Dashboard-fixes (#1)
Browse files- feat: enhanced GUI (63daa8ce6608610cb201ff8faeb918a65aa58d94)
- bug: removed share=True (ed1a96ef44b5680c941da2d9e14e98e4314f30a2)
app.py
CHANGED
@@ -16,109 +16,102 @@ examples = [
|
|
16 |
["Boxing Day ambush & flagship attack Putin has long tried to downplay the true losses his army has faced in the Black Sea."],
|
17 |
]
|
18 |
|
19 |
-
# Custom model class for combining sentiment analysis with subjectivity detection
|
20 |
class CustomModel(PreTrainedModel):
|
21 |
config_class = DebertaV2Config
|
22 |
-
|
23 |
def __init__(self, config, sentiment_dim=3, num_labels=2, *args, **kwargs):
|
24 |
super().__init__(config, *args, **kwargs)
|
25 |
self.deberta = DebertaV2Model(config)
|
26 |
self.pooler = ContextPooler(config)
|
27 |
output_dim = self.pooler.output_dim
|
28 |
self.dropout = nn.Dropout(0.1)
|
29 |
-
|
30 |
self.classifier = nn.Linear(output_dim + sentiment_dim, num_labels)
|
31 |
|
32 |
def forward(self, input_ids, positive, neutral, negative, token_type_ids=None, attention_mask=None, labels=None):
|
33 |
outputs = self.deberta(input_ids=input_ids, attention_mask=attention_mask)
|
34 |
-
|
35 |
encoder_layer = outputs[0]
|
36 |
pooled_output = self.pooler(encoder_layer)
|
37 |
-
|
38 |
-
# Sentiment features as a single tensor
|
39 |
-
sentiment_features = torch.stack((positive, neutral, negative), dim=1) # Shape: (batch_size, 3)
|
40 |
-
|
41 |
-
# Combine CLS embedding with sentiment features
|
42 |
combined_features = torch.cat((pooled_output, sentiment_features), dim=1)
|
43 |
-
|
44 |
-
# Classification head
|
45 |
logits = self.classifier(self.dropout(combined_features))
|
46 |
-
|
47 |
return {'logits': logits}
|
48 |
|
49 |
-
# Load the pre-trained tokenizer
|
50 |
def load_tokenizer(model_name: str):
|
51 |
return AutoTokenizer.from_pretrained(model_name)
|
52 |
|
53 |
-
|
54 |
def load_model(model_name: str):
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
label2id={'OBJ': 0, 'SUBJ': 1},
|
62 |
-
output_attentions=False,
|
63 |
-
output_hidden_states=False
|
64 |
-
)
|
65 |
-
|
66 |
-
model = CustomModel(config=config, sentiment_dim=3, num_labels=2).from_pretrained(model_name)
|
67 |
-
|
68 |
-
else:
|
69 |
-
model = AutoModelForSequenceClassification.from_pretrained(
|
70 |
-
model_name,
|
71 |
-
num_labels=2,
|
72 |
-
id2label={0: 'OBJ', 1: 'SUBJ'},
|
73 |
-
label2id={'OBJ': 0, 'SUBJ': 1},
|
74 |
-
output_attentions=False,
|
75 |
-
output_hidden_states=False
|
76 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
|
78 |
-
|
79 |
-
|
80 |
-
# Get sentiment values using a pre-trained sentiment analysis model
|
81 |
def get_sentiment_values(text: str):
|
82 |
-
|
83 |
-
|
84 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
85 |
|
86 |
def analyze(text):
|
87 |
-
|
88 |
-
|
|
|
|
|
|
|
|
|
|
|
89 |
|
90 |
-
|
91 |
tokenizer = load_tokenizer(model_card)
|
92 |
model_with_sentiment = load_model(sentiment_model)
|
93 |
model_without_sentiment = load_model(subjectivity_only_model)
|
94 |
|
95 |
-
|
96 |
-
|
|
|
|
|
97 |
|
98 |
-
|
99 |
-
outputs_base = model_without_sentiment(**inputs)
|
100 |
logits_base = outputs_base.get('logits')
|
101 |
-
# Calculate probabilities using softmax
|
102 |
prob_base = torch.nn.functional.softmax(logits_base, dim=1)[0]
|
|
|
|
|
|
|
|
|
|
|
103 |
|
104 |
-
|
105 |
-
positive =
|
106 |
-
neutral =
|
107 |
-
negative =
|
108 |
-
|
109 |
-
