vinoku89 commited on
Commit
3621384
·
verified ·
1 Parent(s): 4227547

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +7 -51
README.md CHANGED
@@ -1,23 +1,16 @@
1
  ---
2
  license: apache-2.0
3
- base_model: microsoft/DialoGPT-medium
4
  tags:
5
  - code-generation
6
  - svg
7
  - lora
8
  - fine-tuned
9
  - peft
10
- - graphics
11
- - art
12
  language:
13
  - en
14
  pipeline_tag: text-generation
15
- library_name: peft
16
- datasets:
17
- - custom-svg-dataset
18
- metrics:
19
- - bleu
20
- - rouge
21
  model_type: lora
22
  inference: true
23
  widget:
@@ -27,19 +20,6 @@ widget:
27
  text: "Draw a blue rectangle with black border"
28
  - example_title: "Complex Shape"
29
  text: "Generate a star with 5 points in yellow"
30
- model-index:
31
- - name: svg-code-generator
32
- results:
33
- - task:
34
- type: text-generation
35
- name: SVG Code Generation
36
- metrics:
37
- - type: bleu
38
- value: 0.85
39
- name: BLEU Score
40
- - type: rouge
41
- value: 0.78
42
- name: ROUGE Score
43
  ---
44
 
45
  # SVG Code Generator
@@ -49,12 +29,11 @@ This is a fine-tuned LoRA adapter for generating SVG code from natural language
49
  ## Model Details
50
 
51
  - **Model Name**: model_v10
52
- - **Base Model**: Fine-tuned language model
53
  - **Training Method**: LoRA (Low-Rank Adaptation)
54
  - **Task**: Text-to-SVG code generation
55
  - **Model Type**: Causal Language Model with LoRA adapter
56
- - **Parameters**: ~7B (base) + LoRA parameters
57
- - **Training Framework**: PyTorch with PEFT
58
 
59
  ## Usage
60
 
@@ -62,41 +41,18 @@ Load the model using the transformers library and PEFT for LoRA adapters. Use na
62
 
63
  ## Training Data
64
 
65
- The model was trained on SVG code generation tasks with natural language descriptions covering:
66
- - Basic shapes (circles, rectangles, polygons)
67
- - Complex graphics and patterns
68
- - Color specifications and styling
69
- - Positioning and sizing instructions
70
 
71
  ## Intended Use
72
 
73
- This model is designed to generate SVG code from text descriptions for:
74
- - Educational purposes
75
- - Creative projects
76
- - Rapid prototyping of graphics
77
- - Learning SVG syntax
78
 
79
  ## Limitations
80
 
81
  - Generated SVG may require validation
82
  - Performance depends on prompt clarity
83
  - Limited to SVG syntax and features seen during training
84
- - May not handle very complex geometric calculations
85
 
86
  ## Model Performance
87
 
88
- The model has been fine-tuned specifically for SVG generation tasks and achieves:
89
- - BLEU Score: 0.85
90
- - ROUGE Score: 0.78
91
- - High accuracy on basic shapes and common patterns
92
-
93
- ## Citation
94
-
95
- If you use this model, please cite:
96
- @misc{svg-code-generator,
97
- title={SVG Code Generator},
98
- author={Your Name},
99
- year={2025},
100
- publisher={HuggingFace},
101
- url={https://huggingface.co/your_username/svg-code-generator}
102
- }
 
1
  ---
2
  license: apache-2.0
3
+ base_model: qwen3-0.6B
4
  tags:
5
  - code-generation
6
  - svg
7
  - lora
8
  - fine-tuned
9
  - peft
 
 
10
  language:
11
  - en
12
  pipeline_tag: text-generation
13
+ library_name: transformers
 
 
 
 
 
14
  model_type: lora
15
  inference: true
16
  widget:
 
20
  text: "Draw a blue rectangle with black border"
21
  - example_title: "Complex Shape"
22
  text: "Generate a star with 5 points in yellow"
 
 
 
 
 
 
 
 
 
 
 
 
 
23
  ---
24
 
25
  # SVG Code Generator
 
29
  ## Model Details
30
 
31
  - **Model Name**: model_v10
32
+ - **Base Model**: qwen3-0.6B
33
  - **Training Method**: LoRA (Low-Rank Adaptation)
34
  - **Task**: Text-to-SVG code generation
35
  - **Model Type**: Causal Language Model with LoRA adapter
36
+ - **Library**: Transformers
 
37
 
38
  ## Usage
39
 
 
41
 
42
  ## Training Data
43
 
44
+ The model was trained on SVG code generation tasks with natural language descriptions.
 
 
 
 
45
 
46
  ## Intended Use
47
 
48
+ This model is designed to generate SVG code from text descriptions for educational and creative purposes.
 
 
 
 
49
 
50
  ## Limitations
51
 
52
  - Generated SVG may require validation
53
  - Performance depends on prompt clarity
54
  - Limited to SVG syntax and features seen during training
 
55
 
56
  ## Model Performance
57
 
58
+ The model has been fine-tuned specifically for SVG generation tasks and should be used within this domain for best results.