Hugging Face's logo

ai4bharat
/
indictrans2-en-indic-dist-200M

Translation
Transformers
PyTorch
Safetensors
IndicTrans
text2text-generation
indictrans2
ai4bharat
multilingual
custom_code
Model card Files Community
7
indictrans2-en-indic-dist-200M
File size: 3,855 Bytes 
f4a20bf
ebefd9d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f4a20bf
ebefd9d
 
 
 
 
 
 
 
 
 
f4a20bf
ebefd9d
 
 
 
 
6fcb311
ebefd9d
 
 
 
6827d3d
 
 
 
 
 
 
 
 
3958b6a
6827d3d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0826f8f
ebefd9d
 
 
 
 
 
37f9d74
 
 
 
 
 
 
 
ebefd9d
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
 
---
language:
- as
- bn
- brx
- doi
- en
- gom
- gu
- hi
- kn
- ks
- kas
- mai
- ml
- mr
- mni
- mnb
- ne
- or
- pa
- sa
- sat
- sd
- snd
- ta
- te
- ur
language_details: >-
  asm_Beng, ben_Beng, brx_Deva, doi_Deva, eng_Latn, gom_Deva, guj_Gujr,
  hin_Deva, kan_Knda, kas_Arab, kas_Deva, mai_Deva, mal_Mlym, mar_Deva,
  mni_Beng, mni_Mtei, npi_Deva, ory_Orya, pan_Guru, san_Deva, sat_Olck,
  snd_Arab, snd_Deva, tam_Taml, tel_Telu, urd_Arab
tags:
- indictrans2
- translation
- ai4bharat
- multilingual
license: mit
datasets:
- flores-200
- IN22-Gen
- IN22-Conv
metrics:
- bleu
- chrf
- chrf++
- comet
inference: false
---

# IndicTrans2

This is the model card of IndicTrans2 En-Indic Distilled 200M variant.

Please refer to [section 7.6: Distilled Models](https://openreview.net/forum?id=vfT4YuzAYA) in the TMLR submission for further details on model training, data and metrics.

### Usage Instructions
Please refer to the [github repository](https://github.com/AI4Bharat/IndicTrans2/tree/main/huggingface_inference) for a detail description on how to use HF compatible IndicTrans2 models for inference.

```python
import torch
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
from IndicTransTokenizer import IndicProcessor


model_name = "ai4bharat/indictrans2-en-indic-dist-200M"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)

model = AutoModelForSeq2SeqLM.from_pretrained(model_name, trust_remote_code=True)

ip = IndicProcessor(inference=True)

input_sentences = [
    "When I was young, I used to go to the park every day.",
    "We watched a new movie last week, which was very inspiring.",
    "If you had met me at that time, we would have gone out to eat.",
    "My friend has invited me to his birthday party, and I will give him a gift.",
]

src_lang, tgt_lang = "eng_Latn", "hin_Deva"

batch = ip.preprocess_batch(input_sentences, src_lang=src_lang, tgt_lang=tgt_lang)

DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

# Tokenize the sentences and generate input encodings
inputs = tokenizer(
    batch,
    truncation=True,
    padding="longest",
    return_tensors="pt",
    return_attention_mask=True,
).to(DEVICE)

# Generate translations using the model
with torch.no_grad():
    generated_tokens = model.generate(
        **inputs,
        use_cache=True,
        min_length=0,
        max_length=256,
        num_beams=5,
        num_return_sequences=1,
    )

# Decode the generated tokens into text
with tokenizer.as_target_tokenizer():
    generated_tokens = tokenizer.batch_decode(
        generated_tokens.detach().cpu().tolist(),
        skip_special_tokens=True,
        clean_up_tokenization_spaces=True,
    )

# Postprocess the translations, including entity replacement
translations = ip.postprocess_batch(generated_tokens, lang=tgt_lang)

for input_sentence, translation in zip(input_sentences, translations):
    print(f"{src_lang}: {input_sentence}")
    print(f"{tgt_lang}: {translation}")
```

**Note: IndicTrans2 is now compatible with AutoTokenizer, however you need to use IndicProcessor from [IndicTransTokenizer](https://github.com/VarunGumma/IndicTransTokenizer) for preprocessing before tokenization.**



### Citation

If you consider using our work then please cite using:

```
@article{gala2023indictrans,
title={IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages},
author={Jay Gala and Pranjal A Chitale and A K Raghavan and Varun Gumma and Sumanth Doddapaneni and Aswanth Kumar M and Janki Atul Nawale and Anupama Sujatha and Ratish Puduppully and Vivek Raghavan and Pratyush Kumar and Mitesh M Khapra and Raj Dabre and Anoop Kunchukuttan},
journal={Transactions on Machine Learning Research},
issn={2835-8856},
year={2023},
url={https://openreview.net/forum?id=vfT4YuzAYA},
note={}
}
```