# Convert sentiment values to tensors
|
110 |
-
inputs['positive'] = torch.tensor(positive).unsqueeze(0)
|
111 |
-
inputs['neutral'] = torch.tensor(neutral).unsqueeze(0)
|
112 |
-
inputs['negative'] = torch.tensor(negative).unsqueeze(0)
|
113 |
-
|
114 |
-
# Get the sentiment model outputs
|
115 |
-
outputs_sentiment = model_with_sentiment(**inputs)
|
116 |
-
logits_sentiment = outputs_sentiment.get('logits')
|
117 |
|
118 |
-
|
|
|
|
|
119 |
prob_sentiment = torch.nn.functional.softmax(logits_sentiment, dim=1)[0]
|
120 |
|
121 |
-
# Prepare data for the Dataframe (string values)
|
122 |
table_data = [
|
123 |
["Positive", f"{positive:.2%}"],
|
124 |
["Neutral", f"{neutral:.2%}"],
|
@@ -128,31 +121,70 @@ def analyze(text):
|
|
128 |
["TextOnly OBJ", f"{prob_base[0]:.2%}"],
|
129 |
["TextOnly SUBJ", f"{prob_base[1]:.2%}"]
|
130 |
]
|
131 |
-
|
132 |
return table_data
|
133 |
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
|
139 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
140 |
with gr.Tabs():
|
141 |
with gr.TabItem("Raw Scores π"):
|
142 |
-
table = gr.Dataframe(
|
|
|
|
|
|
|
|
|
143 |
with gr.TabItem("About βΉοΈ"):
|
144 |
-
gr.Markdown(
|
|
|
|
|
|
|
|
|
145 |
with gr.Row():
|
146 |
gr.Markdown("### Examples:")
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
|
|
|
|
156 |
btn.click(fn=analyze, inputs=txt, outputs=[table])
|
157 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
158 |
demo.queue().launch()
|
|
|
16 |
["Boxing Day ambush & flagship attack Putin has long tried to downplay the true losses his army has faced in the Black Sea."],
|
17 |
]
|
18 |
|
|
|
19 |
class CustomModel(PreTrainedModel):
|
20 |
config_class = DebertaV2Config
|
|
|
21 |
def __init__(self, config, sentiment_dim=3, num_labels=2, *args, **kwargs):
|
22 |
super().__init__(config, *args, **kwargs)
|
23 |
self.deberta = DebertaV2Model(config)
|
24 |
self.pooler = ContextPooler(config)
|
25 |
output_dim = self.pooler.output_dim
|
26 |
self.dropout = nn.Dropout(0.1)
|
|
|
27 |
self.classifier = nn.Linear(output_dim + sentiment_dim, num_labels)
|
28 |
|
29 |
def forward(self, input_ids, positive, neutral, negative, token_type_ids=None, attention_mask=None, labels=None):
|
30 |
outputs = self.deberta(input_ids=input_ids, attention_mask=attention_mask)
|
|
|
31 |
encoder_layer = outputs[0]
|
32 |
pooled_output = self.pooler(encoder_layer)
|
33 |
+
sentiment_features = torch.stack((positive, neutral, negative), dim=1).to(pooled_output.dtype)
|
|
|
|
|
|
|
|
|
34 |
combined_features = torch.cat((pooled_output, sentiment_features), dim=1)
|
|
|
|
|
35 |
logits = self.classifier(self.dropout(combined_features))
|
|
|
36 |
return {'logits': logits}
|
37 |
|
|
|
38 |
def load_tokenizer(model_name: str):
|
39 |
return AutoTokenizer.from_pretrained(model_name)
|
40 |
|
41 |
+
load_model_cache = {}
|
42 |
def load_model(model_name: str):
|
43 |
+
if model_name not in load_model_cache:
|
44 |
+
print(f"Loading model: {model_name}")
|
45 |
+
if 'sentiment' in model_name:
|
46 |
+
config = DebertaV2Config.from_pretrained(
|
47 |
+
model_name, num_labels=2, id2label={0: 'OBJ', 1: 'SUBJ'}, label2id={'OBJ': 0, 'SUBJ': 1},
|
48 |
+
output_attentions=False, output_hidden_states=False
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
)
|
50 |
+
model_instance = CustomModel(config=config, sentiment_dim=3, num_labels=2).from_pretrained(model_name)
|
51 |
+
else:
|
52 |
+
