Spaces:
Running
Running
jadehardouin
commited on
Commit
•
ecaa1ea
1
Parent(s):
7769b47
Update app.py
Browse files
app.py
CHANGED
@@ -1,8 +1,11 @@
|
|
1 |
import gradio as gr
|
2 |
import models
|
3 |
import pandas as pd
|
|
|
|
|
|
|
4 |
|
5 |
-
text = "<h1 style='text-align: center; color:
|
6 |
text0 = "<h1 style='text-align: center; color: midnightblue; font-size: 30px;'>Describe your use case"
|
7 |
text1 = "<h1 style='text-align: center; color: midnightblue; font-size: 25px;'>First option"
|
8 |
text2 = "<h1 style='text-align: center; color: midnightblue; font-size: 25px;'>Second option"
|
@@ -43,8 +46,93 @@ def update_plot(tco1, tco2, dropdown, dropdown2, labour_cost1, labour_cost2):
|
|
43 |
}
|
44 |
)
|
45 |
return gr.LinePlot.update(data, x="Number of requests", y="Cost ($)",color="AI model service",color_legend_position="bottom", title="Total Cost of Model Serving for one month", height=300, width=500, tooltip=["Number of requests", "Cost ($)", "AI model service"])
|
46 |
-
|
47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
Models: list[models.BaseTCOModel] = [models.OpenAIModel, models.CohereModel, models.OpenSourceLlama2Model]
|
49 |
model_names = [Model().get_name() for Model in Models]
|
50 |
gr.Markdown(value=text)
|
@@ -55,13 +143,13 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
|
|
55 |
# with gr.Row():
|
56 |
# gr.Markdown(value=text0)
|
57 |
with gr.Row():
|
58 |
-
use_case = gr.Dropdown(["Summarize", "Question-Answering", "Classification"], value="Question-Answering", label="Describe your use case")
|
59 |
-
with gr.Accordion("
|
60 |
with gr.Row():
|
61 |
-
input_tokens = gr.Slider(minimum=1, maximum=1000, value=300, step=1, label="Number of input token", info="We put a value that we find best suit your use case choice but feel free to adjust", interactive=True)
|
62 |
-
output_tokens = gr.Slider(minimum=1, maximum=1000, value=300, step=1, label="Number of output token", info="We put a value that we find best suit your use case choice but feel free to adjust", interactive=True)
|
63 |
with gr.Row(visible=False):
|
64 |
-
num_users = gr.Number(value="1000", interactive = True, label="Number of users for your service")
|
65 |
|
66 |
use_case.change(on_use_case_change, inputs=use_case, outputs=[input_tokens, output_tokens])
|
67 |
|
@@ -69,22 +157,22 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
|
|
69 |
with gr.Column():
|
70 |
#gr.Markdown(value=text1)
|
71 |
page1 = models.ModelPage(Models)
|
72 |
-
dropdown = gr.Dropdown(model_names, interactive=True, label="AI service option
|
73 |
-
with gr.Accordion("
|
74 |
page1.render()
|
75 |
|
76 |
with gr.Column():
|
77 |
#gr.Markdown(value=text2)
|
78 |
page2 = models.ModelPage(Models)
|
79 |
-
dropdown2 = gr.Dropdown(model_names, interactive=True, label="AI service option
|
80 |
-
with gr.Accordion("
|
81 |
page2.render()
|
82 |
|
83 |
dropdown.change(page1.make_model_visible, inputs=[dropdown, use_case], outputs=page1.get_all_components())
|
84 |
dropdown2.change(page2.make_model_visible, inputs=[dropdown2, use_case], outputs=page2.get_all_components())
|
85 |
|
86 |
#gr.Markdown(value=text3)
|
87 |
-
compute_tco_btn = gr.Button("Compute cost/request and TCOs", size="lg")
|
88 |
tco1 = gr.State()
|
89 |
tco2 = gr.State()
|
90 |
labour_cost1 = gr.State()
|
@@ -92,20 +180,20 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
|
|
92 |
|
93 |
with gr.Row():
|
94 |
with gr.Column():
|
95 |
-
tco_output = gr.Text("Output 1: ", label="Cost/request for the first option", info="This is only the infrastructure cost per request for deployment, the labor cost still has to be added for a Total Cost of Model Serving")
|
96 |
latency_info = gr.Markdown()
|
97 |
-
with gr.Accordion("
|
98 |
tco_formula = gr.Markdown()
|
99 |
|
100 |
with gr.Column():
|
101 |
-
tco_output2 = gr.Text("Output 2: ", label="Cost/request for the second option", info="This is only the infrastructure cost per request for deployment, the labor cost still has to be added for a Total Cost of Model Serving")
|
102 |
latency_info2 = gr.Markdown()
|
103 |
