piotr-szleg-bards-ai commited on
Commit
2966484
1 Parent(s): e879520

2024-02-06 13:34:44 Publish script update

Browse files
app.py CHANGED
@@ -5,7 +5,7 @@ import pandas as pd
5
  import plotly
6
  from pandas.api.types import is_numeric_dtype
7
 
8
- from pipeline.config import LLMBoardConfig
9
 
10
  README = """
11
  This projects compares different large language models and their providers for real time applications and mass data processing.
@@ -17,7 +17,7 @@ For this version we chose English, Polish and Japanese languages, with Japanese
17
  We used the following prompt:
18
 
19
  ```
20
- Summarize me this text, the summary should be in {language}
21
  ```
22
 
23
  Where language variable is original language of the text as we wanted to avoid the model translating the text to English during summarization.
@@ -25,9 +25,7 @@ Where language variable is original language of the text as we wanted to avoid t
25
  The model was asked to return the output in three formats: markdown, json and function call. Note that currently function calls are only supported by Open AI API.
26
  To do that we added following text to the query:
27
 
28
- ```
29
- ...
30
- ```
31
 
32
  When measuring execution time we used `time.time()` result saved to variable before making the call to API and compared it to `time.time()` result after receiving the results. We used litellm python library for all of the models which naturally adds some overhead compared to pure curl calls.
33
 
@@ -41,9 +39,16 @@ time_periods_explanation_df = pd.DataFrame({
41
  'hour_range': ["6-8", "9-11", "12-14", "15-17", "18-20", "21-23", "0-2", "3-5"]
42
  })
43
 
 
 
 
 
 
 
 
44
  summary_df: pd.DataFrame = pd.read_csv("data/2024-02-05 23:33:22.947120_summary.csv")
45
  time_of_day_comparison_df = pd.read_csv("data/2024-02-06 09:49:19.637072_time_of_day_comparison.csv")
46
- general_plots = pd.read_csv("data/2024-02-05 12:03:42.452218_general_plot.csv")
47
  model_costs_df = pd.read_csv("data/2024-02-05 12:03:45.281624_model_costs.csv")
48
 
49
 
@@ -129,8 +134,8 @@ def snake_case_to_title(text):
129
  return " ".join(title_words)
130
 
131
 
132
- filter_textbox = gr.Textbox(label="Model name part")
133
- filter_button = gr.Button("Filter dataframes by model name")
134
  collapse_languages_button = gr.Button("Collapse languages")
135
  collapse_output_method_button = gr.Button("Collapse output method")
136
  last_textbox = 0
@@ -143,14 +148,17 @@ with gr.Blocks() as demo:
143
  filter_button.render()
144
 
145
  with gr.Tab("About this project"):
146
- gr.Markdown(README)
 
 
 
147
  with gr.Tab("General plots"):
148
  for index, row in general_plots.iterrows():
149
  plot = plotly.io.from_json(row["plot_json"])
150
  plot.update_layout(autosize=True)
151
- gr.Plot(plot, label=row["description"], scale=1)
152
- if pd.notna(row["comment"]):
153
- gr.Markdown(str(row["comment"]))
154
  with gr.Tab("Output characteristics"):
155
  with gr.Row():
156
  collapse_languages_button.render()
@@ -163,6 +171,8 @@ with gr.Blocks() as demo:
163
  time_periods_explanation_ui = gr.DataFrame(dataframe_style(time_periods_explanation_df), label="Times of day ranges")
164
  gr.Markdown("""
165
  These measurements were made by testing the models using the same dataset as in the other comparisons every hour for 24 hours.
 
 
166
 
167
  Hours and times of day in the table and in the plot are based on Central European Time.
168
 
 
5
  import plotly
6
  from pandas.api.types import is_numeric_dtype
7
 
8
+ from pipeline.config import LLMBoardConfig, QueriesConfig
9
 
10
  README = """
11
  This projects compares different large language models and their providers for real time applications and mass data processing.
 
17
  We used the following prompt:
18
 
19
  ```
20
+ {}
21
  ```
22
 
23
  Where language variable is original language of the text as we wanted to avoid the model translating the text to English during summarization.
 
25
  The model was asked to return the output in three formats: markdown, json and function call. Note that currently function calls are only supported by Open AI API.
26
  To do that we added following text to the query:
27
 
28
+ {}
 
 
29
 
30
  When measuring execution time we used `time.time()` result saved to variable before making the call to API and compared it to `time.time()` result after receiving the results. We used litellm python library for all of the models which naturally adds some overhead compared to pure curl calls.
31
 
 
39
  'hour_range': ["6-8", "9-11", "12-14", "15-17", "18-20", "21-23", "0-2", "3-5"]
40
  })
41
 
