Sergidev commited on
Commit
f6551d5
·
verified ·
1 Parent(s): 5e78e4f
Files changed (1) hide show
  1. app.py +21 -81
app.py CHANGED
@@ -3,7 +3,6 @@ import spaces
3
  import torch
4
  from transformers import AutoTokenizer, AutoModel
5
  import plotly.graph_objects as go
6
- import random
7
 
8
  model_name = "mistralai/Mistral-7B-v0.1"
9
  tokenizer = AutoTokenizer.from_pretrained(model_name)
@@ -28,91 +27,32 @@ def get_embedding(text):
28
  def reduce_to_3d(embedding):
29
  return embedding[:3]
30
 
31
- def generate_random_color():
32
- return f'rgb({random.randint(0, 255)}, {random.randint(0, 255)}, {random.randint(0, 255)})'
33
-
34
  @spaces.GPU
35
- def compare_embeddings(strings):
36
- embeddings = [get_embedding(text) for text in strings if text.strip()]
37
- embeddings_3d = [reduce_to_3d(emb) for emb in embeddings]
38
 
39
- fig = go.Figure()
40
-
41
- # Add origin point
42
- fig.add_trace(go.Scatter3d(x=[0], y=[0], z=[0], mode='markers', name='Origin',
43
- marker=dict(color='black', size=8)))
44
-
45
- # Add lines and points for each string
46
- for i, emb in enumerate(embeddings_3d):
47
- color = generate_random_color()
48
- fig.add_trace(go.Scatter3d(x=[0, emb[0]], y=[0, emb[1]], z=[0, emb[2]],
49
- mode='lines+markers', name=f'Text {i+1}',
50
- line=dict(color=color), marker=dict(color=color)))
51
-
52
- fig.update_layout(scene=dict(xaxis_title='X', yaxis_title='Y', zaxis_title='Z'))
53
 
54
- return fig
55
-
56
- def add_string(string_list):
57
- string_list.append("")
58
- return string_list
59
-
60
- def remove_string(string_list):
61
- if len(string_list) > 1:
62
- string_list.pop()
63
- return string_list
64
-
65
- with gr.Blocks(title="3D Embedding Comparison") as iface:
66
- gr.Markdown("# 3D Embedding Comparison")
67
- gr.Markdown("Compare the embeddings of multiple strings visualized in 3D space using Mistral 7B.")
68
-
69
- with gr.Row():
70
- with gr.Column(scale=1):
71
- string_inputs = gr.State([""])
72
- with gr.Column(variant="panel") as textbox_container:
73
- textboxes = gr.Textbox(label="Text 1")
74
- add_button = gr.Button("Add String")
75
- remove_button = gr.Button("Remove String")
76
-
77
- with gr.Column(scale=2):
78
- plot_output = gr.Plot()
79
 
80
- submit_button = gr.Button("Compare Embeddings")
81
-
82
- def update_textboxes(string_list):
83
- components = []
84
- for i, s in enumerate(string_list):
85
- components.append(gr.Textbox(value=s, label=f"Text {i+1}"))
86
- return gr.Column.update(visible=True, children=components)
87
-
88
- add_button.click(
89
- add_string,
90
- inputs=[string_inputs],
91
- outputs=[string_inputs]
92
- ).then(
93
- update_textboxes,
94
- inputs=[string_inputs],
95
- outputs=[textbox_container]
96
- )
97
-
98
- remove_button.click(
99
- remove_string,
100
- inputs=[string_inputs],
101
- outputs=[string_inputs]
102
- ).then(
103
- update_textboxes,
104
- inputs=[string_inputs],
105
- outputs=[textbox_container]
106
- )
107
 
108
- def process_and_compare(strings):
109
- processed_strings = [t.strip() for t in strings if t.strip()]
110
- return compare_embeddings(processed_strings)
111
 
112
- submit_button.click(
113
- process_and_compare,
114
- inputs=[string_inputs],
115
- outputs=[plot_output]
116
- )
 
 
 
 
 
117
 
118
  iface.launch()
 
3
  import torch
4
  from transformers import AutoTokenizer, AutoModel
5
  import plotly.graph_objects as go
 
6
 
7
  model_name = "mistralai/Mistral-7B-v0.1"
8
  tokenizer = AutoTokenizer.from_pretrained(model_name)
 
27
  def reduce_to_3d(embedding):
28
  return embedding[:3]
29
 
 
 
 
30
  @spaces.GPU
31
+ def compare_embeddings(text1, text2):
32
+ emb1 = get_embedding(text1)
33
+ emb2 = get_embedding(text2)
34
 
35
+ emb1_3d = reduce_to_3d(emb1)
36
+ emb2_3d = reduce_to_3d(emb2)
 
 
 
 
 
 
 
 
 
 
 
 
37
 
38
+ fig = go.Figure(data=[
39
+ go.Scatter3d(x=[0, emb1_3d[0]], y=[0, emb1_3d[1]], z=[0, emb1_3d[2]], mode='lines+markers', name='Text 1'),
40
+ go.Scatter3d(x=[0, emb2_3d[0]], y=[0, emb2_3d[1]], z=[0, emb2_3d[2]], mode='lines+markers', name='Text 2')
41
+ ])
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
42
 
43
+ fig.update_layout(scene=dict(xaxis_title='X', yaxis_title='Y', zaxis_title='Z'))
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
 
45
+ return fig
 
 
46
 
47
+ iface = gr.Interface(
48
+ fn=compare_embeddings,
49
+ inputs=[
50
+ gr.Textbox(label="Text 1"),
51
+ gr.Textbox(label="Text 2")
52
+ ],
53
+ outputs=gr.Plot(),
54
+ title="3D Embedding Comparison",
55
+ description="Compare the embeddings of two strings visualized in 3D space using Mistral 7B."
56
+ )
57
 
58
  iface.launch()