-
with gr.Accordion("
|
104 |
tco_formula2 = gr.Markdown()
|
105 |
|
106 |
with gr.Row():
|
107 |
with gr.Column(scale=1):
|
108 |
-
ratio = gr.Text("Ratio: ", label="Ratio of cost/request for both solutions")
|
109 |
with gr.Column(scale=3):
|
110 |
plot = gr.LinePlot()
|
111 |
|
|
|
1 |
import gradio as gr
|
2 |
import models
|
3 |
import pandas as pd
|
4 |
+
from gradio.themes.base import Base
|
5 |
+
from gradio.themes.utils import colors, fonts, sizes
|
6 |
+
from typing import Iterable
|
7 |
|
8 |
+
text = "<h1 style='text-align: center; color: #f0ba2d; font-size: 40px;'>TCO Comparison Calculator"
|
9 |
text0 = "<h1 style='text-align: center; color: midnightblue; font-size: 30px;'>Describe your use case"
|
10 |
text1 = "<h1 style='text-align: center; color: midnightblue; font-size: 25px;'>First option"
|
11 |
text2 = "<h1 style='text-align: center; color: midnightblue; font-size: 25px;'>Second option"
|
|
|
46 |
}
|
47 |
)
|
48 |
return gr.LinePlot.update(data, x="Number of requests", y="Cost ($)",color="AI model service",color_legend_position="bottom", title="Total Cost of Model Serving for one month", height=300, width=500, tooltip=["Number of requests", "Cost ($)", "AI model service"])
|
49 |
+
|
50 |
+
class Style(Base):
|
51 |
+
def __init__(
|
52 |
+
self,
|
53 |
+
*,
|
54 |
+
primary_hue: colors.Color | str = colors.neutral,
|
55 |
+
secondary_hue: colors.Color | str = colors.neutral,
|
56 |
+
neutral_hue: colors.Color | str = colors.neutral,
|
57 |
+
spacing_size: sizes.Size | str = sizes.spacing_md,
|
58 |
+
radius_size: sizes.Size | str = sizes.radius_md,
|
59 |
+
text_size: sizes.Size | str = sizes.text_md,
|
60 |
+
font: fonts.Font
|
61 |
+
| str
|
62 |
+
| Iterable[fonts.Font | str] = (fonts.GoogleFont("Sora")),
|
63 |
+
font_mono: fonts.Font
|
64 |
+
| str
|
65 |
+
| Iterable[fonts.Font | str] = (fonts.GoogleFont("Sora")),
|
66 |
+
):
|
67 |
+
super().__init__(
|
68 |
+
primary_hue=primary_hue,
|
69 |
+
secondary_hue=secondary_hue,
|
70 |
+
neutral_hue=neutral_hue,
|
71 |
+
spacing_size=spacing_size,
|
72 |
+
radius_size=radius_size,
|
73 |
+
text_size=text_size,
|
74 |
+
font=font,
|
75 |
+
font_mono=font_mono,
|
76 |
+
)
|
77 |
+
super().set(
|
78 |
+
background_fill_primary="#050f19", #The color of the background of blocks
|
79 |
+
background_fill_secondary="#050f19",
|
80 |
+
block_background_fill="#050f19", #The color of the background of blocks
|
81 |
+
block_background_fill_dark="#050f19",
|
82 |
+
|
83 |
+
border_color_primary="#050f19", #The border of a block such as dropdown
|
84 |
+
border_color_primary_dark="#050f19",
|
85 |
+
|
86 |
+
link_text_color="#f0ba2d", #The color for links
|
87 |
+
link_text_color_dark="#f0ba2d",
|
88 |
+
|
89 |
+
block_info_text_color="ffffff",
|
90 |
+
block_info_text_color_dark="ffffff",
|
91 |
+
|
92 |
+
block_border_color="#050f19", #The border color around an item (e.g. Accordion)
|
93 |
+
block_border_color_dark="#050f19",
|
94 |
+
block_shadow="*shadow_drop_lg",
|
95 |
+
#form_gap_width="*spacing_md", #The border gap between form elements, (e.g. consecutive textboxes)
|
96 |
+
|
97 |
+
input_background_fill="#081527", #The background of an input field
|
98 |
+
input_background_fill_dark="#081527",
|
99 |
+
input_border_color="#050f19",
|
100 |
+
input_border_color_dark="#050f19",
|
101 |
+
input_border_width="2px",
|
102 |
+
|
103 |
+
block_label_background_fill="#f0ba2d",
|
104 |
+
block_label_background_fill_dark="#f0ba2d",
|
105 |
+
block_label_border_color=None,
|
106 |
+
block_label_border_color_dark=None,
|
107 |
+
block_label_text_color="#050f19",
|
108 |
+
block_label_text_color_dark="#050f19",
|
109 |
+
|
110 |
+
button_primary_background_fill="#ffffff",
|
111 |
+
button_primary_border_color="#ffffff",
|
112 |
+
button_primary_text_color="#050f19",
|
113 |
+
button_shadow="*shadow_drop_lg",
|
114 |
+
|
115 |
+
block_title_background_fill="#f0ba2d", #The background of the title of a form element (e.g. textbox).
|
116 |
+
block_title_background_fill_dark="#f0ba2d", #The corner radius of the title of a form element (e.g. textbox).