42
+ queries_config = QueriesConfig()
43
+
44
+ output_types_df = pd.DataFrame({
45
+ "Output Type": queries_config.query_template.keys(),
46
+ "Added text": queries_config.query_template.values()
47
+ })
48
+
49
  summary_df: pd.DataFrame = pd.read_csv("data/2024-02-05 23:33:22.947120_summary.csv")
50
  time_of_day_comparison_df = pd.read_csv("data/2024-02-06 09:49:19.637072_time_of_day_comparison.csv")
51
+ general_plots = pd.read_csv("data/2024-02-06 13:25:55.133644_general_plot.csv")
52
  model_costs_df = pd.read_csv("data/2024-02-05 12:03:45.281624_model_costs.csv")
53
 
54
 
 
134
  return " ".join(title_words)
135
 
136
 
137
+ filter_textbox = gr.Textbox(label="Model name part", scale=2)
138
+ filter_button = gr.Button("Filter dataframes by model name", scale=1)
139
  collapse_languages_button = gr.Button("Collapse languages")
140
  collapse_output_method_button = gr.Button("Collapse output method")
141
  last_textbox = 0
 
148
  filter_button.render()
149
 
150
  with gr.Tab("About this project"):
151
+ gr.Markdown(README.format(
152
+ queries_config.base_query_template.replace("```", "'''"),
153
+ output_types_df.to_markdown(index=False)
154
+ ))
155
  with gr.Tab("General plots"):
156
  for index, row in general_plots.iterrows():
157
  plot = plotly.io.from_json(row["plot_json"])
158
  plot.update_layout(autosize=True)
159
+ gr.Plot(plot, label=row["header"], scale=1)
160
+ if pd.notna(row["description"]):
161
+ gr.Markdown(str(row["description"]))
162
  with gr.Tab("Output characteristics"):
163
  with gr.Row():
164
  collapse_languages_button.render()
 
171
  time_periods_explanation_ui = gr.DataFrame(dataframe_style(time_periods_explanation_df), label="Times of day ranges")
172
  gr.Markdown("""
173
  These measurements were made by testing the models using the same dataset as in the other comparisons every hour for 24 hours.
174
+
175
+ Execution time refers to averaged time needed to execute one query.
176
 
