Spaces:
Sleeping
Sleeping
bug fix
Browse files
app.py
CHANGED
@@ -9,23 +9,23 @@ from pathlib import Path
|
|
9 |
|
10 |
import gradio as gr
|
11 |
|
12 |
-
from finetune import finetune_model
|
13 |
|
14 |
from language import languages
|
15 |
from task import tasks
|
16 |
import matplotlib.pyplot as plt
|
17 |
|
18 |
|
19 |
-
os.environ['TEMP_DIR'] = tempfile.mkdtemp()
|
20 |
-
|
21 |
def load_markdown():
|
22 |
with open("intro.md", "r") as f:
|
23 |
return f.read()
|
24 |
|
25 |
|
26 |
-
def read_logs():
|
|
|
|
|
27 |
try:
|
28 |
-
with open(f"output.log", "r") as f:
|
29 |
return f.read()
|
30 |
except:
|
31 |
return None
|
@@ -34,7 +34,10 @@ def read_logs():
|
|
34 |
def plot_loss_acc(temp_dir, log_every):
|
35 |
sys.stdout.flush()
|
36 |
lines = []
|
37 |
-
|
|
|
|
|
|
|
38 |
for line in f.readlines():
|
39 |
if re.match(r"^\[\d+\] - loss: \d+\.\d+ - acc: \d+\.\d+$", line):
|
40 |
lines.append(line)
|
@@ -68,22 +71,28 @@ def upload_file(fileobj, temp_dir):
|
|
68 |
"""
|
69 |
# First check if a file is a zip file.
|
70 |
if not zipfile.is_zipfile(fileobj.name):
|
|
|
71 |
raise gr.Error("Please upload a zip file.")
|
72 |
|
73 |
# Then unzip file
|
|
|
74 |
shutil.unpack_archive(fileobj.name, temp_dir)
|
75 |
|
76 |
# check zip file
|
77 |
if not os.path.exists(os.path.join(temp_dir, "text")):
|
|
|
78 |
raise gr.Error("Please upload a valid zip file.")
|
79 |
|
80 |
if not os.path.exists(os.path.join(temp_dir, "text_ctc")):
|
|
|
81 |
raise gr.Error("Please upload a valid zip file.")
|
82 |
|
83 |
if not os.path.exists(os.path.join(temp_dir, "audio")):
|
|
|
84 |
raise gr.Error("Please upload a valid zip file.")
|
85 |
|
86 |
# check if all texts and audio matches
|
|
|
87 |
audio_ids = []
|
88 |
with open(os.path.join(temp_dir, "text"), "r") as f:
|
89 |
for line in f.readlines():
|
@@ -100,25 +109,39 @@ def upload_file(fileobj, temp_dir):
|
|
100 |
)
|
101 |
|
102 |
if set(audio_ids) != set(ctc_audio_ids):
|
|
|
103 |
raise gr.Error(f"`text` and `text_ctc` have different audio ids.")
|
104 |
|
105 |
for audio_id in glob.glob(os.path.join(temp_dir, "audio", "*")):
|
106 |
if not Path(audio_id).stem in audio_ids:
|
107 |
raise gr.Error(f"Audio id {audio_id} is not in `text` or `text_ctc`.")
|
108 |
|
|
|
109 |
gr.Info("Successfully uploaded and validated zip file.")
|
110 |
|
111 |
return [fileobj]
|
112 |
|
113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
with gr.Blocks(title="OWSM-finetune") as demo:
|
115 |
-
tempdir_path
|
116 |
gr.Markdown(
|
117 |
"""# OWSM finetune demo!
|
118 |
-
|
119 |
Finetune `owsm_v3.1_ebf_base` with your own dataset!
|
120 |
Due to resource limitation, you can only train 10 epochs on maximum.
|
121 |
-
|
122 |
## Upload dataset and define settings
|
123 |
"""
|
124 |
)
|
@@ -153,7 +176,7 @@ Due to resource limitation, you can only train 10 epochs on maximum.
|
|
153 |
with gr.Row():
|
154 |
with gr.Column():
|
155 |
log_every = gr.Number(value=10, label="log_every", interactive=True)
|
156 |
-
max_epoch = gr.Slider(1,
|
157 |
scheduler = gr.Dropdown(
|
158 |
["warmuplr"], label="warmup", value="warmuplr", interactive=True
|
159 |
)
|
@@ -185,7 +208,7 @@ Due to resource limitation, you can only train 10 epochs on maximum.
|
|
185 |
max_lines=23,
|
186 |
lines=23,
|
187 |
)
|
188 |
-
demo.load(read_logs,
|
189 |
|
190 |
with gr.Column():
|
191 |
log_acc = gr.Image(label="Accuracy", show_label=True, interactive=False)
|
@@ -241,7 +264,7 @@ Due to resource limitation, you can only train 10 epochs on maximum.
|
|
241 |
learning_rate,
|
242 |
weight_decay,
|
243 |
],
|
244 |
-
[trained_model, hyp_text]
|
245 |
)
|
246 |
|
247 |
gr.Markdown(load_markdown())
|
|
|
9 |
|
10 |
import gradio as gr
|
11 |
|
12 |
+
from finetune import finetune_model, log
|
13 |
|
14 |
from language import languages
|
15 |
from task import tasks
|
16 |
import matplotlib.pyplot as plt
|
17 |
|
18 |
|
|
|
|
|
19 |
def load_markdown():
|
20 |
with open("intro.md", "r") as f:
|
21 |
return f.read()
|
22 |
|
23 |
|
24 |
+
def read_logs(temp_dir):
|
25 |
+
if not os.path.exists(f"{temp_dir}/output.log"):
|
26 |
+
return "Log file not found."
