Update README.md
Browse files
README.md
CHANGED
@@ -13,3 +13,44 @@ tags:
|
|
13 |
- flan
|
14 |
---
|
15 |

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
- flan
|
14 |
---
|
15 |

|
16 |
+
# **t5-Flan-Prompt-Enhance**
|
17 |
+
T5-Flan-Prompt-Enhance is a fine-tuned model based on **Flan-T5-Small**, specifically designed to **enhance prompts, captions, and annotations**. This means the model has been further trained to improve the quality, clarity, and richness of textual inputs, making them more detailed and expressive.
|
18 |
+
|
19 |
+
### Key Features:
|
20 |
+
1. **Prompt Expansion** – Takes short or vague prompts and enriches them with more context, depth, and specificity.
|
21 |
+
2. **Caption Enhancement** – Improves captions by adding more descriptive details, making them more informative and engaging.
|
22 |
+
3. **Annotation Refinement** – Enhances annotations by making them clearer, more structured, and contextually relevant.
|
23 |
+
|
24 |
+
### Run with Transformers
|
25 |
+
|
26 |
+
```python
|
27 |
+
from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM
|
28 |
+
import torch
|
29 |
+
|
30 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
31 |
+
|
32 |
+
# Model checkpoint
|
33 |
+
model_checkpoint = "prithivMLmods/t5-Flan-Prompt-Enhance"
|
34 |
+
|
35 |
+
# Tokenizer
|
36 |
+
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
|
37 |
+
|
38 |
+
# Model
|
39 |
+
model = AutoModelForSeq2SeqLM.from_pretrained(model_checkpoint)
|
40 |
+
|
41 |
+
enhancer = pipeline('text2text-generation',
|
42 |
+
model=model,
|
43 |
+
tokenizer=tokenizer,
|
44 |
+
repetition_penalty=1.2,
|
45 |
+
device=0 if device == "cuda" else -1)
|
46 |
+
|
47 |
+
max_target_length = 256
|
48 |
+
prefix = "enhance prompt: "
|
49 |
+
|
50 |
+
short_prompt = "three chimneys on the roof, green trees and shrubs in front of the house"
|
51 |
+
answer = enhancer(prefix + short_prompt, max_length=max_target_length)
|
52 |
+
final_answer = answer[0]['generated_text']
|
53 |
+
print(final_answer)
|
54 |
+
```
|
55 |
+
|
56 |
+
This fine-tuning process allows **T5-Flan-Prompt-Enhance** to generate **high-quality, well-structured, and contextually relevant outputs**, which can be particularly useful for tasks such as text generation, content creation, and AI-assisted writing.
|