Update README.md
Browse files
README.md
CHANGED
@@ -49,22 +49,22 @@ inp = tokenizer("I am a boy </s> <2en>", add_special_tokens=False, return_tensor
|
|
49 |
|
50 |
# For generation. Pardon the messiness. Note the decoder_start_token_id.
|
51 |
model.eval() # Set dropouts to zero
|
52 |
-
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
53 |
# Decode to get output strings
|
54 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
55 |
print(decoded_output) # I am a boy
|
56 |
# Note that if your output language is not Hindi or Marathi, you should convert its script from Devanagari to the desired language using the Indic NLP Library.
|
57 |
# What if we mask?
|
58 |
inp = tokenizer("I am [MASK] </s> <2en>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
59 |
-
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
60 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
61 |
print(decoded_output) # I am happy
|
62 |
inp = tokenizer("मैं [MASK] हूठ</s> <2hi>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
63 |
-
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
64 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
65 |
print(decoded_output) # मैं जानता हूà¤
|
66 |
inp = tokenizer("मला [MASK] पाहिजे </s> <2mr>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
67 |
-
model_output=model.generate(inp, use_cache=True,
|
68 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
69 |
print(decoded_output) # मला ओळखलं पाहिजे
|
70 |
```
|
|
|
49 |
|
50 |
# For generation. Pardon the messiness. Note the decoder_start_token_id.
|
51 |
model.eval() # Set dropouts to zero
|
52 |
+
model_output=model.generate(inp, use_cache=True,no_repeat_ngram_size=3,encoder_no_repeat_ngram_size=3, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
53 |
# Decode to get output strings
|
54 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
55 |
print(decoded_output) # I am a boy
|
56 |
# Note that if your output language is not Hindi or Marathi, you should convert its script from Devanagari to the desired language using the Indic NLP Library.
|
57 |
# What if we mask?
|
58 |
inp = tokenizer("I am [MASK] </s> <2en>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
59 |
+
model_output=model.generate(inp, use_cache=True,no_repeat_ngram_size=3,encoder_no_repeat_ngram_size=3, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
60 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
61 |
print(decoded_output) # I am happy
|
62 |
inp = tokenizer("मैं [MASK] हूठ</s> <2hi>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
63 |
+
model_output=model.generate(inp, use_cache=True, num_beams=4,no_repeat_ngram_size=3,encoder_no_repeat_ngram_size=3, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
64 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
65 |
print(decoded_output) # मैं जानता हूà¤
|
66 |
inp = tokenizer("मला [MASK] पाहिजे </s> <2mr>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
|
67 |
+
model_output=model.generate(inp, use_cache=True,no_repeat_ngram_size=3,encoder_no_repeat_ngram_size=3,num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
|
68 |
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
|
69 |
print(decoded_output) # मला ओळखलं पाहिजे
|
70 |
```
|