GGUF
conversational
File size: 8,388 Bytes
9031c8c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0fc06d4
9031c8c
 
 
 
 
 
 
 
 
 
 
 
0fc06d4
9031c8c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
---
license: apache-2.0
base_model:
- Qwen/Qwen3-Coder-480B-A35B-Instruct
datasets:
- codeparrot/github-code-clean
---

## Model Details

This model is a mixed gguf q2ks format of [Qwen/Qwen3-Coder-480B-A35B-Instruct](https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct) generated by [intel/auto-round](https://github.com/intel/auto-round) algorithm.  Embedding layer and lm-head layer are fallback to 8 bits and non expert layers are fallback to 4 bits. Please refer to Section `Generate the model` for more details.

Please follow the license of the original model.

## How To Use

Llamacpp inference

~~~bash
/llama-cli -hf Intel/Qwen3-Coder-480B-A35B-Instruct-gguf-q2ks-mixed-AutoRound:q2_k_s --conversation

~~~

~~~ba
> code a flappy bird in python
Here's a complete Flappy Bird implementation using Pygame:

```python
import pygame
import random
import sys

# Initialize pygame
pygame.init()

# Game constants
WIDTH, HEIGHT = 400, 600
FPS = 60
GRAVITY = 0.5
JUMP = -10
PIPE_SPEED = 3
PIPE_GAP = 150
PIPE_FREQUENCY = 1500  # milliseconds

# Colors
WHITE = (255, 255, 255)
BLACK = (0, 0, 0)
GREEN = (0, 255, 0)
BLUE = (0, 100, 255)

# Create game window
screen = pygame.display.set_mode((WIDTH, HEIGHT))
pygame.display.set_caption("Flappy Bird")
clock = pygame.time.Clock()

# Font
font = pygame.font.SysFont(None, 36)

class Bird:
    def __init__(self):
        self.x = 50
        self.y = HEIGHT // 2
        self.velocity = 0
        self.radius = 15

    def jump(self):
        self.velocity = JUMP

    def update(self):
        self.velocity += GRAVITY
        self.y += self.velocity

    def draw(self):
        pygame.draw.circle(screen, BLUE, (self.x, int(self.y)), self.radius)

    def get_rect(self):
        return pygame.Rect(
            self.x - self.radius,
            self.y - self.radius,
            self.radius * 2,
            self.radius * 2
        )

class Pipe:
    def __init__(self):
        self.x = WIDTH
        self.height = random.randint(50, HEIGHT - PIPE_GAP - 50)
        self.width = 50
        self.passed = False

    def update(self):
        self.x -= PIPE_SPEED

    def draw(self):
        # Top pipe
        pygame.draw.rect(screen, GREEN, (self.x, 0, self.width, self.height))
        # Bottom pipe
        pygame.draw.rect(screen, GREEN, (self.x, self.height + PIPE_GAP, self.width, HEIGHT))

    def collide(self, bird):
        bird_rect = bird.get_rect()
        top_pipe = pygame.Rect(self.x, 0, self.width, self.height)
        bottom_pipe = pygame.Rect(self.x, self.height + PIPE_GAP, self.width, HEIGHT)
        return bird_rect.colliderect(top_pipe) or bird_rect.colliderect(bottom_pipe)

    def off_screen(self):
        return self.x < -self.width

def draw_ground():
    pygame.draw.line(screen, BLACK, (0, HEIGHT - 50), (WIDTH, HEIGHT - 50), 5)

def main():
    bird = Bird()
    pipes = []
    score = 0
    last_pipe = pygame.time.get_ticks()
    game_over = False

    while True:
        # Event handling
        for event in pygame.event.get():
            if event.type == pygame.QUIT:
                pygame.quit()
                sys.exit()
            if event.type == pygame.KEYDOWN:
                if event.key == pygame.K_SPACE:
                    if game_over:
                        # Restart game
                        bird = Bird()
                        pipes = []
                        score = 0
                        last_pipe = pygame.time.get_ticks()
                        game_over = False
                    else:
                        bird.jump()

        if not game_over:
            # Update bird
            bird.update()