|
117 |
+
block_title_radius="*radius_sm",
|
118 |
+
block_title_text_color="#050f19", #The text color of the title of a form element (e.g. textbox).
|
119 |
+
block_title_text_color_dark="#050f19",
|
120 |
+
block_title_text_size="*text_lg",
|
121 |
+
|
122 |
+
body_background_fill="#050f19",
|
123 |
+
body_background_fill_dark="#050f19",
|
124 |
+
body_text_color="#ffffff", #The default text color.
|
125 |
+
body_text_color_dark="#ffffff",
|
126 |
+
body_text_color_subdued="#ffffff",
|
127 |
+
body_text_color_subdued_dark="#ffffff",
|
128 |
+
|
129 |
+
slider_color="*secondary_300",
|
130 |
+
slider_color_dark="*secondary_600",
|
131 |
+
)
|
132 |
+
|
133 |
+
style = Style()
|
134 |
+
|
135 |
+
with gr.Blocks(theme=style) as demo:
|
136 |
Models: list[models.BaseTCOModel] = [models.OpenAIModel, models.CohereModel, models.OpenSourceLlama2Model]
|
137 |
model_names = [Model().get_name() for Model in Models]
|
138 |
gr.Markdown(value=text)
|
|
|
143 |
# with gr.Row():
|
144 |
# gr.Markdown(value=text0)
|
145 |
with gr.Row():
|
146 |
+
use_case = gr.Dropdown(["Summarize", "Question-Answering", "Classification"], value="Question-Answering", label=" Describe your use case ")
|
147 |
+
with gr.Accordion("Click here to customize the number of input and output tokens for your use case", open=False):
|
148 |
with gr.Row():
|
149 |
+
input_tokens = gr.Slider(minimum=1, maximum=1000, value=300, step=1, label=" Number of input token ", info="We put a value that we find best suit your use case choice but feel free to adjust", interactive=True)
|
150 |
+
output_tokens = gr.Slider(minimum=1, maximum=1000, value=300, step=1, label=" Number of output token ", info="We put a value that we find best suit your use case choice but feel free to adjust", interactive=True)
|
151 |
with gr.Row(visible=False):
|
152 |
+
num_users = gr.Number(value="1000", interactive = True, label=" Number of users for your service ")
|
153 |
|
154 |
use_case.change(on_use_case_change, inputs=use_case, outputs=[input_tokens, output_tokens])
|
155 |
|
|
|
157 |
with gr.Column():
|
158 |
#gr.Markdown(value=text1)
|
159 |
page1 = models.ModelPage(Models)
|
160 |
+
dropdown = gr.Dropdown(model_names, interactive=True, label=" First AI service option ")
|
161 |
+
with gr.Accordion("Click here for more information on the computation parameters for your first AI service option", open=False):
|
162 |
page1.render()
|
163 |
|
164 |
with gr.Column():
|
165 |
#gr.Markdown(value=text2)
|
166 |
page2 = models.ModelPage(Models)
|
167 |
+
dropdown2 = gr.Dropdown(model_names, interactive=True, label=" Second AI service option ")
|
168 |
+
with gr.Accordion("Click here for more information on the computation parameters for your second AI service option", open=False):
|
169 |
page2.render()
|
170 |
|
171 |
dropdown.change(page1.make_model_visible, inputs=[dropdown, use_case], outputs=page1.get_all_components())
|
172 |
dropdown2.change(page2.make_model_visible, inputs=[dropdown2, use_case], outputs=page2.get_all_components())
|
173 |
|
174 |
#gr.Markdown(value=text3)
|
175 |
+
compute_tco_btn = gr.Button("Compute cost/request and TCOs", size="lg", variant="primary", scale=1)
|
176 |
tco1 = gr.State()
|
177 |
tco2 = gr.State()
|
178 |
labour_cost1 = gr.State()
|
|
|
180 |
|
181 |
with gr.Row():
|
182 |
with gr.Column():
|
183 |
+
tco_output = gr.Text("Output 1: ", label=" Cost/request for the first option ", info="This is only the infrastructure cost per request for deployment, the labor cost still has to be added for a Total Cost of Model Serving")
|
184 |
latency_info = gr.Markdown()
|
185 |
+
with gr.Accordion("Click here to see the formula", open=False):
|
186 |
tco_formula = gr.Markdown()
|
187 |
|
188 |
with gr.Column():
|
189 |
+
tco_output2 = gr.Text("Output 2: ", label=" Cost/request for the second option ", info="This is only the infrastructure cost per request for deployment, the labor cost still has to be added for a Total Cost of Model Serving")
|
190 |
latency_info2 = gr.Markdown()
|
191 |
+
with gr.Accordion("Click here to see the formula", open=False):
|
192 |
tco_formula2 = gr.Markdown()
|
193 |
|
194 |
with gr.Row():
|
195 |
with gr.Column(scale=1):
|
196 |
+
ratio = gr.Text("Ratio: ", label=" Ratio of cost/request for both solutions ")
|
197 |
with gr.Column(scale=3):
|
198 |
plot = gr.LinePlot()
|
199 |
|