Update app.py
Browse files
app.py
CHANGED
@@ -38,30 +38,52 @@ def load_model():
|
|
38 |
logger.error(f"Failed to load model: {e}")
|
39 |
return False
|
40 |
|
41 |
-
def get_card_info(hub_id: str) -> Tuple[str, str]:
|
42 |
"""Get card information from a Hugging Face hub_id."""
|
43 |
model_exists = False
|
44 |
dataset_exists = False
|
45 |
model_text = None
|
46 |
dataset_text = None
|
47 |
|
48 |
-
#
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
|
66 |
# Handle different cases
|
67 |
if model_exists and dataset_exists:
|
@@ -115,12 +137,12 @@ def generate_summary(card_text: str, card_type: str) -> str:
|
|
115 |
"""Cached wrapper for generate_summary with TTL."""
|
116 |
return _generate_summary_gpu(card_text, card_type)
|
117 |
|
118 |
-
def summarize(hub_id: str = "") -> str:
|
119 |
"""Interface function for Gradio. Returns JSON format."""
|
120 |
try:
|
121 |
if hub_id:
|
122 |
-
# Fetch
|
123 |
-
card_type, card_text = get_card_info(hub_id)
|
124 |
|
125 |
if card_type == "both":
|
126 |
model_text, dataset_text = card_text
|
@@ -148,7 +170,15 @@ def summarize(hub_id: str = "") -> str:
|
|
148 |
def create_interface():
|
149 |
interface = gr.Interface(
|
150 |
fn=summarize,
|
151 |
-
inputs=
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
152 |
outputs=gr.JSON(label="Output"),
|
153 |
title="Hugging Face Hub TLDR Generator",
|
154 |
description="Generate concise summaries of model and dataset cards from the Hugging Face Hub.",
|
@@ -160,4 +190,4 @@ if __name__ == "__main__":
|
|
160 |
interface = create_interface()
|
161 |
interface.launch()
|
162 |
else:
|
163 |
-
print("Failed to load model. Please check the logs for details.")
|
|
|
38 |
logger.error(f"Failed to load model: {e}")
|
39 |
return False
|
40 |
|
41 |
+
def get_card_info(hub_id: str, repo_type: str = "auto") -> Tuple[str, str]:
|
42 |
"""Get card information from a Hugging Face hub_id."""
|
43 |
model_exists = False
|
44 |
dataset_exists = False
|
45 |
model_text = None
|
46 |
dataset_text = None
|
47 |
|
48 |
+
# Handle based on repo type
|
49 |
+
if repo_type == "auto":
|
50 |
+
# Try getting model card
|
51 |
+
try:
|
52 |
+
info = model_info(hub_id)
|
53 |
+
card = ModelCard.load(hub_id)
|
54 |
+
model_exists = True
|
55 |
+
model_text = card.text
|
56 |
+
except Exception as e:
|
57 |
+
logger.debug(f"No model card found for {hub_id}: {e}")
|
58 |
+
|
59 |
+
# Try getting dataset card
|
60 |
+
try:
|
61 |
+
info = dataset_info(hub_id)
|
62 |
+
card = DatasetCard.load(hub_id)
|
63 |
+
dataset_exists = True
|
64 |
+
dataset_text = card.text
|
65 |
+
except Exception as e:
|
66 |
+
logger.debug(f"No dataset card found for {hub_id}: {e}")
|
67 |
+
elif repo_type == "model":
|
68 |
+
try:
|
69 |
+
info = model_info(hub_id)
|
70 |
+
card = ModelCard.load(hub_id)
|
71 |
+
model_exists = True
|
72 |
+
model_text = card.text
|
73 |
+
except Exception as e:
|
74 |
+
logger.error(f"Failed to get model card for {hub_id}: {e}")
|
75 |
+
raise ValueError(f"Could not find model with id {hub_id}")
|
76 |
+
elif repo_type == "dataset":
|
77 |
+
try:
|
78 |
+
info = dataset_info(hub_id)
|
79 |
+
card = DatasetCard.load(hub_id)
|
80 |
+
dataset_exists = True
|
81 |
+
dataset_text = card.text
|
82 |
+
except Exception as e:
|
83 |
+
logger.error(f"Failed to get dataset card for {hub_id}: {e}")
|
84 |
+
raise ValueError(f"Could not find dataset with id {hub_id}")
|
85 |
+
else:
|
86 |
+
raise ValueError(f"Invalid repo_type: {repo_type}. Must be 'auto', 'model', or 'dataset'")
|
87 |
|
88 |
# Handle different cases
|
89 |
if model_exists and dataset_exists:
|
|
|
137 |
"""Cached wrapper for generate_summary with TTL."""
|
138 |
return _generate_summary_gpu(card_text, card_type)
|
139 |
|
140 |
+
def summarize(hub_id: str = "", repo_type: str = "auto") -> str:
|
141 |
"""Interface function for Gradio. Returns JSON format."""
|
142 |
try:
|
143 |
if hub_id:
|
144 |
+
# Fetch card information with specified repo_type
|
145 |
+
card_type, card_text = get_card_info(hub_id, repo_type)
|
146 |
|
147 |
if card_type == "both":
|
148 |
model_text, dataset_text = card_text
|
|
|
170 |
def create_interface():
|
171 |
interface = gr.Interface(
|
172 |
fn=summarize,
|
173 |
+
inputs=[
|
174 |
+
gr.Textbox(label="Hub ID", placeholder="e.g., huggingface/llama-7b"),
|
175 |
+
gr.Radio(
|
176 |
+
choices=["auto", "model", "dataset"],
|
177 |
+
value="auto",
|
178 |
+
label="Repository Type",
|
179 |
+
info="Choose 'auto' to detect automatically, or specify the repository type"
|
180 |
+
)
|
181 |
+
],
|
182 |
outputs=gr.JSON(label="Output"),
|
183 |
title="Hugging Face Hub TLDR Generator",
|
184 |
description="Generate concise summaries of model and dataset cards from the Hugging Face Hub.",
|
|
|
190 |
interface = create_interface()
|
191 |
interface.launch()
|
192 |
else:
|
193 |
+
print("Failed to load model. Please check the logs for details.")
|