Instructions to use Brunobkr/OFFELLIA_Quantis with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use Brunobkr/OFFELLIA_Quantis with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="Brunobkr/OFFELLIA_Quantis", filename="OFFELLIA_GELab-Engine-7B_IQ4_XS.gguf", )
llm.create_chat_completion( messages = "No input example has been defined for this model task." )
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- llama.cpp
How to use Brunobkr/OFFELLIA_Quantis with llama.cpp:
Install from brew
brew install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Brunobkr/OFFELLIA_Quantis:Q4_K_M # Run inference directly in the terminal: llama-cli -hf Brunobkr/OFFELLIA_Quantis:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Brunobkr/OFFELLIA_Quantis:Q4_K_M # Run inference directly in the terminal: llama-cli -hf Brunobkr/OFFELLIA_Quantis: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 Brunobkr/OFFELLIA_Quantis:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf Brunobkr/OFFELLIA_Quantis: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 Brunobkr/OFFELLIA_Quantis:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf Brunobkr/OFFELLIA_Quantis:Q4_K_M
Use Docker
docker model run hf.co/Brunobkr/OFFELLIA_Quantis:Q4_K_M
- LM Studio
- Jan
- Ollama
How to use Brunobkr/OFFELLIA_Quantis with Ollama:
ollama run hf.co/Brunobkr/OFFELLIA_Quantis:Q4_K_M
- Unsloth Studio new
How to use Brunobkr/OFFELLIA_Quantis 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 Brunobkr/OFFELLIA_Quantis 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 Brunobkr/OFFELLIA_Quantis to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Brunobkr/OFFELLIA_Quantis to start chatting
- Pi new
How to use Brunobkr/OFFELLIA_Quantis with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama-server -hf Brunobkr/OFFELLIA_Quantis:Q4_K_M
Configure the model in Pi
# Install Pi: npm install -g @mariozechner/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "Brunobkr/OFFELLIA_Quantis:Q4_K_M" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent new
How to use Brunobkr/OFFELLIA_Quantis with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama-server -hf Brunobkr/OFFELLIA_Quantis: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 Brunobkr/OFFELLIA_Quantis:Q4_K_M
Run Hermes
hermes
- Docker Model Runner
How to use Brunobkr/OFFELLIA_Quantis with Docker Model Runner:
docker model run hf.co/Brunobkr/OFFELLIA_Quantis:Q4_K_M
- Lemonade
How to use Brunobkr/OFFELLIA_Quantis with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Brunobkr/OFFELLIA_Quantis:Q4_K_M
Run and chat with the model
lemonade run user.OFFELLIA_Quantis-Q4_K_M
List all available models
lemonade list
Upload ΩFFΣLLIαHLX_Qwen3.5-35B_MXFP4_MOE.gguf
Browse files3,5 tokens seg constant, more 200000 ctx full




, in Ryzen5 AMD only CPU, 16gb ram, itś a proof of FN HELICOIDAL, enjoy this for free and opensource.
/home/userk21/llama-pure/build/bin/llama-server \
-m /home/userk21/GGUFS/ΩFFΣLLIαHLX_Qwen3.5-35B_MXFP4_MOE.gguf
--ignore-eos \
--temp 0.8 \
--repeat-penalty 1.15 \




Antes analise o programa feito ja em 20 minutos kkkkkkkkkk: <<<reasoning_content_start>>>Here's a thinking process that leads to the suggested Flappy Bird clone:
1. **Analyze the Request:**
* **Goal:** Create a clone of the game "Flappy Bird".
* **Style:** Modern and functional.
* **Platform:** Notebook (PC) without a dedicated GPU (meaning performance needs to be optimized for CPU-only rendering, standard resolution, no heavy effects).
* **Language:** Python.
* **Library:** `pygame`.
* **Constraint:** Single `.py` file (all code, assets, logic in one).
* **Language of Response:** Portuguese.
2. **Determine Key Features for a "Modern" Feel (within constraints):**
* **Graphics:** Since I can't load external images (single file constraint usually implies no dependencies unless I base64 encode them, but standard practice for these requests is using geometric shapes or simple drawing), I will use `pygame.draw` to create clean, vector-like graphics. This ensures it runs fast without a GPU.
* **Physics:** Smooth gravity, velocity, collision detection.
* **UI:** Score display, game over screen, restart mechanic.
* **Optimization:** Minimal frame rate checks (or standard `clock.tick`), efficient collision logic.