|
27 |
try:
|
28 |
+
with open(f"{temp_dir}/output.log", "r") as f:
|
29 |
return f.read()
|
30 |
except:
|
31 |
return None
|
|
|
34 |
def plot_loss_acc(temp_dir, log_every):
|
35 |
sys.stdout.flush()
|
36 |
lines = []
|
37 |
+
if not os.path.exists(f"{temp_dir}/output.log"):
|
38 |
+
return None, None
|
39 |
+
|
40 |
+
with open(f"{temp_dir}/output.log", "r") as f:
|
41 |
for line in f.readlines():
|
42 |
if re.match(r"^\[\d+\] - loss: \d+\.\d+ - acc: \d+\.\d+$", line):
|
43 |
lines.append(line)
|
|
|
71 |
"""
|
72 |
# First check if a file is a zip file.
|
73 |
if not zipfile.is_zipfile(fileobj.name):
|
74 |
+
log(temp_dir, "Please upload a zip file.")
|
75 |
raise gr.Error("Please upload a zip file.")
|
76 |
|
77 |
# Then unzip file
|
78 |
+
log(temp_dir, "Unzipping file...")
|
79 |
shutil.unpack_archive(fileobj.name, temp_dir)
|
80 |
|
81 |
# check zip file
|
82 |
if not os.path.exists(os.path.join(temp_dir, "text")):
|
83 |
+
log(temp_dir, "Please upload a valid zip file.")
|
84 |
raise gr.Error("Please upload a valid zip file.")
|
85 |
|
86 |
if not os.path.exists(os.path.join(temp_dir, "text_ctc")):
|
87 |
+
log(temp_dir, "Please upload a valid zip file.")
|
88 |
raise gr.Error("Please upload a valid zip file.")
|
89 |
|
90 |
if not os.path.exists(os.path.join(temp_dir, "audio")):
|
91 |
+
log(temp_dir, "Please upload a valid zip file.")
|
92 |
raise gr.Error("Please upload a valid zip file.")
|
93 |
|
94 |
# check if all texts and audio matches
|
95 |
+
log(temp_dir, "Checking if all texts and audio matches...")
|
96 |
audio_ids = []
|
97 |
with open(os.path.join(temp_dir, "text"), "r") as f:
|
98 |
for line in f.readlines():
|
|
|
109 |
)
|
110 |
|
111 |
if set(audio_ids) != set(ctc_audio_ids):
|
112 |
+
log(temp_dir, f"`text` and `text_ctc` have different audio ids.")
|
113 |
raise gr.Error(f"`text` and `text_ctc` have different audio ids.")
|
114 |
|
115 |
for audio_id in glob.glob(os.path.join(temp_dir, "audio", "*")):
|
116 |
if not Path(audio_id).stem in audio_ids:
|
117 |
raise gr.Error(f"Audio id {audio_id} is not in `text` or `text_ctc`.")
|
118 |
|
119 |
+
log(temp_dir, "Successfully uploaded and validated zip file.")
|
120 |
gr.Info("Successfully uploaded and validated zip file.")
|
121 |
|
122 |
return [fileobj]
|
123 |
|
124 |
|
125 |
+
def delete_tmp_dir(tmp_dir):
|
126 |
+
if os.path.exists(tmp_dir):
|
127 |
+
shutil.rmtree(tmp_dir)
|
128 |
+
print(f"Deleted temporary directory: {tmp_dir}")
|
129 |
+
else:
|
130 |
+
print("Temporary directory already deleted")
|
131 |
+
|
132 |
+
|
133 |
+
def create_tmp_dir():
|
134 |
+
tmp_dir = tempfile.mkdtemp()
|
135 |
+
print(f"Created temporary directory: {tmp_dir}")
|
136 |
+
return tmp_dir
|
137 |
+
|
138 |
+
|
139 |
with gr.Blocks(title="OWSM-finetune") as demo:
|
140 |
+
tempdir_path=gr.State(create_tmp_dir, delete_callback=delete_tmp_dir, time_to_live=600)
|
141 |
gr.Markdown(
|
142 |
"""# OWSM finetune demo!
|
|
|
143 |
Finetune `owsm_v3.1_ebf_base` with your own dataset!
|
144 |
Due to resource limitation, you can only train 10 epochs on maximum.
|
|
|
145 |
## Upload dataset and define settings
|
146 |
"""
|
147 |
)
|
|
|
176 |
with gr.Row():
|
177 |
with gr.Column():
|
178 |
log_every = gr.Number(value=10, label="log_every", interactive=True)
|
179 |
+
max_epoch = gr.Slider(1, 30, step=1, label="max_epoch", interactive=True)
|
180 |
scheduler = gr.Dropdown(
|
181 |
["warmuplr"], label="warmup", value="warmuplr", interactive=True
|
182 |
)
|
|
|
208 |
max_lines=23,
|
209 |
lines=23,
|
210 |
)
|
211 |
+
demo.load(read_logs, [tempdir_path], log_output, every=2)
|
212 |
|
213 |
with gr.Column():
|
214 |
log_acc = gr.Image(label="Accuracy", show_label=True, interactive=False)
|
|
|
264 |
learning_rate,
|
265 |
weight_decay,
|
266 |
],
|
267 |
+
[trained_model, ref_text, base_text, hyp_text]
|
268 |
)
|
269 |
|
270 |
gr.Markdown(load_markdown())
|