Update README.md
Browse files
README.md
CHANGED
@@ -85,6 +85,9 @@ if input_ids != None:
|
|
85 |
|
86 |
print("Updated max_len = " + str(max_len))
|
87 |
|
|
|
|
|
|
|
88 |
sample_outputs = model.generate(
|
89 |
input_ids,
|
90 |
do_sample=True,
|
@@ -94,13 +97,18 @@ sample_outputs = model.generate(
|
|
94 |
num_return_sequences=sample_output_num
|
95 |
)
|
96 |
|
97 |
-
print(100 * '-' + "\
|
98 |
-
Output:\
|
99 |
-
" + 100 * '-')
|
100 |
for i, sample_output in enumerate(sample_outputs):
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
|
106 |
```
|
|
|
85 |
|
86 |
print("Updated max_len = " + str(max_len))
|
87 |
|
88 |
+
stop_token = "<|endoftext|>"
|
89 |
+
new_lines = "\n\n\n"
|
90 |
+
|
91 |
sample_outputs = model.generate(
|
92 |
input_ids,
|
93 |
do_sample=True,
|
|
|
97 |
num_return_sequences=sample_output_num
|
98 |
)
|
99 |
|
100 |
+
print(100 * '-' + "\n\t\tOutput\n" + 100 * '-')
|
|
|
|
|
101 |
for i, sample_output in enumerate(sample_outputs):
|
102 |
+
|
103 |
+
text = tokenizer.decode(sample_output, skip_special_tokens=True)
|
104 |
+
|
105 |
+
# Remove all text after the stop token
|
106 |
+
text = text[: text.find(stop_token) if stop_token else None]
|
107 |
+
|
108 |
+
# Remove all text after 3 newlines
|
109 |
+
text = text[: text.find(new_lines) if new_lines else None]
|
110 |
+
|
111 |
+
print("\n{}: {}".format(i, text))
|
112 |
+
print("\n" + 100 * '-')
|
113 |
|
114 |
```
|