Update README.md
Browse files
README.md
CHANGED
|
@@ -7,30 +7,67 @@ library_name: transformers
|
|
| 7 |
|
| 8 |
|
| 9 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
|
|
|
|
|
|
|
|
|
|
| 12 |
from transformers import BartTokenizer, BartForConditionalGeneration
|
| 13 |
from datasets import load_dataset
|
| 14 |
|
| 15 |
-
|
| 16 |
tokenizer = BartTokenizer.from_pretrained('ayjays132/EnhancerModel')
|
| 17 |
model = BartForConditionalGeneration.from_pretrained('ayjays132/EnhancerModel')
|
| 18 |
|
| 19 |
-
|
| 20 |
dataset = load_dataset("cnn_dailymail", "3.0.0")
|
| 21 |
|
| 22 |
-
|
| 23 |
def summarize(text):
|
| 24 |
inputs = tokenizer(text, return_tensors="pt", max_length=1024, truncation=True)
|
| 25 |
summary_ids = model.generate(inputs['input_ids'], max_length=150, min_length=40, length_penalty=2.0, num_beams=4, early_stopping=True)
|
| 26 |
return tokenizer.decode(summary_ids[0], skip_special_tokens=True)
|
| 27 |
|
| 28 |
-
|
| 29 |
print("Type of dataset['test']:", type(dataset['test']))
|
| 30 |
print("Type of the first element in dataset['test']:", type(dataset['test'][0]))
|
| 31 |
print("Content of the first element in dataset['test']:", dataset['test'][0])
|
| 32 |
|
| 33 |
-
|
| 34 |
for example in dataset['test'][:5]:
|
| 35 |
try:
|
| 36 |
# If the example is a string, then it's likely that 'dataset['test']' is not loaded as expected
|
|
@@ -47,3 +84,5 @@ for example in dataset['test'][:5]:
|
|
| 47 |
print("No 'article' field found in this example.")
|
| 48 |
except Exception as e:
|
| 49 |
print(f"Error processing example: {e}")
|
|
|
|
|
|
|
|
|
| 7 |
|
| 8 |
|
| 9 |
|
| 10 |
+
<style>
|
| 11 |
+
body {
|
| 12 |
+
font-family: Arial, sans-serif;
|
| 13 |
+
line-height: 1.6;
|
| 14 |
+
margin: 0;
|
| 15 |
+
padding: 20px;
|
| 16 |
+
background-color: #f4f4f4;
|
| 17 |
+
}
|
| 18 |
+
.model-description {
|
| 19 |
+
background: #fff;
|
| 20 |
+
border-radius: 8px;
|
| 21 |
+
padding: 20px;
|
| 22 |
+
box-shadow: 0 0 10px rgba(0,0,0,0.1);
|
| 23 |
+
max-width: 800px;
|
| 24 |
+
margin: auto;
|
| 25 |
+
}
|
| 26 |
+
.model-description h2 {
|
| 27 |
+
margin-top: 0;
|
| 28 |
+
color: #333;
|
| 29 |
+
}
|
| 30 |
+
.model-description p {
|
| 31 |
+
margin: 0 0 10px;
|
| 32 |
+
color: #666;
|
| 33 |
+
}
|
| 34 |
+
</style>
|
| 35 |
+
</head>
|
| 36 |
+
<body>
|
| 37 |
|
| 38 |
+
<div class="model-description">
|
| 39 |
+
<h2>Enhanced Model Features</h2>
|
| 40 |
+
<p><strong>Adaptability:</strong> Adjusts to diverse contexts and user needs, ensuring relevant and precise interactions.</p>
|
| 41 |
+
<p><strong>Contextual Intelligence:</strong> Provides contextually aware responses, improving engagement and interaction quality.</p>
|
| 42 |
+
<p><strong>Advanced Algorithms:</strong> Employs cutting-edge algorithms for sophisticated and intelligent responses.</p>
|
| 43 |
+
<p><strong>User Experience:</strong> Designed with a focus on seamless interaction, offering an intuitive and refined user experience.</p>
|
| 44 |
+
</div>
|
| 45 |
|
| 46 |
+
|
| 47 |
+
<div class="code-container">
|
| 48 |
+
<code>
|
| 49 |
from transformers import BartTokenizer, BartForConditionalGeneration
|
| 50 |
from datasets import load_dataset
|
| 51 |
|
| 52 |
+
Load pre-trained BART model for summarization
|
| 53 |
tokenizer = BartTokenizer.from_pretrained('ayjays132/EnhancerModel')
|
| 54 |
model = BartForConditionalGeneration.from_pretrained('ayjays132/EnhancerModel')
|
| 55 |
|
| 56 |
+
Load dataset
|
| 57 |
dataset = load_dataset("cnn_dailymail", "3.0.0")
|
| 58 |
|
| 59 |
+
Function to generate summary
|
| 60 |
def summarize(text):
|
| 61 |
inputs = tokenizer(text, return_tensors="pt", max_length=1024, truncation=True)
|
| 62 |
summary_ids = model.generate(inputs['input_ids'], max_length=150, min_length=40, length_penalty=2.0, num_beams=4, early_stopping=True)
|
| 63 |
return tokenizer.decode(summary_ids[0], skip_special_tokens=True)
|
| 64 |
|
| 65 |
+
ebugging: Print the type and content of the first example
|
| 66 |
print("Type of dataset['test']:", type(dataset['test']))
|
| 67 |
print("Type of the first element in dataset['test']:", type(dataset['test'][0]))
|
| 68 |
print("Content of the first element in dataset['test']:", dataset['test'][0])
|
| 69 |
|
| 70 |
+
Test the model on a few examples
|
| 71 |
for example in dataset['test'][:5]:
|
| 72 |
try:
|
| 73 |
# If the example is a string, then it's likely that 'dataset['test']' is not loaded as expected
|
|
|
|
| 84 |
print("No 'article' field found in this example.")
|
| 85 |
except Exception as e:
|
| 86 |
print(f"Error processing example: {e}")
|
| 87 |
+
</code>
|
| 88 |
+
</div>
|