model_instance = AutoModelForSequenceClassification.from_pretrained(
|
53 |
+
model_name, num_labels=2, id2label={0: 'OBJ', 1: 'SUBJ'}, label2id={'OBJ': 0, 'SUBJ': 1},
|
54 |
+
output_attentions=False, output_hidden_states=False
|
55 |
+
)
|
56 |
+
load_model_cache[model_name] = model_instance
|
57 |
+
return load_model_cache[model_name]
|
58 |
|
59 |
+
sentiment_pipeline_cache = None #
|
|
|
|
|
60 |
def get_sentiment_values(text: str):
|
61 |
+
global sentiment_pipeline_cache
|
62 |
+
if sentiment_pipeline_cache is None:
|
63 |
+
print("Loading sentiment pipeline...")
|
64 |
+
sentiment_pipeline_cache = pipeline(
|
65 |
+
"sentiment-analysis",
|
66 |
+
model="cardiffnlp/twitter-xlm-roberta-base-sentiment",
|
67 |
+
tokenizer="cardiffnlp/twitter-xlm-roberta-base-sentiment",
|
68 |
+
top_k=None
|
69 |
+
)
|
70 |
+
sentiments_output = sentiment_pipeline_cache(text)
|
71 |
+
if sentiments_output and isinstance(sentiments_output, list) and sentiments_output[0]:
|
72 |
+
sentiments = sentiments_output[0]
|
73 |
+
return {s['label'].lower(): s['score'] for s in sentiments}
|
74 |
+
return {}
|
75 |
+
|
76 |
|
77 |
def analyze(text):
|
78 |
+
if not text or not text.strip():
|
79 |
+
empty_data = [
|
80 |
+
["Positive", ""], ["Neutral", ""], ["Negative", ""],
|
81 |
+
["Sent-Subj OBJ", ""], ["Sent-Subj SUBJ", ""],
|
82 |
+
["TextOnly OBJ", ""], ["TextOnly SUBJ", ""]
|
83 |
+
]
|
84 |
+
return empty_data
|
85 |
|
86 |
+
sentiment_values = get_sentiment_values(text)
|
87 |
tokenizer = load_tokenizer(model_card)
|
88 |
model_with_sentiment = load_model(sentiment_model)
|
89 |
model_without_sentiment = load_model(subjectivity_only_model)
|
90 |
|
91 |
+
inputs_dict = tokenizer(text, padding=True, truncation=True, max_length=256, return_tensors='pt')
|
92 |
+
|
93 |
+
device = next(model_without_sentiment.parameters()).device
|
94 |
+
inputs_dict_on_device = {k: v.to(device) for k, v in inputs_dict.items()}
|
95 |
|
96 |
+
outputs_base = model_without_sentiment(**inputs_dict_on_device)
|
|
|
97 |
logits_base = outputs_base.get('logits')
|
|
|
98 |
prob_base = torch.nn.functional.softmax(logits_base, dim=1)[0]
|
99 |
+
|
100 |
+
positive = sentiment_values.get('positive', 0.0)
|
101 |
+
neutral = sentiment_values.get('neutral', 0.0)
|
102 |
+
negative = sentiment_values.get('negative', 0.0)
|
103 |
+
|
104 |
|
105 |
+
current_inputs_for_sentiment_model = inputs_dict_on_device.copy()
|
106 |
+
current_inputs_for_sentiment_model['positive'] = torch.tensor(positive, device=device).unsqueeze(0).float()
|
107 |
+
current_inputs_for_sentiment_model['neutral'] = torch.tensor(neutral, device=device).unsqueeze(0).float()
|
108 |
+
current_inputs_for_sentiment_model['negative'] = torch.tensor(negative, device=device).unsqueeze(0).float()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
109 |
|
110 |
+
|
111 |
+
outputs_sentiment = model_with_sentiment(**current_inputs_for_sentiment_model)
|
112 |
+
logits_sentiment = outputs_sentiment.get('logits')
|
113 |
prob_sentiment = torch.nn.functional.softmax(logits_sentiment, dim=1)[0]
|
114 |
|
|
|
115 |
table_data = [
|
116 |
["Positive", f"{positive:.2%}"],
|
117 |
["Neutral", f"{neutral:.2%}"],
|
|
|
121 |
["TextOnly OBJ", f"{prob_base[0]:.2%}"],
|
122 |
["TextOnly SUBJ", f"{prob_base[1]:.2%}"]
|
123 |
]
|
|
|
124 |
return table_data
|
125 |
|
126 |
+
def load_default_example_on_startup():
|
127 |
+
print("Loading default example on startup...")
