madlad400-3b-mt / README.md
jbochi's picture
Fix tokenizer path
7504c16
|
raw
history blame
993 Bytes
---
license: apache-2.0
language:
- en
- pt
- es
- de
- it
library_name: transformers
tags:
- text-generation-inference
---
T5ForConditionalGeneration files for the [Madlad-400](https://github.com/google-research/google-research/tree/master/madlad_400) 3B parameter MT model.
```python
from transformers import T5ForConditionalGeneration, T5Tokenizer, GenerationConfig
model = T5ForConditionalGeneration.from_pretrained('jbochi/madlad400-3b-mt')
tokenizer = T5Tokenizer.from_pretrained('jbochi/madlad400-3b-mt')
text = "<2es> how do you say torch in portuguese?"
input_ids = tokenizer(text, return_tensors="pt").input_ids
outputs = model.generate(
input_ids=input_ids,
generation_config=GenerationConfig(
decoder_start_token_id=2,
))
tokenizer.decode(outputs[0], skip_special_tokens=True)
# como se dice antorcha en portugués?
```
Colab to generate these files is [here](https://colab.research.google.com/drive/1rZ2NRyl2zwmg0sQ2Wi-uZZF48iVYulTC#scrollTo=pVODoE6gA9sw).