Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -29,6 +29,15 @@ st.markdown("""
|
|
29 |
border-radius: 10px;
|
30 |
padding: 15px;
|
31 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
</style>
|
33 |
""", unsafe_allow_html=True)
|
34 |
|
@@ -57,7 +66,6 @@ def plot_model_comparison(selected_model):
|
|
57 |
fig, ax = plt.subplots(figsize=(10, 6))
|
58 |
bars = ax.bar(model_names, params)
|
59 |
|
60 |
-
# Highlight selected model
|
61 |
index = list(MODELS.keys()).index(selected_model)
|
62 |
bars[index].set_color('#00ff00')
|
63 |
|
@@ -70,8 +78,70 @@ def plot_model_comparison(selected_model):
|
|
70 |
|
71 |
st.pyplot(fig)
|
72 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
73 |
def visualize_attention_patterns():
|
74 |
-
# Simplified attention patterns visualization
|
75 |
fig, ax = plt.subplots(figsize=(8, 6))
|
76 |
data = torch.randn(5, 5)
|
77 |
ax.imshow(data, cmap='viridis')
|
@@ -83,14 +153,10 @@ def visualize_attention_patterns():
|
|
83 |
def main():
|
84 |
st.title("π§ Transformer Model Visualizer")
|
85 |
|
86 |
-
# Model selection
|
87 |
selected_model = st.sidebar.selectbox("Select Model", list(MODELS.keys()))
|
88 |
-
|
89 |
-
# Model details
|
90 |
model_info = MODELS[selected_model]
|
91 |
config = get_model_config(selected_model)
|
92 |
|
93 |
-
# Display metrics
|
94 |
col1, col2, col3, col4 = st.columns(4)
|
95 |
with col1:
|
96 |
st.metric("Model Type", model_info["type"])
|
@@ -101,13 +167,20 @@ def main():
|
|
101 |
with col4:
|
102 |
st.metric("Parameters", f"{model_info['params']}M")
|
103 |
|
104 |
-
|
105 |
-
tab1, tab2, tab3 = st.tabs(["Model Structure", "Comparison", "Model Specific"])
|
106 |
|
107 |
with tab1:
|
108 |
st.subheader("Architecture Diagram")
|
109 |
-
|
110 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
111 |
|
112 |
with tab2:
|
113 |
st.subheader("Model Size Comparison")
|
|
|
29 |
border-radius: 10px;
|
30 |
padding: 15px;
|
31 |
}
|
32 |
+
.architecture {
|
33 |
+
font-family: monospace;
|
34 |
+
color: #00ff00;
|
35 |
+
white-space: pre-wrap;
|
36 |
+
background-color: #1a1a1a;
|
37 |
+
padding: 20px;
|
38 |
+
border-radius: 10px;
|
39 |
+
border: 1px solid #00ff00;
|
40 |
+
}
|
41 |
</style>
|
42 |
""", unsafe_allow_html=True)
|
43 |
|
|
|
66 |
fig, ax = plt.subplots(figsize=(10, 6))
|
67 |
bars = ax.bar(model_names, params)
|
68 |
|
|
|
69 |
index = list(MODELS.keys()).index(selected_model)
|
70 |
bars[index].set_color('#00ff00')
|
71 |
|
|
|
78 |
|
79 |
st.pyplot(fig)
|
80 |
|
81 |
+
def visualize_architecture(model_info):
|
82 |
+
architecture = []
|
83 |
+
model_type = model_info["type"]
|
84 |
+
layers = model_info["layers"]
|
85 |
+
heads = model_info["heads"]
|
86 |
+
|
87 |
+
architecture.append("Input")
|
88 |
+
architecture.append("β")
|
89 |
+
architecture.append("βΌ")
|
90 |
+
|
91 |
+
if model_type == "Encoder":
|
92 |
+
architecture.append("[Embedding Layer]")
|
93 |
+
for i in range(layers):
|