            # Generate new pipes
            current_time = pygame.time.get_ticks()
            if current_time - last_pipe > PIPE_FREQUENCY:
                pipes.append(Pipe())
                last_pipe = current_time


            # Update pipes and check for score
            for pipe in pipes:
                pipe.update()
                if not pipe.passed and pipe.x < bird.x:
                    pipe.passed = True
                    score += 1

            # Remove off-screen pipes
            pipes = [pipe for pipe in pipes if not pipe.off_screen()]

            # Check collisions
            if bird.y > HEIGHT - 50 - bird.radius or bird.y < bird.radius:
                game_over = True
            for pipe in pipes:
                if pipe.collide(bird):
                    game_over = True

        # Draw everything
        screen.fill(WHITE)
        for pipe in pipes:
            pipe.draw()
        bird.draw()
        draw_ground()

        # Draw score
        score_text = font.render(f"Score: {score}", True, BLACK)
        screen.blit(score_text, (10, 10))

        # Draw game over message
        if game_over:
            game_over_text = font.render("Game Over! Press SPACE to restart", True, BLACK)
            screen.blit(game_over_text, (WIDTH//2 - game_over_text.get_width()//2, HEIGHT//2))

        pygame.display.flip()
        clock.tick(FPS)

if __name__ == "__main__":
    main()
```

### Features:
1. **Bird Physics**:
   - Gravity pulls the bird down
   - Space bar makes the bird jump upward

2. **Pipe System**:
   - Pipes generate randomly with gaps
   - Pipes move from right to left
   - Collision detection with bird

3. **Scoring**:
   - Score increases when bird passes through pipes
   - Displayed at top-left of screen

4. **Game States**:
   - Active gameplay
   - Game over screen with restart option

### Controls:
- **SPACE**: Jump (during gameplay) / Restart (after game over)

### Requirements:
- Python 3.x
- Pygame (`pip install pygame`)

### Game Mechanics:
1. The bird automatically falls due to gravity
2. Players must press SPACE to make the bird jump
3. Avoid hitting pipes or the ground
4. Each pipe passed increases the score by 1
5. When game over occurs, press SPACE to restart

This implementation includes all core Flappy Bird mechanics in a single file with minimal dependencies. The game features smooth animations, collision detection, and score tracking.


~~~



### Generate the model

Here is the sample command to reproduce the model

```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from auto_round import AutoRound

model_name = "Qwen/Qwen3-Coder-480B-A35B-Instruct"

model = AutoModelForCausalLM.from_pretrained(model_name,
                                             device_map="cpu", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
layer_config = {}
for n, m in model.named_modules():
    if n == "lm_head" or isinstance(m, torch.nn.Embedding):
        layer_config[n] = {"bits": 8}
    elif isinstance(m, torch.nn.Linear) and (not "expert" in n or "shared_experts" in n) and n != "lm_head":
        layer_config[n] = {"bits": 4}

autoround = AutoRound(model, tokenizer, iters=0, layer_config=layer_config, nsamples=512, dataset="github-code-clean")
autoround.quantize_and_save("./Qwen3-Coder-480B-A35B-Instruct-q2ks", format="gguf:q2_k_s")

```



## Ethical Considerations and Limitations

The model can produce factually incorrect output, and should not be relied on to produce factually accurate information. Because of the limitations of the pretrained model and the finetuning datasets, it is possible that this model could generate lewd, biased or otherwise offensive outputs.

Therefore, before deploying any applications of the model, developers should perform safety testing.

## Caveats and Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.

Here are a couple of useful links to learn more about Intel's AI software:

- Intel Neural Compressor [link](https://github.com/intel/neural-compressor)

## Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.

## Cite

@article{cheng2023optimize, title={Optimize weight rounding via signed gradient descent for the quantization of llms}, author={Cheng, Wenhua and Zhang, Weiwei and Shen, Haihao and Cai, Yiyang and He, Xin and Lv, Kaokao and Liu, Yi}, journal={arXiv preprint arXiv:2309.05516}, year={2023} }

[arxiv](https://arxiv.org/abs/2309.05516) [github](https://github.com/intel/auto-round)