Hugging Face's logo

CubicLabs
/
AXL-Vision-v2

Text Generation
Transformers
PyTorch
GGUF
code
multiscale_transformer
code-generation
multi-scale-transformer
cpu-optimized
cubic
llama
byte-level
vision
multi-modal
Eval Results (legacy)
Model card Files
xet
 Community
Restrictated commited on
Commit
af27b53
·
verified ·
1 Parent(s): c217fa1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -9
README.md CHANGED
@@ -6,7 +6,7 @@ tags:
6
  - code-generation
7
  - multi-scale-transformer
8
  - cpu-optimized
9
- - koinic
10
  - pytorch
11
  - llama
12
  - gguf
@@ -16,7 +16,7 @@ tags:
16
  pipeline_tag: text-generation
17
  library_name: transformers
18
  datasets:
19
- - koinic/axl-synthetic-ui
20
  model-index:
21
  - name: AXL-Vision-v2
22
  results:
@@ -30,13 +30,13 @@ model-index:
30
 
31
  # AXL-Vision-v2
32
 
33
- UI vision. 4.1M params. Screen-to-code prototype. Trained on synthetic UI screenshots. Part of the AXL model family by [KoinicLabs](https://huggingface.co/KoinicLabs).
34
 
35
  ## Model Details
36
 
37
  | Property | Value |
38
  |----------|-------|
39
- | Developed by | [KoinicLabs](https://huggingface.co/KoinicLabs) |
40
  | Architecture | Multi-Scale Transformer |
41
  | Parameters | 4M |
42
  | Optimizer | SGD |
@@ -51,8 +51,8 @@ UI vision. 4.1M params. Screen-to-code prototype. Trained on synthetic UI screen
51
 
52
  ### Sources
53
 
54
- - **Repository:** [GitHub](https://github.com/Koinic/AXL)
55
- - **Organization:** [KoinicLabs](https://huggingface.co/KoinicLabs)
56
 
57
  ## Uses
58
 
@@ -152,9 +152,9 @@ Multi-Scale Transformer with three parallel encoder stacks at resolution scales
152
  ```bibtex
153
  @misc{axl_2026,
154
  title={AXL: AXL-Vision-v2 - Multi-Scale Transformer for CPU Code Generation},
155
- author={Koinic},
156
  year={2026},
157
- url={https://huggingface.co/KoinicLabs}
158
  }
159
  ```
160
 
@@ -185,4 +185,4 @@ ids = torch.tensor([tokenizer.encode(prompt)], dtype=torch.long)
185
  with torch.no_grad():
186
  out = model.generate(ids, max_new_tokens=100, temperature=0.8, top_k=40)
187
  print(tokenizer.decode(out[0].tolist()))
188
- ```
 
6
  - code-generation
7
  - multi-scale-transformer
8
  - cpu-optimized
9
+ - cubic
10
  - pytorch
11
  - llama
12
  - gguf
 
16
  pipeline_tag: text-generation
17
  library_name: transformers
18
  datasets:
19
+ - cubic/axl-synthetic-ui
20
  model-index:
21
  - name: AXL-Vision-v2
22
  results:
 
30
 
31
  # AXL-Vision-v2
32
 
33
+ UI vision. 4.1M params. Screen-to-code prototype. Trained on synthetic UI screenshots. Part of the AXL model family by [CubicLabs](https://huggingface.co/CubicLabs).
34
 
35
  ## Model Details
36
 
37
  | Property | Value |
38
  |----------|-------|
39
+ | Developed by | [CubicLabs](https://huggingface.co/CubicLabs) |
40
  | Architecture | Multi-Scale Transformer |
41
  | Parameters | 4M |
42
  | Optimizer | SGD |
 
51
 
52
  ### Sources
53
 
54
+ - **Repository:** [GitHub](https://github.com/Cubic/AXL)
55
+ - **Organization:** [CubicLabs](https://huggingface.co/CubicLabs)
56
 
57
  ## Uses
58
 
 
152
  ```bibtex
153
  @misc{axl_2026,
154
  title={AXL: AXL-Vision-v2 - Multi-Scale Transformer for CPU Code Generation},
155
+ author={Cubic},
156
  year={2026},
157
+ url={[https://huggingface.co/CubicLabs](https://huggingface.co/CubicLabs)}
158
  }
159
  ```
160
 
 
185
  with torch.no_grad():
186
  out = model.generate(ids, max_new_tokens=100, temperature=0.8, top_k=40)
187
  print(tokenizer.decode(out[0].tolist()))
188
+ ```