3. **Drafting the Code Structure:**
* **Imports:** `pygame`, `sys`, `random`, `math` (maybe not needed).
* **Constants:** Screen size, colors, gravity, jump strength, pipe speed, gap size.
* **Classes:**
* `Bird`: Position, velocity, drawing, jumping logic.
* `Pipe`: Position, moving logic, drawing.
* `Game`: State management (Start, Playing, Game Over), Score, Loop.
* **Main Loop:** Event handling, updating, drawing, clock.
4. **Addressing the "No GPU" Constraint:**
* Avoid complex particle effects or heavy texture scaling.
* Use solid colors for background and objects.
* Keep the resolution reasonable (e.g., 480x640 or 800x600). 480x640 is classic mobile aspect ratio, 800x600 is standard desktop. Let's go with 480x640 for the Flappy Bird feel, but scaled if needed. Actually, standard desktop is safer for "Notebook". Let's stick to 320x480 (scaled up) or 480x640. Let's choose 480 width x 640 height for the classic look.
* *Self-Correction:* To make it look "modern" without external assets, I'll use smooth gradients or nice border-radius style drawing (circles/rects).
5. **Drafting the Logic (Mental or Scratchpad):**
* *Bird:* `y`, `velocity`. `update()`: velocity += gravity, `y += velocity`. `jump()`: velocity = -jump_strength.
* *Pipe:* `x` moves left. Two rects (top, bottom). Random `height` for the gap.
* *Collision:* Bird rect intersects Pipe rect OR Bird hits floor/ceiling.
* *Score:* Increment when passing a pipe.
* *Game Loop:* Clear screen -> Update Bird -> Update Pipes -> Check Collisions -> Draw everything.
6. **Refining for "Single File" & "No External Assets":**
* I cannot include `.png` files. I must draw everything using `pygame.draw`.
* This makes the file lightweight.
7. **Writing the Code (Iterative refinement):**
* *Setup:* `pygame.init()`, `screen = pygame.display.set_mode()`, `clock = pygame.time.Clock()`.
* *Bird:* Yellow circle-ish.
* *Pipes:* Green rectangles.
* *Background:* Sky blue.
* *Game Over:* Text overlay.
* *Optimization:* Ensure `clock.tick(60)` is used.
8. **Adding "Modern" Touches:**
* Smooth transitions for score.
* Clean fonts (system default or loaded from memory if possible, but standard `pygame.font` is safer).
* Visual feedback (maybe a slight flash on collision or just a clear Game Over screen). Let's stick to a clean Game Over screen with a "Press Space to Restart".
9. **Reviewing against constraints:**
* *Python?* Yes.
* *Pygame?* Yes.
* *One .py?* Yes.
* *No GPU?* Yes, it's pure 2D drawing, very CPU friendly.
* *Portuguese?* Yes.
10. **Final Code Assembly:** (This involves writing the actual Python script, ensuring indentation is correct, comments are in Portuguese, and logic is sound.)
* *Wait, one detail:* High DPI on some notebooks can make things blurry. I'll stick to standard integer coordinates.
* *Performance:* Use a simple list for pipes to manage collision.
* *Assets:* None needed.
11. **Final Polish:** Add comments explaining how to run and the design choices.
*(Self-Correction during coding)*: I should make sure the bird rotation isn't too complex to avoid lag. A simple vertical rotation or just a static shape is better for a basic clone. I'll add a slight tilt based on velocity for that "modern" feel without heavy calculation.
Let's write the response. I'll provide the code block and instructions on how to save and run it.
12. **Output Generation:** (Proceed to generate the Portuguese response with the code).
*(Wait, check specific request details again)*: "Clone moderno e funcional... sem GPU".
Okay, I will emphasize that the code is optimized for CPU usage by avoiding heavy image loading or complex shader effects. I will use geometric primitives.
Let's build the script.
* `BIRD_COLOR`: Yellow.
* `PIPE_COLOR`: Green.
* `BIRD_SIZE`: Small circle.
* `GRAVITY`: Standard float.
* `JUMP`: Standard negative velocity.
Ready to generate.cs<<<reasoning_content_end>>>Com certeza! Abaixo está um código completo, otimizado para rodar em qualquer notebook (CPU-only), utilizando apenas **Pygame** e **primitivas de desenho** (sem necessidade de imagens externas), garantindo leveza e performance.
O design é "neon/clean" para dar o toque moderno que você pediu.