|
128 |
+
if examples and examples[0] and isinstance(examples[0], list) and examples[0]:
|
129 |
+
default_text = examples[0][0]
|
130 |
+
default_analysis_results = analyze(default_text)
|
131 |
+
return default_text, default_analysis_results
|
132 |
+
print("Warning: No valid default example found. Loading empty.")
|
133 |
+
empty_text = ""
|
134 |
+
empty_results = analyze(empty_text)
|
135 |
+
return empty_text, empty_results
|
136 |
+
|
137 |
+
with gr.Blocks(theme=gr.themes.Ocean(), title="Subjectivity & Sentiment Dashboard") as demo:
|
138 |
+
gr.Markdown("# π Subjectivity & Sentiment Analysis Dashboard π")
|
139 |
+
|
140 |
+
with gr.Column():
|
141 |
+
txt = gr.Textbox(
|
142 |
+
label="Enter text to analyze",
|
143 |
+
placeholder="Paste news sentence here...",
|
144 |
+
lines=2,
|
145 |
+
)
|
146 |
+
with gr.Row():
|
147 |
+
gr.Column(scale=1, min_width=0)
|
148 |
+
btn = gr.Button(
|
149 |
+
"Analyze π",
|
150 |
+
variant="primary",
|
151 |
+
size="md",
|
152 |
+
scale=0
|
153 |
+
)
|
154 |
+
|
155 |
with gr.Tabs():
|
156 |
with gr.TabItem("Raw Scores π"):
|
157 |
+
table = gr.Dataframe(
|
158 |
+
headers=["Metric", "Value"],
|
159 |
+
datatype=["str", "str"],
|
160 |
+
interactive=False
|
161 |
+
)
|
162 |
with gr.TabItem("About βΉοΈ"):
|
163 |
+
gr.Markdown(
|
164 |
+
"This dashboard uses two DeBERTa-based models (with and without sentiment integration) "
|
165 |
+
"to detect subjectivity, alongside sentiment scores from an XLM-RoBERTa model."
|
166 |
+
)
|
167 |
+
|
168 |
with gr.Row():
|
169 |
gr.Markdown("### Examples:")
|
170 |
+
|
171 |
+
|
172 |
+
gr.Examples(
|
173 |
+
examples=examples,
|
174 |
+
inputs=txt,
|
175 |
+
outputs=[table],
|
176 |
+
fn=analyze,
|
177 |
+
label="Click an example to analyze",
|
178 |
+
cache_examples=True,
|
179 |
+
)
|
180 |
+
|
181 |
btn.click(fn=analyze, inputs=txt, outputs=[table])
|
182 |
|
183 |
+
|
184 |
+
demo.load(
|
185 |
+
fn=load_default_example_on_startup,
|
186 |
+
inputs=None,
|
187 |
+
outputs=[txt, table]
|
188 |
+
)
|
189 |
+
|
190 |
demo.queue().launch()
|