
| 214 |
+
Requirements:
|
| 215 |
+
- HTML, CSS, and JavaScript in the same file
|
| 216 |
+
- interactive charts
|
| 217 |
+
- responsive
|
| 218 |
+
- sample data
|
| 219 |
+
- no mandatory external dependencies
|
| 220 |
|
| 221 |
+
Return only the complete HTML.
|
| 222 |
```
|
| 223 |
|
| 224 |
---
|
| 225 |
|
| 226 |
+
## Implement a Feature
|
| 227 |
|
| 228 |
```text
|
| 229 |
+
Analyze the code below and implement JWT authentication.
|
| 230 |
|
| 231 |
+
Requirements:
|
| 232 |
+
- preserve the current architecture
|
| 233 |
+
- validate expired tokens
|
| 234 |
+
- add authentication middleware
|
| 235 |
+
- do not modify public endpoints
|
| 236 |
+
- explain only important decisions
|
| 237 |
|
| 238 |
+
Code:
|
| 239 |
+
[paste the code here]
|
| 240 |
```
|
| 241 |
|
| 242 |
---
|
|
|
|
| 244 |
## Debug
|
| 245 |
|
| 246 |
```text
|
| 247 |
+
Find the cause of the bug in the code below.
|
| 248 |
|
| 249 |
+
Expected behavior:
|
| 250 |
+
[describe]
|
| 251 |
|
| 252 |
+
Current behavior:
|
| 253 |
+
[describe]
|
| 254 |
|
| 255 |
+
Error:
|
| 256 |
+
[paste the error]
|
| 257 |
|
| 258 |
+
Code:
|
| 259 |
+
[paste the code]
|
| 260 |
|
| 261 |
+
Identify the cause and return the complete fix.
|
| 262 |
```
|
| 263 |
|
| 264 |
---
|
| 265 |
|
| 266 |
+
## Refactoring
|
| 267 |
|
| 268 |
```text
|
| 269 |
+
Refactor this code.
|
| 270 |
|
| 271 |
+
Goals:
|
| 272 |
+
- reduce duplication
|
| 273 |
+
- improve readability
|
| 274 |
+
- preserve current behavior
|
| 275 |
+
- preserve the public API
|
| 276 |
+
- avoid unnecessary abstractions
|
| 277 |
|
| 278 |
+
Return the refactored code first, followed by a short summary of the changes.
|
| 279 |
```
|
| 280 |
|
| 281 |
---
|
| 282 |
|
| 283 |
+
# Prompt Format
|
| 284 |
|
| 285 |
+
For larger tasks, a simple structure usually produces more predictable results:
|
| 286 |
|
| 287 |
```text
|
| 288 |
+
OBJECTIVE
|
| 289 |
+
What needs to be built.
|
| 290 |
|
| 291 |
+
CONTEXT
|
| 292 |
+
Stack, existing files, and architecture.
|
| 293 |
|
| 294 |
+
REQUIREMENTS
|
| 295 |
+
Mandatory behaviors.
|
| 296 |
|
| 297 |
+
CONSTRAINTS
|
| 298 |
+
What must not be changed.
|
| 299 |
|
| 300 |
+
OUTPUT
|
| 301 |
+
Exact expected format.
|
| 302 |
```
|
| 303 |
|
| 304 |
+
For example:
|
| 305 |
|
| 306 |
```text
|
| 307 |
+
OBJECTIVE
|
| 308 |
+
Create an analytics page.
|
| 309 |
|
| 310 |
+
CONTEXT
|
| 311 |
+
React + TypeScript + Tailwind project.
|
| 312 |
|
| 313 |
+
REQUIREMENTS
|
| 314 |
+
- revenue chart
|
| 315 |
+
- period filters
|
| 316 |
+
- metric cards
|
| 317 |
+
- transactions table
|
| 318 |
|
| 319 |
+
CONSTRAINTS
|
| 320 |
+
- do not add new dependencies
|
| 321 |
+
- reuse existing components
|
| 322 |
|
| 323 |
+
OUTPUT
|
| 324 |
+
Return the complete files that need to be created or modified.
|
| 325 |
```
|
| 326 |
|
| 327 |
---
|
| 328 |
|
| 329 |
# Code-first
|
| 330 |
|
| 331 |
+
LEVUM was designed to work within the normal development cycle:
|
| 332 |
|
| 333 |
```text
|
| 334 |
+
PROMPT / CODE
|
| 335 |
β
|
| 336 |
LEVUM
|
| 337 |
β
|
| 338 |
+
UNDERSTAND
|
| 339 |
β
|
| 340 |
+
GENERATE
|
| 341 |
β
|
| 342 |
DEBUG
|
| 343 |
β
|
| 344 |
+
REFACTOR
|
| 345 |
β
|
| 346 |
+
RUNNING SOFTWARE
|
| 347 |
```
|
| 348 |
|
| 349 |
+
### Generation
|
| 350 |
|
| 351 |
+
Projects, components, APIs, scripts, and features.
|
| 352 |
|
| 353 |
### Debug
|
| 354 |
|
| 355 |
+
Analysis of errors, unexpected behavior, and fixes.
|
| 356 |
|
| 357 |
+
### Refactoring
|
| 358 |
|
| 359 |
+
Structure, readability, and maintenance of existing code.