### Como usar:
1. Certifique-se de ter o Python instalado.
2. Instale o Pygame: `pip install pygame`
3. Salve o código abaixo como `flappy.py`.
4. Execute no terminal: `python flappy.py`
```python
import pygame
import sys
import random
# --- Configurações Iniciais ---
pygame.init()
# Dimensões (Focadas em performance, baixa resolução interna é melhor para CPU)
SCREEN_WIDTH = 480
SCREEN_HEIGHT = 640
FPS = 60
# Cores (Estilo Moderno Clean)
COLOR_BG = (200, 220, 255) # Céu Azul Pastel
COLOR_BIRD = (255, 215, 0) # Dourado
COLOR_PIPE = (46, 204, 113) # Verde Esmeralda
COLOR_PIPE_BORDER = (39, 174, 96)
COLOR_TEXT = (255, 255, 255)
COLOR_TEXT_SHADOW = (0, 0, 0)
# Constantes do Jogo
GRAVITY = 0.25
JUMP_STRENGTH = -6
PIPE_SPEED = 3
PIPE_GAP = 150
PIPE_FREQUENCY = 1000 # Millissegundos
# Configuração da Tela
screen = pygame.display.set_mode((SCREEN_WIDTH, SCREEN_HEIGHT))
pygame.display.set_caption("Flappy Bird - Moderno CPU")
clock = pygame.time.Clock()
font = pygame.font.SysFont("Arial", 30, bold=True)
font_small = pygame.font.SysFont("Arial", 20)
# --- Classes ---
class Bird:
def __init__(self):
self.x = SCREEN_WIDTH // 3
self.y = SCREEN_HEIGHT // 2
self.velocity = 0
self.radius = 20
self.color = COLOR_BIRD
self.rotation = 0
def jump(self):
self.velocity = JUMP_STRENGTH
def update(self):
self.velocity += GRAVITY
self.y += self.velocity
# Rotação baseada na velocidade
self.rotation = min(pygame.math.Vector2(0, self.velocity).angle_to((0, -1)), 90)
self.rotation = max(self.rotation, -90)
def draw(self, surface):
# Desenha o pássaro como um círculo simples (leve para a CPU)
points = []
for i in range(6):
angle = (i / 6) * 360
rad = pygame.math.Vector2(1, 0).rotate(-self.rotation)
x = self.x + rad.x * self.radius
y = self.y + rad.y * self.radius
points.append((x, y))
# Arredondar para parecer suave
pygame.draw.circle(surface, self.color, (int(self.x), int(self.y)), self.radius)
# Olho
eye_offset_x = self.x + self.radius * 0.5
eye_offset_y = self.y - self.radius * 0.2
pygame.draw.circle(surface, (0, 0, 0), (int(eye_offset_x), int(eye_offset_y)), 5)
# Bico
beak_x = self.x + self.radius * 0.8
beak_y = self.y + self.radius * 0.2
pygame.draw.polygon(surface, (255, 100, 0), [
(int(beak_x), int(beak_y)),
(int(beak_x + 10), int(beak_y - 5)),
(int(beak_x + 10), int(beak_y + 5))
])
def get_rect(self):
return pygame.Rect(self.x - self.radius, self.y
- .gitattributes +1 -0
- ΩFFΣLLIαHLX_Qwen3.5-35B_MXFP4_MOE.gguf +3 -0
|
@@ -95,3 +95,4 @@ OFFELLIA_Qwen2.5-Coder-7B-Instruct_Q4_0.gguf filter=lfs diff=lfs merge=lfs -text
|
|
| 95 |
ΩFFΣLLIαHLX_LFM2-24B-A2B-A2B_Q8.gguf filter=lfs diff=lfs merge=lfs -text
|
| 96 |
ΩFFΣLLIαHLX_LFM2-8B-A1B_MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
|
| 97 |
ΩFFΣLLIαHLX_LFM2-24B-A2B_MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 95 |
ΩFFΣLLIαHLX_LFM2-24B-A2B-A2B_Q8.gguf filter=lfs diff=lfs merge=lfs -text
|
| 96 |
ΩFFΣLLIαHLX_LFM2-8B-A1B_MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
|
| 97 |
ΩFFΣLLIαHLX_LFM2-24B-A2B_MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
|
| 98 |
+
ΩFFΣLLIαHLX_Qwen3.5-35B_MXFP4_MOE.gguf filter=lfs diff=lfs merge=lfs -text
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ffa9dcd035ec3f81505c38eb7ccb06bb000e6404e5549bc7b0e92d16c1445e7
|
| 3 |
+
size 19790379360
|