Update README.md
Browse files
README.md
CHANGED
@@ -3,7 +3,6 @@ license: apache-2.0
|
|
3 |
base_model:
|
4 |
- openai/gpt-oss-20b
|
5 |
---
|
6 |
-
|
7 |
# GPT-OSS PlantUML Generation Model V1
|
8 |
|
9 |
**Model ID:** `chrisrutherford/gpt-oss-pumlGenV1`
|
@@ -17,17 +16,13 @@ GPT-OSS PlantUML Generation Model V1 is a fine-tuned language model specialised
|
|
17 |
- **Base Model:** GPT-OSS architecture
|
18 |
- **Model Type:** Causal Language Model
|
19 |
- **Language(s):** English
|
20 |
-
- **License:**
|
21 |
- **Fine-tuned from:** [Base model name]
|
22 |
|
23 |
## Training Details
|
24 |
|
25 |
### Training Data
|
26 |
-
The model was fine-tuned on
|
27 |
-
- Philosophical concept mappings
|
28 |
-
- Historical evolution of ideas
|
29 |
-
- Cross-disciplinary connections
|
30 |
-
- Complex abstract relationships
|
31 |
|
32 |
### Training Configuration
|
33 |
- **Optimiser:** AdamW 8-bit
|
@@ -59,53 +54,7 @@ Map the evolution of the concept of 'nothing' from Parmenides through Buddhist
|
|
59 |
```
|
60 |
|
61 |
**Expected Output:**
|
62 |
-
The model generates PlantUML code that creates a comprehensive diagram
|
63 |
-
- Historical progression from ancient Greek philosophy to modern physics
|
64 |
-
- Philosophical interpretations (Parmenides' "nothing cannot exist")
|
65 |
-
- Buddhist concepts (śūnyatā as emptiness/interdependence)
|
66 |
-
- Mathematical formalisations (zero, null sets, vacuum states)
|
67 |
-
- Physical manifestations (quantum vacuum fluctuations, virtual particles)
|
68 |
-
- Interconnections and conceptual bridges between domains
|
69 |
-
|
70 |
-
## Performance
|
71 |
-
|
72 |
-
### Evaluation Metrics
|
73 |
-
- **Diagram Validity:** Generated PlantUML code compiles successfully
|
74 |
-
- **Conceptual Accuracy:** Diagrams accurately represent described relationships
|
75 |
-
- **Completeness:** All requested elements are included in the output
|
76 |
-
- **Clarity:** Diagrams are visually comprehensible and well-structured
|
77 |
-
|
78 |
-
### Benchmark Results
|
79 |
-
The model demonstrates strong performance in:
|
80 |
-
- Cross-disciplinary concept mapping
|
81 |
-
- Historical timeline visualisation
|
82 |
-
- Abstract relationship representation
|
83 |
-
- Complex hierarchical structures
|
84 |
-
|
85 |
-
## Limitations
|
86 |
-
|
87 |
-
- **Domain Specificity:** Optimised for conceptual and philosophical diagrams
|
88 |
-
- **PlantUML Syntax:** Limited to PlantUML's diagramming capabilities
|
89 |
-
- **Language:** Primarily trained on English descriptions
|
90 |
-
- **Complexity:** Very complex diagrams may require iterative refinement
|
91 |
-
|
92 |
-
## Bias and Ethical Considerations
|
93 |
-
|
94 |
-
- **Cultural Bias:** Training data may reflect Western philosophical traditions more heavily
|
95 |
-
- **Historical Perspective:** Interpretations of historical concepts may vary
|
96 |
-
- **Simplification:** Complex philosophical ideas may be oversimplified for visualisation
|
97 |
-
|
98 |
-
## Technical Specifications
|
99 |
-
|
100 |
-
### Model Architecture
|
101 |
-
- **Parameters:** [Specify parameter count]
|
102 |
-
- **Context Length:** [Specify context window]
|
103 |
-
- **Vocabulary Size:** [Specify vocabulary size]
|
104 |
-
|
105 |
-
### Inference Requirements
|
106 |
-
- **Memory:** [Specify RAM requirements]
|
107 |
-
- **Compute:** [Specify GPU/CPU requirements]
|
108 |
-
- **Dependencies:** PlantUML rendering engine for diagram generation
|
109 |
|
110 |
## Usage Examples
|
111 |
|
@@ -167,3 +116,14 @@ generated_text = puml_code[len(prompt):]
|
|
167 |
}
|
168 |
```
|
169 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
base_model:
|
4 |
- openai/gpt-oss-20b
|
5 |
---
|
|
|
6 |
# GPT-OSS PlantUML Generation Model V1
|
7 |
|
8 |
**Model ID:** `chrisrutherford/gpt-oss-pumlGenV1`
|
|
|
16 |
- **Base Model:** GPT-OSS architecture
|
17 |
- **Model Type:** Causal Language Model
|
18 |
- **Language(s):** English
|
19 |
+
- **License:** Apache 2.0
|
20 |
- **Fine-tuned from:** [Base model name]
|
21 |
|
22 |
## Training Details
|
23 |
|
24 |
### Training Data
|
25 |
+
The model was fine-tuned on the PumlGenV1 dataset of natural language descriptions paired with corresponding PlantUML diagram code.
|
|
|
|
|
|
|
|
|
26 |
|
27 |
### Training Configuration
|
28 |
- **Optimiser:** AdamW 8-bit
|
|
|
54 |
```
|
55 |
|
56 |
**Expected Output:**
|
57 |
+
The model generates PlantUML code that creates a comprehensive diagram.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
|
59 |
## Usage Examples
|
60 |
|
|
|
116 |
}
|
117 |
```
|
118 |
|
119 |
+
## Contact
|
120 |
+
|
121 |
+
For questions, issues, or collaboration opportunities, please contact [contact information].
|
122 |
+
|
123 |
+
## Changelog
|
124 |
+
|
125 |
+
### Version 1.0
|
126 |
+
- Initial release
|
127 |
+
- Fine-tuned on philosophical and scientific concept mapping dataset
|
128 |
+
- Optimised for cross-disciplinary diagram generation
|
129 |
+
|