94 |
+
architecture.extend([
|
95 |
+
f"Encoder Layer {i+1}",
|
96 |
+
"ββ Multi-Head Attention",
|
97 |
+
f"β ββ {heads} Heads",
|
98 |
+
"ββ Layer Normalization",
|
99 |
+
"ββ Feed Forward Network",
|
100 |
+
"β",
|
101 |
+
"βΌ"
|
102 |
+
])
|
103 |
+
architecture.append("[Output]")
|
104 |
+
|
105 |
+
elif model_type == "Decoder":
|
106 |
+
architecture.append("[Embedding Layer]")
|
107 |
+
for i in range(layers):
|
108 |
+
architecture.extend([
|
109 |
+
f"Decoder Layer {i+1}",
|
110 |
+
"ββ Masked Multi-Head Attention",
|
111 |
+
f"β ββ {heads} Heads",
|
112 |
+
"ββ Layer Normalization",
|
113 |
+
"ββ Feed Forward Network",
|
114 |
+
"β",
|
115 |
+
"βΌ"
|
116 |
+
])
|
117 |
+
architecture.append("[Output]")
|
118 |
+
|
119 |
+
elif model_type == "Seq2Seq":
|
120 |
+
architecture.append("Encoder Stack")
|
121 |
+
for i in range(layers):
|
122 |
+
architecture.extend([
|
123 |
+
f"Encoder Layer {i+1}",
|
124 |
+
"ββ Self-Attention",
|
125 |
+
"ββ Feed Forward Network",
|
126 |
+
"β",
|
127 |
+
"βΌ"
|
128 |
+
])
|
129 |
+
architecture.append("βββ [Context] βββ")
|
130 |
+
architecture.append("Decoder Stack")
|
131 |
+
for i in range(layers):
|
132 |
+
architecture.extend([
|
133 |
+
f"Decoder Layer {i+1}",
|
134 |
+
"ββ Masked Self-Attention",
|
135 |
+
"ββ Encoder-Decoder Attention",
|
136 |
+
"ββ Feed Forward Network",
|
137 |
+
"β",
|
138 |
+
"βΌ"
|
139 |
+
])
|
140 |
+
architecture.append("[Output]")
|
141 |
+
|
142 |
+
return "\n".join(architecture)
|
143 |
+
|
144 |
def visualize_attention_patterns():
|
|
|
145 |
fig, ax = plt.subplots(figsize=(8, 6))
|
146 |
data = torch.randn(5, 5)
|
147 |
ax.imshow(data, cmap='viridis')
|
|
|
153 |
def main():
|
154 |
st.title("π§ Transformer Model Visualizer")
|
155 |
|
|
|
156 |
selected_model = st.sidebar.selectbox("Select Model", list(MODELS.keys()))
|
|
|
|
|
157 |
model_info = MODELS[selected_model]
|
158 |
config = get_model_config(selected_model)
|
159 |
|
|
|
160 |
col1, col2, col3, col4 = st.columns(4)
|
161 |
with col1:
|
162 |
st.metric("Model Type", model_info["type"])
|
|
|
167 |
with col4:
|
168 |
st.metric("Parameters", f"{model_info['params']}M")
|
169 |
|
170 |
+
tab1, tab2, tab3 = st.tabs(["Model Structure", "Comparison", "Model Attention"])
|
|
|
171 |
|
172 |
with tab1:
|
173 |
st.subheader("Architecture Diagram")
|
174 |
+
architecture = visualize_architecture(model_info)
|
175 |
+
st.markdown(f"<div class='architecture'>{architecture}</div>", unsafe_allow_html=True)
|
176 |
+
|
177 |
+
st.markdown("""
|
178 |
+
**Legend:**
|
179 |
+
- **Multi-Head Attention**: Self-attention mechanism with multiple parallel heads
|
180 |
+
- **Layer Normalization**: Normalization operation between layers
|
181 |
+
- **Feed Forward Network**: Position-wise fully connected network
|
182 |
+
- **Masked Attention**: Attention with future token masking
|
183 |
+
""")
|
184 |
|
185 |
with tab2:
|
186 |
st.subheader("Model Size Comparison")
|