|
| 360 |
|
| 361 |
+
### Prototyping
|
| 362 |
|
| 363 |
+
Turn an idea into something executable quickly.
|
| 364 |
|
| 365 |
---
|
| 366 |
|
| 367 |
+
# Example
|
| 368 |
|
| 369 |
**Prompt**
|
| 370 |
|
| 371 |
```text
|
| 372 |
+
Create a Flappy Bird-style game in a single HTML file.
|
| 373 |
|
| 374 |
+
Use only HTML, CSS, and JavaScript.
|
| 375 |
|
| 376 |
+
Include:
|
| 377 |
+
- physics
|
| 378 |
+
- obstacles
|
| 379 |
+
- collision detection
|
| 380 |
+
- scoring
|
| 381 |
+
- restart
|
| 382 |
+
- keyboard and click controls
|
| 383 |
|
| 384 |
+
Return only the complete HTML.
|
| 385 |
```
|
| 386 |
|
| 387 |
+
**Expected result**
|
| 388 |
|
| 389 |
```text
|
| 390 |
prompt
|
|
|
|
| 395 |
β
|
| 396 |
browser
|
| 397 |
β
|
| 398 |
+
running software
|
| 399 |
```
|
| 400 |
|
| 401 |
---
|
| 402 |
|
| 403 |
# Hardware
|
| 404 |
|
| 405 |
+
Actual resource usage depends on several factors, including:
|
| 406 |
|
| 407 |
+
* quantization;
|
| 408 |
+
* context size;
|
| 409 |
* runtime;
|
| 410 |
* KV cache;
|
| 411 |
* CPU;
|
| 412 |
* GPU;
|
| 413 |
+
* number of layers offloaded to the GPU;
|
| 414 |
+
* inference configuration.
|
| 415 |
|
| 416 |
+
Because of this, memory requirements can vary significantly between systems.
|
| 417 |
|
| 418 |
+
As a general rule:
|
| 419 |
|
| 420 |
```text
|
| 421 |
+
more bits
|
| 422 |
β
|
| 423 |
+
more fidelity
|
| 424 |
β
|
| 425 |
+
more memory
|
| 426 |
|
| 427 |
+
fewer bits
|
| 428 |
β
|
| 429 |
+
less memory
|
| 430 |
β
|
| 431 |
+
more compression
|
| 432 |
```
|
| 433 |
|
| 434 |
+
Choose the quantization based on the available hardware and the quality required for the task.
|
| 435 |
|
| 436 |
---
|
| 437 |
|
| 438 |
+
# Limitations
|
| 439 |
|
| 440 |
+
LEVUM is still a language model.
|
| 441 |
|
| 442 |
+
This means it may:
|
| 443 |
|
| 444 |
+
* generate incorrect code;
|
| 445 |
+
* hallucinate APIs or libraries;
|
| 446 |
+
* produce insecure solutions;
|
| 447 |
+
* misinterpret requirements;
|
| 448 |
+
* introduce regressions;
|
| 449 |
+
* suggest nonexistent dependencies;
|
| 450 |
+
* generate code that looks correct without actually working.
|
| 451 |
|
| 452 |
+
For important software, review, test, and validate the code before putting it into production.
|
| 453 |
|
| 454 |
+
AI-generated code does not gain magical powers just because it compiled once.
|
| 455 |
|
| 456 |
---
|
| 457 |
|
| 458 |
+
# Responsible Use
|
| 459 |
|
| 460 |
+
Before running code generated by the model:
|
| 461 |
|
| 462 |
+
1. review the changes;
|
| 463 |
+
2. verify dependencies;
|
| 464 |
+
3. run tests;
|
| 465 |
+
4. validate external inputs;
|
| 466 |
+
5. review filesystem, network, and database operations;
|
| 467 |
+
6. do not expose secrets or credentials unnecessarily;
|
| 468 |
+
7. use isolated environments when testing unknown code.
|
| 469 |
|
| 470 |
+
For critical applications, the model should serve as an assistance tool, not as the only layer of review.
|
| 471 |
|
| 472 |
---
|
| 473 |
|
| 474 |
+
# Base Model
|
| 475 |
|
| 476 |
+
LEVUM is based on:
|
| 477 |
|
| 478 |
**Qwen 3.5 9B**
|
| 479 |
|
| 480 |
+
The base model provides the general capabilities upon which LEVUM is built.
|
| 481 |
|
| 482 |
+
Also review the base model's model card and license before distributing or using derivatives.
|
| 483 |
|
| 484 |
---
|
| 485 |
|
| 486 |
+
# License
|
| 487 |
|
| 488 |
+
Use of LEVUM is subject to the license published in this repository and, where applicable, the terms and conditions associated with the base model.
|
| 489 |
|
| 490 |
+
Review the `LICENSE` file before commercial use, redistribution, or creating derivatives.
|
| 491 |
|
| 492 |
---
|
| 493 |
|
| 494 |
+
# Brazil
|
| 495 |
|
| 496 |
<div align="center">
|
| 497 |
|
| 498 |
+
### π§π· MADE IN BRAZIL
|
| 499 |
|
| 500 |
+
**Intelligence that stays close.**
|
| 501 |
|
| 502 |
Local-first Β· Code-first Β· Software-first
|
| 503 |
|
|
|
|
| 509 |
|
| 510 |
**LEVUM Β© 2026**
|
| 511 |
|
| 512 |
+
</div>
|