chrisrutherford commited on
Commit
10a1fa7
·
verified ·
1 Parent(s): 5fc75ae

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -56
README.md CHANGED
@@ -52,7 +52,9 @@ Map the evolution of the concept of 'nothing' from Parmenides through Buddhist
52
  ```
53
 
54
  **Expected Output:**
55
- The model generates PlantUML code that creates a comprehensive diagram.
 
 
56
 
57
  ## Usage Examples
58
 
@@ -70,58 +72,3 @@ inputs = tokenizer(prompt, return_tensors="pt")
70
  outputs = model.generate(**inputs, max_length=1000)
71
  puml_code = tokenizer.decode(outputs[0], skip_special_tokens=True)
72
  ```
73
-
74
- #### Python with Chat Template
75
- ```python
76
- from transformers import AutoTokenizer, AutoModelForCausalLM
77
-
78
- tokenizer = AutoTokenizer.from_pretrained("chrisrutherford/gpt-oss-pumlGenV1")
79
- model = AutoModelForCausalLM.from_pretrained("chrisrutherford/gpt-oss-pumlGenV1")
80
-
81
- # Format with chat template
82
- messages = [
83
- {"role": "user", "content": "Map the evolution of the concept of 'nothing' from Parmenides through Buddhist śūnyatā to quantum vacuum fluctuations, showing philosophical, mathematical, and physical interpretations"}
84
- ]
85
-
86
- # Apply chat template
87
- prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
88
- inputs = tokenizer(prompt, return_tensors="pt")
89
-
90
- # Generate with stop tokens
91
- outputs = model.generate(
92
- **inputs,
93
- max_length=1000,
94
- do_sample=True,
95
- temperature=0.7,
96
- pad_token_id=tokenizer.eos_token_id,
97
- eos_token_id=tokenizer.eos_token_id
98
- )
99
-
100
- puml_code = tokenizer.decode(outputs[0], skip_special_tokens=True)
101
- # Extract only the generated portion (after the prompt)
102
- generated_text = puml_code[len(prompt):]
103
- ```
104
-
105
- ## Citation
106
-
107
- ```bibtex
108
- @misc{rutherford2024gptossplantumv1,
109
- title={GPT-OSS PlantUML Generation Model V1},
110
- author={Chris Rutherford},
111
- year={2024},
112
- publisher={Hugging Face},
113
- url={https://huggingface.co/chrisrutherford/gpt-oss-pumlGenV1}
114
- }
115
- ```
116
-
117
- ## Contact
118
-
119
- For questions, issues, or collaboration opportunities, please contact [contact information].
120
-
121
- ## Changelog
122
-
123
- ### Version 1.0
124
- - Initial release
125
- - Fine-tuned on philosophical and scientific concept mapping dataset
126
- - Optimised for cross-disciplinary diagram generation
127
-
 
52
  ```
53
 
54
  **Expected Output:**
55
+
56
+ ![image/png](https://cdn-uploads.huggingface.co/production/uploads/65be16980a0c57943fbe8b00/qzuoX_hBMSJsFi_VfpK9M.png)
57
+
58
 
59
  ## Usage Examples
60
 
 
72
  outputs = model.generate(**inputs, max_length=1000)
73
  puml_code = tokenizer.decode(outputs[0], skip_special_tokens=True)
74
  ```