177
  Hours and times of day in the table and in the plot are based on Central European Time.
178
 
data/2024-02-06 13:25:55.133644_general_plot.csv ADDED
@@ -0,0 +1,284 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ path,plot_object,header,description,plot_json
2
+ ./html/plots/response_times.html,"Figure({
3
+ 'data': [{'alignmentgroup': 'True',
4
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
5
+ 'legendgroup': 'Mistral-7B-Instruct-v0.2',
6
+ 'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
7
+ 'name': 'Mistral-7B-Instruct-v0.2',
8
+ 'offsetgroup': 'Mistral-7B-Instruct-v0.2',
9
+ 'orientation': 'v',
10
+ 'showlegend': True,
11
+ 'textposition': 'auto',
12
+ 'type': 'bar',
13
+ 'x': array(['Mistral-7B-Instruct-v0.2'], dtype=object),
14
+ 'xaxis': 'x',
15
+ 'y': array([3.27102681]),
16
+ 'yaxis': 'y'},
17
+ {'alignmentgroup': 'True',
18
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
19
+ 'legendgroup': 'Mixtral-8x7B-Instruct-v0.1',
20
+ 'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
21
+ 'name': 'Mixtral-8x7B-Instruct-v0.1',
22
+ 'offsetgroup': 'Mixtral-8x7B-Instruct-v0.1',
23
+ 'orientation': 'v',
24
+ 'showlegend': True,
25
+ 'textposition': 'auto',
26
+ 'type': 'bar',
27
+ 'x': array(['Mixtral-8x7B-Instruct-v0.1'], dtype=object),
28
+ 'xaxis': 'x',
29
+ 'y': array([4.83500647]),
30
+ 'yaxis': 'y'},
31
+ {'alignmentgroup': 'True',
32
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
33
+ 'legendgroup': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
34
+ 'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
35
+ 'name': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
36
+ 'offsetgroup': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
37
+ 'orientation': 'v',
38
+ 'showlegend': True,
39
+ 'textposition': 'auto',
40
+ 'type': 'bar',
41
+ 'x': array(['TinyLlama/TinyLlama-1.1B-Chat-v1.0'], dtype=object),
42
+ 'xaxis': 'x',
43
+ 'y': array([1.79273835]),
44
+ 'yaxis': 'y'},
45
+ {'alignmentgroup': 'True',
46
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
47
+ 'legendgroup': 'chat-bison',
48
+ 'marker': {'color': '#ab63fa', 'pattern': {'shape': ''}},
49
+ 'name': 'chat-bison',
50
+ 'offsetgroup': 'chat-bison',
51
+ 'orientation': 'v',
52
+ 'showlegend': True,
53
+ 'textposition': 'auto',
54
+ 'type': 'bar',
55
+ 'x': array(['chat-bison'], dtype=object),
56
+ 'xaxis': 'x',
57
+ 'y': array([3.89024363]),
58
+ 'yaxis': 'y'},
59
+ {'alignmentgroup': 'True',
60
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
61
+ 'legendgroup': 'chat-bison (PaLM 2)',
62
+ 'marker': {'color': '#FFA15A', 'pattern': {'shape': ''}},
63
+ 'name': 'chat-bison (PaLM 2)',
64
+ 'offsetgroup': 'chat-bison (PaLM 2)',
65
+ 'orientation': 'v',
66
+ 'showlegend': True,
67
+ 'textposition': 'auto',
68
+ 'type': 'bar',
69
+ 'x': array(['chat-bison (PaLM 2)'], dtype=object),
70
+ 'xaxis': 'x',
71
+ 'y': array([2.66830386]),
72
+ 'yaxis': 'y'},
73
+ {'alignmentgroup': 'True',
74
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
75
+ 'legendgroup': 'chat-bison-32k',
76
+ 'marker': {'color': '#19d3f3', 'pattern': {'shape': ''}},
77
+ 'name': 'chat-bison-32k',
78
+ 'offsetgroup': 'chat-bison-32k',
79
+ 'orientation': 'v',
80
+ 'showlegend': True,
81
+ 'textposition': 'auto',
82
+ 'type': 'bar',
83
+ 'x': array(['chat-bison-32k'], dtype=object),
84
+ 'xaxis': 'x',
85
+ 'y': array([4.12817777]),
86
+ 'yaxis': 'y'},
87
+ {'alignmentgroup': 'True',
88
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
89
+ 'legendgroup': 'chat-bison-32k (PaLM 2 32K)',
90
+ 'marker': {'color': '#FF6692', 'pattern': {'shape': ''}},
91
+ 'name': 'chat-bison-32k (PaLM 2 32K)',
92
+ 'offsetgroup': 'chat-bison-32k (PaLM 2 32K)',
93
+ 'orientation': 'v',
94
+ 'showlegend': True,
95
+ 'textposition': 'auto',
96
+ 'type': 'bar',
97
+ 'x': array(['chat-bison-32k (PaLM 2 32K)'], dtype=object),
98
+ 'xaxis': 'x',
99
+ 'y': array([6.73886213]),
100
+ 'yaxis': 'y'},
101
+ {'alignmentgroup': 'True',
102
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
103
+ 'legendgroup': 'gemini-pro',
104
+ 'marker': {'color': '#B6E880', 'pattern': {'shape': ''}},
105
+ 'name': 'gemini-pro',
106
+ 'offsetgroup': 'gemini-pro',
107
+ 'orientation': 'v',
108
+ 'showlegend': True,
109
+ 'textposition': 'auto',
110
+ 'type': 'bar',
111
+ 'x': array(['gemini-pro'], dtype=object),
112
+ 'xaxis': 'x',
113
+ 'y': array([3.03678889]),
114
+ 'yaxis': 'y'},
115
+ {'alignmentgroup': 'True',
116
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
117
+ 'legendgroup': 'gpt-3.5-turbo',
118
+ 'marker': {'color': '#FF97FF', 'pattern': {'shape': ''}},
119
+ 'name': 'gpt-3.5-turbo',
120
+ 'offsetgroup': 'gpt-3.5-turbo',
121
+ 'orientation': 'v',
122
+ 'showlegend': True,
123
+ 'textposition': 'auto',
124
+ 'type': 'bar',
125
+ 'x': array(['gpt-3.5-turbo'], dtype=object),
126
+ 'xaxis': 'x',
127
+ 'y': array([4.29307391]),
128
+ 'yaxis': 'y'},
129
+ {'alignmentgroup': 'True',
130
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
131
+ 'legendgroup': 'gpt-4',
132
+ 'marker': {'color': '#FECB52', 'pattern': {'shape': ''}},
133
+ 'name': 'gpt-4',
134
+ 'offsetgroup': 'gpt-4',
135
+ 'orientation': 'v',
136
+ 'showlegend': True,
137
+ 'textposition': 'auto',
138
+ 'type': 'bar',
139
+ 'x': array(['gpt-4'], dtype=object),
140
