piotr-szleg-bards-ai commited on
Commit
95117cd
·
1 Parent(s): 74c94d1

2024-02-15 13:26:46 Publish script update

Browse files
Files changed (4) hide show
  1. app.py +26 -34
  2. data/general_plots.csv +55 -39
  3. data/time_of_day_plots.csv +0 -0
  4. pipeline/models.py +70 -56
app.py CHANGED
@@ -1,12 +1,11 @@
 
 
1
  import re
2
 
3
  import gradio as gr
4
  import pandas as pd
5
  import plotly
6
- import io
7
  import plotly.express as px
8
- import json
9
-
10
  from pandas.api.types import is_numeric_dtype
11
 
12
  from pipeline.config import LLMBoardConfig, QueriesConfig
@@ -120,15 +119,12 @@ def dataframes():
120
  def for_dataframe(df):
121
  if not searched_model_names:
122
  return df
123
- return dataframe_style(pd.concat(
124
- df[df.model.str.lower().str.contains(n)]
125
- for n in searched_model_names
126
- ))
127
 
128
  return (
129
  for_dataframe(summary_df_processed),
130
  for_dataframe(time_of_day_comparison_df),
131
- for_dataframe(model_costs_df)
132
  )
133
 
134
 
@@ -161,7 +157,8 @@ last_textbox = 0
161
  plots = []
162
  single_model_plots = []
163
 
164
- def filter_plots(searched_query:str):
 
165
  searched_model_names = searched_query.split("|")
166
  searched_model_names = [n.lower().strip() for n in searched_model_names]
167
  searched_model_names = [n for n in searched_model_names if n]
@@ -169,11 +166,8 @@ def filter_plots(searched_query:str):
169
  def filter_dataframe(df):
170
  if not searched_model_names:
171
  return df
172
- return pd.concat(
173
- df[df.model.str.lower().str.contains(n)]
174
- for n in searched_model_names
175
- )
176
-
177
  results = []
178
  for plot_display, plot, row in plots:
179
  visible = True
@@ -181,31 +175,27 @@ def filter_plots(searched_query:str):
181
  buffer = io.StringIO(row["df"])
182
  df = pd.read_csv(buffer)
183
  df = filter_dataframe(df)
184
- plot = px.bar(
185
- df,
186
- **json.loads(row["arguments"])
187
- )
188
  plot.update_layout(autosize=True)
189
  elif "for model" in row["header"] and searched_model_names:
190
  plot_model = row["header"].split("for model")[1].lower()
191
  if not any(n in plot_model for n in searched_model_names):
192
  visible = False
193
-
194
  results.append(gr.Plot(plot, visible=visible))
195
 
196
  return results
197
 
 
198
  def display_plot(plot_df_row):
199
  row = dict(plot_df_row)
200
  plot = plotly.io.from_json(row["plot_json"])
201
  plot.update_layout(autosize=True)
202
- plots.append((
203
- gr.Plot(plot, label=row["header"], scale=1),
204
- plot,
205
- row))
206
  if "description" in row and pd.notna(row["description"]):
207
  gr.Markdown(str(row["description"]))
208
 
 
209
  with gr.Blocks() as demo:
210
  gr.HTML("<h1>Performance LLM Board</h1>")
211
 
@@ -213,7 +203,7 @@ with gr.Blocks() as demo:
213
  filter_textbox.render()
214
  filter_button.render()
215
  gr.Markdown(
216
- "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;\* You can use `|` operator to display multiple models at once, for example \"gpt|mistral|zephyr\""
217
  )
218
 
219
  with gr.Tab("About this project"):
@@ -238,8 +228,6 @@ Execution time refers to averaged time needed to execute one query.
238
  Hours and times of day in the table and in the plot are based on Central European Time.
239
 
240
  Measurements were made during a normal work week.
241
-
242
- To compare providers we checked execution times for the same model `Mistral (7B) Instruct v0.2` on both Hugging Face and Together AI.
243
  """
244
  )
245
  # display rest of the plots
@@ -272,14 +260,18 @@ for models hosted this way we calculated "Cost Per Token" column using data coll
272
  Note that pause and resume time cost was not included in the "Cost Per Token" column calculation.
273
  """
274
  )
275
- general_plots[general_plots.plot_name=="execution_costs"].apply(display_plot, axis=1)
276
- with gr.Tab("Model dimensions"):
277
- general_plots[general_plots.plot_name!="execution_costs"].apply(display_plot, axis=1)
278
- gr.Markdown("""
279
- Model dimensions based on release blogs and documentation of their respective developers.
280
-
281
- Note that models, which dimensions were not disclosed publicly, were omitted from the plots.
282
- """)
 
