Spaces:
Running
Running
Fixed file load with files including capital letters
Browse files- funcs/embeddings.py +1 -1
- funcs/helper_functions.py +4 -4
funcs/embeddings.py
CHANGED
|
@@ -18,7 +18,7 @@ def make_or_load_embeddings(docs, file_list, embeddings_out, embedding_model, em
|
|
| 18 |
if embeddings_out.size == 0:
|
| 19 |
print("Embeddings not found. Loading or generating new ones.")
|
| 20 |
|
| 21 |
-
embeddings_file_names = [string
|
| 22 |
|
| 23 |
if embeddings_file_names:
|
| 24 |
embeddings_file_name = embeddings_file_names[0]
|
|
|
|
| 18 |
if embeddings_out.size == 0:
|
| 19 |
print("Embeddings not found. Loading or generating new ones.")
|
| 20 |
|
| 21 |
+
embeddings_file_names = [string for string in file_list if "embedding" in string.lower()]
|
| 22 |
|
| 23 |
if embeddings_file_names:
|
| 24 |
embeddings_file_name = embeddings_file_names[0]
|
funcs/helper_functions.py
CHANGED
|
@@ -96,7 +96,7 @@ def initial_file_load(in_file):
|
|
| 96 |
|
| 97 |
file_list = [string.name for string in in_file]
|
| 98 |
|
| 99 |
-
data_file_names = [string
|
| 100 |
if data_file_names:
|
| 101 |
data_file_name = data_file_names[0]
|
| 102 |
df = read_file(data_file_name)
|
|
@@ -110,19 +110,19 @@ def initial_file_load(in_file):
|
|
| 110 |
print(error)
|
| 111 |
output_text = error
|
| 112 |
|
| 113 |
-
model_file_names = [string
|
| 114 |
if model_file_names:
|
| 115 |
model_file_name = model_file_names[0]
|
| 116 |
topic_model = read_file(model_file_name)
|
| 117 |
output_text = "Bertopic model loaded."
|
| 118 |
|
| 119 |
-
embedding_file_names = [string
|
| 120 |
if embedding_file_names:
|
| 121 |
embedding_file_name = embedding_file_names[0]
|
| 122 |
embeddings = read_file(embedding_file_name)
|
| 123 |
output_text = "Embeddings loaded."
|
| 124 |
|
| 125 |
-
label_file_names = [string
|
| 126 |
if label_file_names:
|
| 127 |
label_file_name = label_file_names[0]
|
| 128 |
custom_labels = read_file(label_file_name)
|
|
|
|
| 96 |
|
| 97 |
file_list = [string.name for string in in_file]
|
| 98 |
|
| 99 |
+
data_file_names = [string for string in file_list if "npz" not in string.lower() and "pkl" not in string.lower() and "topic_list.csv" not in string.lower()]
|
| 100 |
if data_file_names:
|
| 101 |
data_file_name = data_file_names[0]
|
| 102 |
df = read_file(data_file_name)
|
|
|
|
| 110 |
print(error)
|
| 111 |
output_text = error
|
| 112 |
|
| 113 |
+
model_file_names = [string for string in file_list if "pkl" in string.lower()]
|
| 114 |
if model_file_names:
|
| 115 |
model_file_name = model_file_names[0]
|
| 116 |
topic_model = read_file(model_file_name)
|
| 117 |
output_text = "Bertopic model loaded."
|
| 118 |
|
| 119 |
+
embedding_file_names = [string for string in file_list if "npz" in string.lower()]
|
| 120 |
if embedding_file_names:
|
| 121 |
embedding_file_name = embedding_file_names[0]
|
| 122 |
embeddings = read_file(embedding_file_name)
|
| 123 |
output_text = "Embeddings loaded."
|
| 124 |
|
| 125 |
+
label_file_names = [string for string in file_list if "topic_list" in string.lower()]
|
| 126 |
if label_file_names:
|
| 127 |
label_file_name = label_file_names[0]
|
| 128 |
custom_labels = read_file(label_file_name)
|