+ 'xaxis': 'x',
141
+ 'y': array([14.32373439]),
142
+ 'yaxis': 'y'},
143
+ {'alignmentgroup': 'True',
144
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
145
+ 'legendgroup': 'gpt-4-turbo',
146
+ 'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
147
+ 'name': 'gpt-4-turbo',
148
+ 'offsetgroup': 'gpt-4-turbo',
149
+ 'orientation': 'v',
150
+ 'showlegend': True,
151
+ 'textposition': 'auto',
152
+ 'type': 'bar',
153
+ 'x': array(['gpt-4-turbo'], dtype=object),
154
+ 'xaxis': 'x',
155
+ 'y': array([15.5354884]),
156
+ 'yaxis': 'y'},
157
+ {'alignmentgroup': 'True',
158
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
159
+ 'legendgroup': 'llama-2-70b-chat',
160
+ 'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
161
+ 'name': 'llama-2-70b-chat',
162
+ 'offsetgroup': 'llama-2-70b-chat',
163
+ 'orientation': 'v',
164
+ 'showlegend': True,
165
+ 'textposition': 'auto',
166
+ 'type': 'bar',
167
+ 'x': array(['llama-2-70b-chat'], dtype=object),
168
+ 'xaxis': 'x',
169
+ 'y': array([3.6742315]),
170
+ 'yaxis': 'y'},
171
+ {'alignmentgroup': 'True',
172
+ 'hovertemplate': 'Model=%{x}<br>Execution time (s)=%{y}<extra></extra>',
173
+ 'legendgroup': 'zephyr-7b-beta',
174
+ 'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
175
+ 'name': 'zephyr-7b-beta',
176
+ 'offsetgroup': 'zephyr-7b-beta',
177
+ 'orientation': 'v',
178
+ 'showlegend': True,
179
+ 'textposition': 'auto',
180
+ 'type': 'bar',
181
+ 'x': array(['zephyr-7b-beta'], dtype=object),
182
+ 'xaxis': 'x',
183
+ 'y': array([3.6370095]),
184
+ 'yaxis': 'y'}],
185
+ 'layout': {'barmode': 'relative',
186
+ 'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
187
+ 'template': '...',
188
+ 'title': {'text': 'Model endpoints response times'},
189
+ 'xaxis': {'anchor': 'y',
190
+ 'categoryarray': [Mistral-7B-Instruct-v0.2,
191
+ Mixtral-8x7B-Instruct-v0.1,
192
+ TinyLlama/TinyLlama-1.1B-Chat-v1.0,
193
+ chat-bison, chat-bison (PaLM 2), chat-
194
+ bison-32k, chat-bison-32k (PaLM 2 32K),
195
+ gemini-pro, gpt-3.5-turbo, gpt-4,
196
+ gpt-4-turbo, llama-2-70b-chat,
197
+ zephyr-7b-beta],
198
+ 'categoryorder': 'array',
199
+ 'domain': [0.0, 1.0],
200
+ 'title': {'text': 'Model'}},
201
+ 'yaxis': {'anchor': 'x', 'domain': [0.0, 1.0], 'title': {'text': 'Execution time (s)'}}}
202
+ })",Model endpoints response times,"This plot shows averaged response time for one query for each model.
203
+
204
+ We used `time.time()` result saved to variable before making the call to API and compared it to `time.time()` result after receiving the results.
205
+
206
+ It is also important to note that wee used litellm python library for all of the models which adds some overhead compared to pure curl calls.","{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mistral-7B-Instruct-v0.2"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""Mistral-7B-Instruct-v0.2"",""offsetgroup"":""Mistral-7B-Instruct-v0.2"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""Mistral-7B-Instruct-v0.2""],""xaxis"":""x"",""y"":[3.271026806036631],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mixtral-8x7B-Instruct-v0.1"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""Mixtral-8x7B-Instruct-v0.1"",""offsetgroup"":""Mixtral-8x7B-Instruct-v0.1"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""Mixtral-8x7B-Instruct-v0.1""],""xaxis"":""x"",""y"":[4.835006467343727],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""offsetgroup"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0""],""xaxis"":""x"",""y"":[1.7927383464436197],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""chat-bison"",""offsetgroup"":""chat-bison"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""chat-bison""],""xaxis"":""x"",""y"":[3.890243631601334],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison (PaLM 2)"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""chat-bison (PaLM 2)"",""offsetgroup"":""chat-bison (PaLM 2)"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""chat-bison (PaLM 2)""],""xaxis"":""x"",""y"":[2.668303858522159],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison-32k"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""chat-bison-32k"",""offsetgroup"":""chat-bison-32k"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""chat-bison-32k""],""xaxis"":""x"",""y"":[4.128177767992019],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison-32k (PaLM 2 32K)"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""chat-bison-32k (PaLM 2 32K)"",""offsetgroup"":""chat-bison-32k (PaLM 2 32K)"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""chat-bison-32k (PaLM 2 