README.md

---
license: cc-by-nc-nd-4.0
dataset_info:
- config_name: default
  features:
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 678202759.44
    num_examples: 27000
  - name: val
    num_bytes: 75355862.16
    num_examples: 3000
  - name: test
    num_bytes: 188389655.4
    num_examples: 7500
  download_size: 520788766
  dataset_size: 941948277.0
- config_name: email-corpus
  features:
  - name: file
    dtype: string
  - name: message
    dtype: string
  splits:
  - name: train
    num_bytes: 1424661264
    num_examples: 517401
  download_size: 606473179
  dataset_size: 1424661264
- config_name: indic-corpus
  features:
  - name: lang_id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 242818701
    num_examples: 11
  download_size: 105665834
  dataset_size: 242818701
- config_name: wiki-topics
  features:
  - name: article
    dtype: string
  - name: category
    dtype: string
  - name: text
    dtype: string
  - name: __index_level_0__
    dtype: int64
  splits:
  - name: train
    num_bytes: 338290047
    num_examples: 8000
  - name: test
    num_bytes: 84610785
    num_examples: 2000
  download_size: 236620494
  dataset_size: 422900832
- config_name: Assignment-3-word2vec
  features:
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 59964905
    num_examples: 200000
  download_size: 59964905
  dataset_size: 59964905
- config_name: Assignment-3-word2vec-analogy
  features:
  - name: word1
    dtype: string
  - name: word2
    dtype: string
  - name: word3
    dtype: string
  splits:
  - name: test
    num_bytes: 25776
    num_examples: 1000
  download_size: 25776
  dataset_size: 25776
- config_name: Assignment-3-naive-bayes
  features:
  - name: text
    dtype: string
  - name: category
    dtype: string
  splits:
  - name: train
    num_bytes: 690247310
    num_examples: 8000
  - name: test
    num_bytes: 171875690
    num_examples: 2000
  download_size: 862123000
  dataset_size: 862123000
- config_name: Assignment-3-em
  features:
  - name: text
    dtype: string
  - name: category
    dtype: string
  splits:
  - name: train
    num_bytes: 683095685
    num_examples: 8000
  - name: test
    num_bytes: 171875690
    num_examples: 2000
  download_size: 854971375
  dataset_size: 854971375
- config_name: Assignment-4
  features:
    - name: id
      dtype: string
    - name: tokens
      sequence: string
    - name: ner_tags
      sequence: int32
  splits:
    - name: train
      num_bytes: 38232055
      num_examples: 75827
    - name: val
      num_bytes: 5464732
      num_examples: 10851
    - name: test
      num_bytes: 10939918
      num_examples: 21657
  download_size: 54636705
  dataset_size: 54636705
- config_name: Deep-learning-assignment
  features:
    - name: Category
      dtype: string
    - name: Description
      sequence: string
  splits:
    - name: train
      num_bytes: 15089359
      num_examples: 120000
    - name: test
      num_bytes: 1014560
      num_examples: 7600
  download_size: 16103919
  dataset_size: 16103919
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
  - split: test
    path: data/test-*
- config_name: email-corpus
  data_files:
  - split: train
    path: email-corpus/train-*
- config_name: indic-corpus
  data_files:
  - split: train
    path: indic-corpus/train-*
- config_name: wiki-topics
  data_files:
  - split: train
    path: wiki-topics/train-*
  - split: test
    path: wiki-topics/test-*
- config_name: Assignment-3-word2vec
  data_files:
  - split: train
    path: Assignment-3/word2vec/train*
- config_name: Assignment-3-word2vec-analogy
  data_files:
  - split: test
    path: Assignment-3/word2vec/test*
- config_name: Assignment-3-naive-bayes
  data_files:
  - split: train
    path: Assignment-3/naive_bayes/train*
  - split: test
    path: Assignment-3/naive_bayes/test_nb_with_labels*
- config_name: Assignment-3-em
  data_files:
  - split: train
    path: Assignment-3/em/train*
  - split: test
    path: Assignment-3/em/test*
- config_name: Assignment-4
  data_files:
  - split: train
    path: Assignment-4/train*
  - split: test
    path: Assignment-4/test*
  - split: val
    path: Assignment-4/val*

- config_name: Deep-learning-assignment
  data_files:
  - split: train
    path: Deep-learning-assignment/train*
  - split: test
    path: Deep-learning-assignment/test*
---
# CS779-Fall 2025 IIT-Kanpur
Instructor: Dr. Ashutosh Modi

## Assignment-3
There are 3 main tasks in Assignment-3:
1. Neural Network Implementation from Scratch for Word2Vec using Wikipedia Text
2. Naive Bayes Classifier for Topic classification on Wikipedia Articles
3. Expectation-Maximization Based clustering on Wikipedia Articles

The data can be fetched using the datasets API as follows:
```python
from datasets import load_dataset

# Word2Vec Dataset
word2vec_train = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-word2vec", split="train")
word2vec_test = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-word2vec-analogy", split="test")

# Naive Bayes
naive_bayes_train = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-naive-bayes", split="train")
naive_bayes_test = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-naive-bayes", split="test")

# Expectation-Maximization
em_train = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-em", split="train")
em_test = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-3-em", split="test")

```

## Assignment-4
This assignment involves Named Entity Recognition (NER) on a Hindi dataset (NED) on a custom dataset. The dataset consists of sentences with tokens and their corresponding NER tags. The list of NER tags includes:
- B-FESTIVAL
- B-GAME
- B-LANGUAGE
- B-LITERATURE
- B-LOCATION
- B-MISC
- B-NUMEX
- B-ORGANIZATION
- B-PERSON
- B-RELIGION
- B-TIMEX
- I-FESTIVAL
- I-GAME
- I-LANGUAGE
- I-LITERATURE
- I-LOCATION
- I-MISC
- I-NUMEX
- I-ORGANIZATION
- I-PERSON
- I-RELIGION
- I-TIMEX
- O

The data can be fetched using the datasets API as follows:
```python
from datasets import load_dataset

# Train
train = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-4", split="train")
# Test
test = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-4", split="test")
# Validation
val = load_dataset("Exploration-Lab/CS779-Fall25", "Assignment-4", split="val")

```


## Deep Learning Assignment
This assignment involves text classification on a dataset containing product descriptions and their corresponding categories. The dataset consists of two columns: "Category" and "Description". The "Category" column contains the category labels, while the "Description" column contains the product descriptions. The data can be fetched using the datasets API as follows:
```python
from datasets import load_dataset

# Train
train = load_dataset("Exploration-Lab/CS779-Fall25", "Deep-learning-assignment", split="train")
# Test
test = load_dataset("Exploration-Lab/CS779-Fall25", "Deep-learning-assignment", split="test")

```