Commit
•
ff77171
1
Parent(s):
a63edff
Update README.md
Browse files
README.md
CHANGED
|
@@ -94,7 +94,49 @@ outputs = model.generate(inputs.to("cuda:0"), max_new_tokens=40, do_sample=True,
|
|
| 94 |
pred = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
| 95 |
```
|
| 96 |
|
| 97 |
-
## Linking
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
|
| 99 |
## Information
|
| 100 |
|
|
|
|
| 94 |
pred = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
| 95 |
```
|
| 96 |
|
| 97 |
+
## Linking the predicted sequence of tokens to the corresponding ARASAAC pictograms
|
| 98 |
+
|
| 99 |
+
```python
|
| 100 |
+
import pandas as pd
|
| 101 |
+
|
| 102 |
+
def process_output_trad(pred):
|
| 103 |
+
return pred.split()
|
| 104 |
+
|
| 105 |
+
def read_lexicon(lexicon):
|
| 106 |
+
df = pd.read_csv(lexicon, sep='\t')
|
| 107 |
+
df['keyword_no_cat'] = df['lemma'].str.split(' #').str[0].str.strip().str.replace(' ', '_')
|
| 108 |
+
return df
|
| 109 |
+
|
| 110 |
+
def get_id_picto_from_predicted_lemma(df_lexicon, lemma):
|
| 111 |
+
id_picto = df_lexicon.loc[df_lexicon['keyword_no_cat'] == lemma, 'id_picto'].tolist()
|
| 112 |
+
return (id_picto[0], lemma) if id_picto else (0, lemma)
|
| 113 |
+
|
| 114 |
+
lexicon = read_lexicon("lexicon.csv")
|
| 115 |
+
sentence_to_map = process_output_trad(pred)
|
| 116 |
+
pictogram_ids = [get_id_picto_from_predicted_lemma(lexicon, lemma) for lemma in sentence_to_map]
|
| 117 |
+
```
|
| 118 |
+
|
| 119 |
+
## Viewing the predicted sequence of ARASAAC pictograms
|
| 120 |
+
|
| 121 |
+
```python
|
| 122 |
+
def generate_html(ids):
|
| 123 |
+
html_content = '<html><body>'
|
| 124 |
+
for picto_id, lemma in ids:
|
| 125 |
+
if picto_id != 0: # ignore invalid IDs
|
| 126 |
+
img_url = f"https://static.arasaac.org/pictograms/{picto_id}/{picto_id}_500.png"
|
| 127 |
+
html_content += f'''
|
| 128 |
+
<figure style="display:inline-block; margin:1px;">
|
| 129 |
+
<img src="{img_url}" alt="{lemma}" width="200" height="200" />
|
| 130 |
+
<figcaption>{lemma}</figcaption>
|
| 131 |
+
</figure>
|
| 132 |
+
'''
|
| 133 |
+
html_content += '</body></html>'
|
| 134 |
+
return html_content
|
| 135 |
+
|
| 136 |
+
html = generate_html(pictogram_ids)
|
| 137 |
+
with open("pictograms.html", "w") as file:
|
| 138 |
+
file.write(html)
|
| 139 |
+
```
|
| 140 |
|
| 141 |
## Information
|
| 142 |
|