32K)""],""xaxis"":""x"",""y"":[6.738862129823485],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gemini-pro"",""marker"":{""color"":""#B6E880"",""pattern"":{""shape"":""""}},""name"":""gemini-pro"",""offsetgroup"":""gemini-pro"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""gemini-pro""],""xaxis"":""x"",""y"":[3.036788886637026],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-3.5-turbo"",""marker"":{""color"":""#FF97FF"",""pattern"":{""shape"":""""}},""name"":""gpt-3.5-turbo"",""offsetgroup"":""gpt-3.5-turbo"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""gpt-3.5-turbo""],""xaxis"":""x"",""y"":[4.293073908173212],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4"",""marker"":{""color"":""#FECB52"",""pattern"":{""shape"":""""}},""name"":""gpt-4"",""offsetgroup"":""gpt-4"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""gpt-4""],""xaxis"":""x"",""y"":[14.323734385105828],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4-turbo"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""gpt-4-turbo"",""offsetgroup"":""gpt-4-turbo"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""gpt-4-turbo""],""xaxis"":""x"",""y"":[15.535488395149827],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""llama-2-70b-chat"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""llama-2-70b-chat"",""offsetgroup"":""llama-2-70b-chat"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""llama-2-70b-chat""],""xaxis"":""x"",""y"":[3.6742314998615178],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution time (s)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""zephyr-7b-beta"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""zephyr-7b-beta"",""offsetgroup"":""zephyr-7b-beta"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""zephyr-7b-beta""],""xaxis"":""x"",""y"":[3.6370095004383316],""yaxis"":""y"",""type"":""bar""}],""layout"":{""template"":{""data"":{""histogram2dcontour"":[{""type"":""histogram2dcontour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""choropleth"":[{""type"":""choropleth"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""histogram2d"":[{""type"":""histogram2d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmap"":[{""type"":""heatmap"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmapgl"":[{""type"":""heatmapgl"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""contourcarpet"":[{""type"":""contourcarpet"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""contour"":[{""type"":""contour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""surface"":[{""type"":""surface"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""mesh3d"":[{""type"":""mesh3d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""scatter"":[{""fillpattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2},""type"":""scatter""}],""parcoords"":[{""type"":""parcoords"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolargl"":[{""type"":""scatterpolargl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""bar"":[{""error_x"":{""color"":""#2a3f5f""},""error_y"":{""color"":""#2a3f5f""},""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""bar""}],""scattergeo"":[{""type"":""scattergeo"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolar"":[{""type"":""scatterpolar"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""histogram"":[{""marker"":{""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""histogram""}],""scattergl"":[{""type"":""scattergl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatter3d"":[{""type"":""scatter3d"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattermapbox"":[{""type"":""scattermapbox"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterternary"":[{""type"":""scatterternary"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattercarpet"":[{""type"":""scattercarpet"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""carpet"":[{""aaxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""baxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""type"":""carpet""}],""table"":[{""cells"":{""fill"":{""color"":""#EBF0F8""},""line"":{""color"":""white""}},""header"":{""fill"":{""color"":""#C8D4E3""},""line"":{""color"":""white""}},""type"":""table""}],""barpolar"":[{""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""barpolar""}],""pie"":[{""automargin"":true,""type"":""pie""}]},""layout"":{""autotypenumbers"":""strict"",""colorway"":[""#636efa"",""#EF553B"",""#00cc96"",""#ab63fa"",""#FFA15A"",""#19d3f3"",""#FF6692"",""#B6E880"",""#FF97FF"",""#FECB52""],""font"":{""color"":""#2a3f5f""},""hovermode"":""closest"",""hoverlabel"":{""align"":""left""},""paper_bgcolor"":""white"",""plot_bgcolor"":""#E5ECF6"",""polar"":{""bgcolor"":""#E5ECF6"",""angularaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""radialaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""ternary"":{""bgcolor"":""#E5ECF6"",""aaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""baxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""caxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""coloraxis"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""colorscale"":{""sequential"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""sequentialminus"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""diverging"":[[0,""#8e0152""],[0.1,""#c51b7d""],[0.2,""#de77ae""],[0.3,""#f1b6da""],[0.4,""#fde0ef""],[0.5,""#f7f7f7""],[0.6,""#e6f5d0""],[0.7,""#b8e186""],[0.8,""#7fbc41""],[0.9,""#4d9221""],[1,""#276419""]]},""xaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""yaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""scene"":{""xaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""yaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""zaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2}},""shapedefaults"":{""line"":{""color"":""#2a3f5f""}},""annotationdefaults"":{""arrowcolor"":""#2a3f5f"",""arrowhead"":0,""arrowwidth"":1},""geo"":{""bgcolor"":""white"",""landcolor"":""#E5ECF6"",""subunitcolor"":""white"",""showland"":true,""showlakes"":true,""lakecolor"":""white""},""title"":{""x"":0.05},""mapbox"":{""style"":""light""}}},""xaxis"":{""anchor"":""y"",""domain"":[0.0,1.0],""title"":{""text"":""Model""},""categoryorder"":""array"",""categoryarray"":[""Mistral-7B-Instruct-v0.2"",""Mixtral-8x7B-Instruct-v0.1"",""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""chat-bison"",""chat-bison (PaLM 2)"",""chat-bison-32k"",""chat-bison-32k (PaLM 2 32K)"",""gemini-pro"",""gpt-3.5-turbo"",""gpt-4"",""gpt-4-turbo"",""llama-2-70b-chat"",""zephyr-7b-beta""]},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Execution time (s)""}},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Model endpoints response times""},""barmode"":""relative""}}"