 
 
 
283
  filter_button.click(
284
  fn=filter_dataframes,
285
  inputs=filter_textbox,
 
1
+ import io
2
+ import json
3
  import re
4
 
5
  import gradio as gr
6
  import pandas as pd
7
  import plotly
 
8
  import plotly.express as px
 
 
9
  from pandas.api.types import is_numeric_dtype
10
 
11
  from pipeline.config import LLMBoardConfig, QueriesConfig
 
119
  def for_dataframe(df):
120
  if not searched_model_names:
121
  return df
122
+ return dataframe_style(pd.concat(df[df.model.str.lower().str.contains(n)] for n in searched_model_names))
 
 
 
123
 
124
  return (
125
  for_dataframe(summary_df_processed),
126
  for_dataframe(time_of_day_comparison_df),
127
+ for_dataframe(model_costs_df),
128
  )
129
 
130
 
 
157
  plots = []
158
  single_model_plots = []
159
 
160
+
161
+ def filter_plots(searched_query: str):
162
  searched_model_names = searched_query.split("|")
163
  searched_model_names = [n.lower().strip() for n in searched_model_names]
164
  searched_model_names = [n for n in searched_model_names if n]
 
166
  def filter_dataframe(df):
167
  if not searched_model_names:
168
  return df
169
+ return pd.concat(df[df.model.str.lower().str.contains(n)] for n in searched_model_names)
170
+
 
 
 
171
  results = []
172
  for plot_display, plot, row in plots:
173
  visible = True
 
175
  buffer = io.StringIO(row["df"])
176
  df = pd.read_csv(buffer)
177
  df = filter_dataframe(df)
178
+ plot = px.bar(df, **json.loads(row["arguments"]))
 
 
 
179
  plot.update_layout(autosize=True)
180
  elif "for model" in row["header"] and searched_model_names:
181
  plot_model = row["header"].split("for model")[1].lower()
182
  if not any(n in plot_model for n in searched_model_names):
183
  visible = False
184
+
185
  results.append(gr.Plot(plot, visible=visible))
186
 
187
  return results
188
 
189
+
190
  def display_plot(plot_df_row):
191
  row = dict(plot_df_row)
192
  plot = plotly.io.from_json(row["plot_json"])
193
  plot.update_layout(autosize=True)
194
+ plots.append((gr.Plot(plot, label=row["header"], scale=1), plot, row))
 
 
 
195
  if "description" in row and pd.notna(row["description"]):
196
  gr.Markdown(str(row["description"]))
197
 
198
+
199
  with gr.Blocks() as demo:
200
  gr.HTML("<h1>Performance LLM Board</h1>")
201
 
 
203
  filter_textbox.render()
204
  filter_button.render()
205
  gr.Markdown(
206
+ '&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;\* You can use `|` operator to display multiple models at once, for example "gpt|mistral|zephyr"'
207
  )
208
 
209
  with gr.Tab("About this project"):
 
