Updating README
Browse files
README.md
CHANGED
|
@@ -51,7 +51,7 @@ model = AutoModelForSeq2SeqLM.from_pretrained(
|
|
| 51 |
)
|
| 52 |
tokenizer = AutoTokenizer.from_pretrained(model_path)
|
| 53 |
# MASKED LANGUAGE MODELING
|
| 54 |
-
sentence = "Ansiktsuttrykket [MASK_1]
|
| 55 |
encoding = tokenizer(sentence, return_tensors="pt")
|
| 56 |
mask_1 = tokenizer.convert_tokens_to_ids("[MASK_1]")
|
| 57 |
mask_2 = tokenizer.convert_tokens_to_ids("[MASK_2]")
|
|
@@ -61,7 +61,7 @@ output_tensor = model.generate(
|
|
| 61 |
eos_token_id=mask_2,
|
| 62 |
)
|
| 63 |
print(tokenizer.decode(output_tensor.squeeze(), skip_special_tokens=False))
|
| 64 |
-
# should output: '[MASK_1]
|
| 65 |
```
|
| 66 |
|
| 67 |
## Intermediate checkpoints
|
|
|
|
| 51 |
)
|
| 52 |
tokenizer = AutoTokenizer.from_pretrained(model_path)
|
| 53 |
# MASKED LANGUAGE MODELING
|
| 54 |
+
sentence = "Ansiktsuttrykket duckface har [MASK_1] seg til et utbredt kulturelt fenomen."
|
| 55 |
encoding = tokenizer(sentence, return_tensors="pt")
|
| 56 |
mask_1 = tokenizer.convert_tokens_to_ids("[MASK_1]")
|
| 57 |
mask_2 = tokenizer.convert_tokens_to_ids("[MASK_2]")
|
|
|
|
| 61 |
eos_token_id=mask_2,
|
| 62 |
)
|
| 63 |
print(tokenizer.decode(output_tensor.squeeze(), skip_special_tokens=False))
|
| 64 |
+
# should output: '[MASK_1]utviklet[MASK_2]'
|
| 65 |
```
|
| 66 |
|
| 67 |
## Intermediate checkpoints
|