207
+ ./html/plots/execution_costs.html,"Figure({
208
+ 'data': [{'alignmentgroup': 'True',
209
+ 'hovertemplate': 'Model=%{x}<br>Execution cost ($)=%{y}<extra></extra>',
210
+ 'legendgroup': 'Mixtral-8x7B-Instruct-v0.1',
211
+ 'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
212
+ 'name': 'Mixtral-8x7B-Instruct-v0.1',
213
+ 'offsetgroup': 'Mixtral-8x7B-Instruct-v0.1',
214
+ 'orientation': 'v',
215
+ 'showlegend': True,
216
+ 'textposition': 'auto',
217
+ 'type': 'bar',
218
+ 'x': array(['Mixtral-8x7B-Instruct-v0.1'], dtype=object),
219
+ 'xaxis': 'x',
220
+ 'y': array([0.072612]),
221
+ 'yaxis': 'y'},
222
+ {'alignmentgroup': 'True',
223
+ 'hovertemplate': 'Model=%{x}<br>Execution cost ($)=%{y}<extra></extra>',
224
+ 'legendgroup': 'zephyr-7b-beta',
225
+ 'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
226
+ 'name': 'zephyr-7b-beta',
227
+ 'offsetgroup': 'zephyr-7b-beta',
228
+ 'orientation': 'v',
229
+ 'showlegend': True,
230
+ 'textposition': 'auto',
231
+ 'type': 'bar',
232
+ 'x': array(['zephyr-7b-beta'], dtype=object),
233
+ 'xaxis': 'x',
234
+ 'y': array([0.05180257]),
235
+ 'yaxis': 'y'}],
236
+ 'layout': {'barmode': 'relative',
237
+ 'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
238
+ 'template': '...',
239
+ 'title': {'text': 'Costs of execution of 800 test queries per model'},
240
+ 'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Model'}},
241
+ 'yaxis': {'anchor': 'x', 'domain': [0.0, 1.0], 'title': {'text': 'Execution cost ($)'}}}
242
+ })",Costs of execution of 800 test queries per model,"Costs of experiments carried out to create this table, 800 test queries per model.
243
+
244
+ Note that pause and resume time cost was not included in the cost calculation for the models billed hourly.
245
+
246
+ Also although Gemini-Pro model is free for under 60 queries per minute we calculated the full cost for generating the tokens.","{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution cost ($)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mixtral-8x7B-Instruct-v0.1"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""Mixtral-8x7B-Instruct-v0.1"",""offsetgroup"":""Mixtral-8x7B-Instruct-v0.1"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""Mixtral-8x7B-Instruct-v0.1""],""xaxis"":""x"",""y"":[0.07261200000000001],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eExecution cost ($)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""zephyr-7b-beta"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""zephyr-7b-beta"",""offsetgroup"":""zephyr-7b-beta"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""zephyr-7b-beta""],""xaxis"":""x"",""y"":[0.05180256798532274],""yaxis"":""y"",""type"":""bar""}],""layout"":{""template"":{""data"":{""histogram2dcontour"":[{""type"":""histogram2dcontour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""choropleth"":[{""type"":""choropleth"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""histogram2d"":[{""type"":""histogram2d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmap"":[{""type"":""heatmap"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmapgl"":[{""type"":""heatmapgl"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""contourcarpet"":[{""type"":""contourcarpet"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""contour"":[{""type"":""contour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""surface"":[{""type"":""surface"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""mesh3d"":[{""type"":""mesh3d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""scatter"":[{""fillpattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2},""type"":""scatter""}],""parcoords"":[{""type"":""parcoords"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolargl"":[{""type"":""scatterpolargl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""bar"":[{""error_x"":{""color"":""#2a3f5f""},""error_y"":{""color"":""#2a3f5f""},""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""bar""}],""scattergeo"":[{""type"":""scattergeo"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolar"":[{""type"":""scatterpolar"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""histogram"":[{""marker"":{""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""histogram""}],""scattergl"":[{""type"":""scattergl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatter3d"":[{""type"":""scatter3d"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattermapbox"":[{""type"":""scattermapbox"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterternary"":[{""type"":""scatterternary"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattercarpet"":[{""type"":""scattercarpet"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""carpet"":[{""aaxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""baxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""type"":""carpet""}],""table"":[{""cells"":{""fill"":{""color"":""#EBF0F8""},""line"":{""color"":""white""}},""header"":{""fill"":{""color"":""#C8D4E3""},""line"":{""color"":""white""}},""type"":""table""}],""barpolar"":[{""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""barpolar""}],""pie"":[{""automargin"":true,""type"":""pie""}]},""layout"":{""autotypenumbers"":""strict"",""colorway"":[""#636efa"",""#EF553B"",""#00cc96"",""#ab63fa"",""#FFA15A"",""#19d3f3"",""#FF6692"",""#B6E880"",""#FF97FF"",""#FECB52""],""font"":{""color"":""#2a3f5f""},""hovermode"":""closest"",""hoverlabel"":{""align"":""left""},""paper_bgcolor"":""white"",""plot_bgcolor"":""#E5ECF6"",""polar"":{""bgcolor"":""#E5ECF6"",""angularaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""radialaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""ternary"":{""bgcolor"":""#E5ECF6"",""aaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""baxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""caxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""coloraxis"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""colorscale"":{""sequential"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""sequentialminus"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""diverging"":[[0,""#8e0152""],[0.1,""#c51b7d""],[0.2,""#de77ae""],[0.3,""#f1b6da""],[0.4,""#fde0ef""],[0.5,""#f7f7f7""],[0.6,""#e6f5d0""],[0.7,""#b8e186""],[0.8,""#7fbc41""],[0.9,""#4d9221""],[1,""#276419""]]},""xaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""yaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""scene"":{""xaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""yaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""zaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2}},""shapedefaults"":{""line"":{""color"":""#2a3f5f""}},""annotationdefaults"":{""arrowcolor"":""#2a3f5f"",""arrowhead"":0,""arrowwidth"":1},""geo"":{""bgcolor"":""white"",""landcolor"":""#E5ECF6"",""subunitcolor"":""white"",""showland"":true,""showlakes"":true,""lakecolor"":""white""},""title"":{""x"":0.05},""mapbox"":{""style"":""light""}}},""xaxis"":{""anchor"":""y"",""domain"":[0.0,1.0],""title"":{""text"":""Model""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Execution cost ($)""}},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Costs of execution of 800 test queries per model""},""barmode"":""relative""}}"
247
+ ./html/plots/model_sizes.html,"Figure({
248
+ 'data': [{'alignmentgroup': 'True',
249
+ 'hovertemplate': 'Model=%{x}<br>Model size (B params)=%{y}<extra></extra>',
250
+ 'legendgroup': 'Mixtral-8x7B-Instruct-v0.1',
251
+ 'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
252
+ 'name': 'Mixtral-8x7B-Instruct-v0.1',
253
+ 'offsetgroup': 'Mixtral-8x7B-Instruct-v0.1',
254
+ 'orientation': 'v',
255
+ 'showlegend': True,
256
+ 'textposition': 'auto',
257
+ 'type': 'bar',
258
+ 'x': array(['Mixtral-8x7B-Instruct-v0.1'], dtype=object),
259
+ 'xaxis': 'x',
260
+ 'y': array([56]),
261
+ 'yaxis': 'y'},
262
+ {'alignmentgroup': 'True',
263
+ 'hovertemplate': 'Model=%{x}<br>Model size (B params)=%{y}<extra></extra>',
264
+ 'legendgroup': 'zephyr-7b-beta',
265
+ 'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
266
+ 'name': 'zephyr-7b-beta',
267
+ 'offsetgroup': 'zephyr-7b-beta',
268
+ 'orientation': 'v',
269
+ 'showlegend': True,
270
+ 'textposition': 'auto',
271
+ 'type': 'bar',
272
+ 'x': array(['zephyr-7b-beta'], dtype=object),
273
+ 'xaxis': 'x',
274
+ 'y': array([7]),
275
+ 'yaxis': 'y'}],
276
+ 'layout': {'barmode': 'relative',
277
+ 'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
278
+ 'template': '...',
279
+ 'title': {'text': 'Model sizes in billions of parameters'},
280
+ 'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Model'}},
281
+ 'yaxis': {'anchor': 'x', 'domain': [0.0, 1.0], 'title': {'text': 'Model size (B params)'}}}
282
+ })",Model sizes in billions of parameters,"Model sizes based on release blogs and documentation of their respective developers.
283
+
284
+ Note that models, which sizes were not disclosed publicly were omitted from this comparison.","{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eModel size (B params)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mixtral-8x7B-Instruct-v0.1"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""Mixtral-8x7B-Instruct-v0.1"",""offsetgroup"":""Mixtral-8x7B-Instruct-v0.1"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""Mixtral-8x7B-Instruct-v0.1""],""xaxis"":""x"",""y"":[56],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{x}\u003cbr\u003eModel size (B