Upload 8 files

LICENSE

@@ -0,0 +1,9 @@
+Apache License 2.0
+
+Copyright 2025 hmnshudhmn24
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0

README.md

@@ -1,3 +1,58 @@
----
-license: apache-2.0
----
+---
+language: en
+license: apache-2.0
+datasets: cnn_dailymail
+pipeline_tag: summarization
+library_name: transformers
+tags:
+  - t5
+  - summarization
+  - nlp
+  - text-generation
+base_model: t5-small
+---
+
+# 🧠 T5 News Summarizer
+
+A fine-tuned **T5-small** model trained on the **CNN/DailyMail dataset** for **news summarization**.
+This model converts long news articles into concise, readable summaries.
+
+---
+
+## 📊 Model Details
+
+- **Base model:** t5-small
+- **Dataset:** cnn_dailymail v3.0.0
+- **Task:** Summarization
+- **Language:** English
+- **Framework:** PyTorch
+
+---
+
+## 🚀 Usage
+
+```python
+from transformers import pipeline
+summarizer = pipeline("summarization", model="hmnshudhmn24/t5-news-summarizer")
+
+text = "The economy has seen a major shift due to advances in artificial intelligence..."
+print(summarizer(text))
+```
+
+---
+
+## 🧩 Example
+
+| Input | Output |
+|-------|---------|
+| "AI is transforming industries across the world..." | "AI is changing how industries operate globally." |
+
+---
+
+## ⚖️ License
+Licensed under the [Apache 2.0 License](./LICENSE).
+
+---
+
+## 🏷️ Tags
+`t5` `summarization` `nlp` `transformers` `cnn_dailymail`

config.json

@@ -0,0 +1,11 @@
+{
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "model_type": "t5",
+  "d_model": 512,
+  "num_heads": 8,
+  "num_layers": 6,
+  "vocab_size": 32128,
+  "decoder_start_token_id": 0
+}

inference.py

@@ -0,0 +1,9 @@
+from transformers import pipeline
+
+summarizer = pipeline("summarization", model="hmnshudhmn24/t5-news-summarizer")
+
+article = """The rapid development of artificial intelligence has raised questions about its impact on jobs and society.
+Experts believe AI will enhance productivity but may disrupt traditional industries.
+"""
+summary = summarizer(article, max_length=60, min_length=10, do_sample=False)
+print(summary)

requirements.txt

@@ -0,0 +1,4 @@
+transformers>=4.44.0
+datasets>=2.21.0
+torch>=2.2.0
+evaluate>=0.4.2

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "pad_token": "<pad>",
+  "eos_token": "</s>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,5 @@
+{
+  "model_max_length": 512,
+  "truncation_side": "right",
+  "padding_side": "right"
+}

train_summarizer.py

@@ -0,0 +1,59 @@
+from datasets import load_dataset
+from transformers import T5ForConditionalGeneration, T5TokenizerFast, Trainer, TrainingArguments
+import evaluate
+import numpy as np
+
+# Load dataset
+dataset = load_dataset("cnn_dailymail", "3.0.0")
+
+# Load tokenizer and model
+tokenizer = T5TokenizerFast.from_pretrained("t5-small")
+model = T5ForConditionalGeneration.from_pretrained("t5-small")
+
+# Preprocess function
+def preprocess_function(examples):
+    inputs = ["summarize: " + doc for doc in examples["article"]]
+    model_inputs = tokenizer(inputs, max_length=512, truncation=True)
+    labels = tokenizer(text_target=examples["highlights"], max_length=128, truncation=True)
+    model_inputs["labels"] = labels["input_ids"]
+    return model_inputs
+
+tokenized_datasets = dataset.map(preprocess_function, batched=True, remove_columns=["article", "highlights", "id"])
+
+# Metrics
+rouge = evaluate.load("rouge")
+
+def compute_metrics(eval_pred):
+    predictions, labels = eval_pred
+    decoded_preds = tokenizer.batch_decode(predictions, skip_special_tokens=True)
+    labels = np.where(labels != -100, labels, tokenizer.pad_token_id)
+    decoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=True)
+    result = rouge.compute(predictions=decoded_preds, references=decoded_labels)
+    return {k: v * 100 for k, v in result.items()}
+
+# Training arguments
+training_args = TrainingArguments(
+    output_dir="./results",
+    evaluation_strategy="epoch",
+    learning_rate=3e-4,
+    per_device_train_batch_size=2,
+    per_device_eval_batch_size=2,
+    num_train_epochs=1,
+    save_strategy="epoch",
+    predict_with_generate=True,
+    push_to_hub=False
+)
+
+# Trainer
+trainer = Trainer(
+    model=model,
+    args=training_args,
+    train_dataset=tokenized_datasets["train"].select(range(2000)),
+    eval_dataset=tokenized_datasets["validation"].select(range(500)),
+    tokenizer=tokenizer,
+    compute_metrics=compute_metrics
+)
+
+trainer.train()
+trainer.save_model("./t5-news-summarizer")
+tokenizer.save_pretrained("./t5-news-summarizer")