params)=%{y}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""zephyr-7b-beta"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""zephyr-7b-beta"",""offsetgroup"":""zephyr-7b-beta"",""orientation"":""v"",""showlegend"":true,""textposition"":""auto"",""x"":[""zephyr-7b-beta""],""xaxis"":""x"",""y"":[7],""yaxis"":""y"",""type"":""bar""}],""layout"":{""template"":{""data"":{""histogram2dcontour"":[{""type"":""histogram2dcontour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""choropleth"":[{""type"":""choropleth"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""histogram2d"":[{""type"":""histogram2d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmap"":[{""type"":""heatmap"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""heatmapgl"":[{""type"":""heatmapgl"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""contourcarpet"":[{""type"":""contourcarpet"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""contour"":[{""type"":""contour"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""surface"":[{""type"":""surface"",""colorbar"":{""outlinewidth"":0,""ticks"":""""},""colorscale"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]]}],""mesh3d"":[{""type"":""mesh3d"",""colorbar"":{""outlinewidth"":0,""ticks"":""""}}],""scatter"":[{""fillpattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2},""type"":""scatter""}],""parcoords"":[{""type"":""parcoords"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolargl"":[{""type"":""scatterpolargl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""bar"":[{""error_x"":{""color"":""#2a3f5f""},""error_y"":{""color"":""#2a3f5f""},""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""bar""}],""scattergeo"":[{""type"":""scattergeo"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterpolar"":[{""type"":""scatterpolar"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""histogram"":[{""marker"":{""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""histogram""}],""scattergl"":[{""type"":""scattergl"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatter3d"":[{""type"":""scatter3d"",""line"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattermapbox"":[{""type"":""scattermapbox"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scatterternary"":[{""type"":""scatterternary"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""scattercarpet"":[{""type"":""scattercarpet"",""marker"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}}}],""carpet"":[{""aaxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""baxis"":{""endlinecolor"":""#2a3f5f"",""gridcolor"":""white"",""linecolor"":""white"",""minorgridcolor"":""white"",""startlinecolor"":""#2a3f5f""},""type"":""carpet""}],""table"":[{""cells"":{""fill"":{""color"":""#EBF0F8""},""line"":{""color"":""white""}},""header"":{""fill"":{""color"":""#C8D4E3""},""line"":{""color"":""white""}},""type"":""table""}],""barpolar"":[{""marker"":{""line"":{""color"":""#E5ECF6"",""width"":0.5},""pattern"":{""fillmode"":""overlay"",""size"":10,""solidity"":0.2}},""type"":""barpolar""}],""pie"":[{""automargin"":true,""type"":""pie""}]},""layout"":{""autotypenumbers"":""strict"",""colorway"":[""#636efa"",""#EF553B"",""#00cc96"",""#ab63fa"",""#FFA15A"",""#19d3f3"",""#FF6692"",""#B6E880"",""#FF97FF"",""#FECB52""],""font"":{""color"":""#2a3f5f""},""hovermode"":""closest"",""hoverlabel"":{""align"":""left""},""paper_bgcolor"":""white"",""plot_bgcolor"":""#E5ECF6"",""polar"":{""bgcolor"":""#E5ECF6"",""angularaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""radialaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""ternary"":{""bgcolor"":""#E5ECF6"",""aaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""baxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""},""caxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":""""}},""coloraxis"":{""colorbar"":{""outlinewidth"":0,""ticks"":""""}},""colorscale"":{""sequential"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""sequentialminus"":[[0.0,""#0d0887""],[0.1111111111111111,""#46039f""],[0.2222222222222222,""#7201a8""],[0.3333333333333333,""#9c179e""],[0.4444444444444444,""#bd3786""],[0.5555555555555556,""#d8576b""],[0.6666666666666666,""#ed7953""],[0.7777777777777778,""#fb9f3a""],[0.8888888888888888,""#fdca26""],[1.0,""#f0f921""]],""diverging"":[[0,""#8e0152""],[0.1,""#c51b7d""],[0.2,""#de77ae""],[0.3,""#f1b6da""],[0.4,""#fde0ef""],[0.5,""#f7f7f7""],[0.6,""#e6f5d0""],[0.7,""#b8e186""],[0.8,""#7fbc41""],[0.9,""#4d9221""],[1,""#276419""]]},""xaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""yaxis"":{""gridcolor"":""white"",""linecolor"":""white"",""ticks"":"""",""title"":{""standoff"":15},""zerolinecolor"":""white"",""automargin"":true,""zerolinewidth"":2},""scene"":{""xaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""yaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2},""zaxis"":{""backgroundcolor"":""#E5ECF6"",""gridcolor"":""white"",""linecolor"":""white"",""showbackground"":true,""ticks"":"""",""zerolinecolor"":""white"",""gridwidth"":2}},""shapedefaults"":{""line"":{""color"":""#2a3f5f""}},""annotationdefaults"":{""arrowcolor"":""#2a3f5f"",""arrowhead"":0,""arrowwidth"":1},""geo"":{""bgcolor"":""white"",""landcolor"":""#E5ECF6"",""subunitcolor"":""white"",""showland"":true,""showlakes"":true,""lakecolor"":""white""},""title"":{""x"":0.05},""mapbox"":{""style"":""light""}}},""xaxis"":{""anchor"":""y"",""domain"":[0.0,1.0],""title"":{""text"":""Model""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Model size (B params)""}},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Model sizes in billions of parameters""},""barmode"":""relative""}}"