228
  Hours and times of day in the table and in the plot are based on Central European Time.
229
 
230
  Measurements were made during a normal work week.
 
 
231
  """
232
  )
233
  # display rest of the plots
 
260
  Note that pause and resume time cost was not included in the "Cost Per Token" column calculation.
261
  """
262
  )
263
+ general_plots[general_plots.plot_name == "execution_costs"].apply(display_plot, axis=1)
264
+ with gr.Tab("Context length and parameters count"):
265
+ general_plots[general_plots.plot_name != "execution_costs"].apply(display_plot, axis=1)
266
+ gr.Markdown(
267
+ """
268
+ LLM models context length and parameters count are based on release blogs and documentation of their respective developers.
269
+
270
+ A lot of models had to be omitted due to their developers not disclosing their parameters count.
271
+
272
+ Mainly OpenAI's GPT models and Google's Palm 2.
273
+ """
274
+ )
275
  filter_button.click(
276
  fn=filter_dataframes,
277
  inputs=filter_textbox,
data/general_plots.csv CHANGED
@@ -10,7 +10,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
10
  'showlegend': True,
11
  'textposition': 'auto',
12
  'type': 'bar',
13
- 'x': array([0.0249]),
14
  'xaxis': 'x',
15
  'y': array(['gpt-4'], dtype=object),
16
  'yaxis': 'y'},
@@ -24,7 +24,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
24
  'showlegend': True,
25
  'textposition': 'auto',
26
  'type': 'bar',
27
- 'x': array([0.0229]),
28
  'xaxis': 'x',
29
  'y': array(['gpt-4-turbo'], dtype=object),
30
  'yaxis': 'y'},
@@ -38,7 +38,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
38
  'showlegend': True,
39
  'textposition': 'auto',
40
  'type': 'bar',
41
- 'x': array([0.00083]),
42
  'xaxis': 'x',
43
  'y': array(['gpt-3.5-turbo'], dtype=object),
44
  'yaxis': 'y'},
@@ -52,7 +52,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
52
  'showlegend': True,
53
  'textposition': 'auto',
54
  'type': 'bar',
55
- 'x': array([0.000711]),
56
  'xaxis': 'x',
57
  'y': array(['llama-2-70b-chat'], dtype=object),
58
  'yaxis': 'y'},
@@ -66,7 +66,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
66
  'showlegend': True,
67
  'textposition': 'auto',
68
  'type': 'bar',
69
- 'x': array([0.000711]),
70
  'xaxis': 'x',
71
  'y': array(['Mixtral-8x7B-Instruct-v0.1'], dtype=object),
72
  'yaxis': 'y'},
@@ -80,7 +80,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
80
  'showlegend': True,
81
  'textposition': 'auto',
82
  'type': 'bar',
83
- 'x': array([0.0001975]),
84
  'xaxis': 'x',
85
  'y': array(['gemini-pro'], dtype=object),
86
  'yaxis': 'y'},
@@ -94,7 +94,7 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
94
  'showlegend': True,
95
  'textposition': 'auto',
96
  'type': 'bar',
97
- 'x': array([0.0001975]),
98
  'xaxis': 'x',
99
  'y': array(['chat-bison (PaLM 2)'], dtype=object),
100
  'yaxis': 'y'},
@@ -108,9 +108,23 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
108
  'showlegend': True,
109
  'textposition': 'auto',
110
  'type': 'bar',
111
- 'x': array([0.0001975]),
112
  'xaxis': 'x',
113
  'y': array(['chat-bison-32k (PaLM 2 32K)'], dtype=object),
 
 
 
 
 
 
 
 
 
 
 
 
 
 
114
  'yaxis': 'y'}],
115
  'layout': {'barmode': 'relative',
116
  'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
@@ -118,7 +132,8 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
118
  'title': {'text': 'Costs of execution of 20 test queries per model'},
119
  'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Execution cost ($)'}},
120
  'yaxis': {'anchor': 'x',
121
- 'categoryarray': [chat-bison-32k (PaLM 2 32K), chat-bison
 
122
  (PaLM 2), gemini-pro,
123
  Mixtral-8x7B-Instruct-v0.1,
124
  llama-2-70b-chat, gpt-3.5-turbo,
@@ -126,15 +141,16 @@ execution_costs,./html/plots/execution_costs.html,"Figure({
126
  'categoryorder': 'array',
127
  'domain': [0.0, 1.0],
128
  'title': {'text': 'Model'}}}
129
- })",Costs of execution of 20 test queries per model,,"{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""gpt-4"",""offsetgroup"":""gpt-4"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0249],""xaxis"":""x"",""y"":[""gpt-4""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4-turbo"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""gpt-4-turbo"",""offsetgroup"":""gpt-4-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0229],""xaxis"":""x"",""y"":[""gpt-4-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-3.5-turbo"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""gpt-3.5-turbo"",""offsetgroup"":""gpt-3.5-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.00083],""xaxis"":""x"",""y"":[""gpt-3.5-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""llama-2-70b-chat"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""llama-2-70b-chat"",""offsetgroup"":""llama-2-70b-chat"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.000711],""xaxis"":""x"",""y"":[""llama-2-70b-chat""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mixtral-8x7B-Instruct-v0.1"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""Mixtral-8x7B-Instruct-v0.1"",""offsetgroup"":""Mixtral-8x7B-Instruct-v0.1"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.000711],""xaxis"":""x"",""y"":[""Mixtral-8x7B-Instruct-v0.1""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gemini-pro"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""gemini-pro"",""offsetgroup"":""gemini-pro"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0001975],""xaxis"":""x"",""y"":[""gemini-pro""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison (PaLM 2)"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""chat-bison (PaLM 2)"",""offsetgroup"":""chat-bison (PaLM 2)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0001975],""xaxis"":""x"",""y"":[""chat-bison (PaLM 2)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison-32k (PaLM 2 32K)"",""marker"":{""color"":""#B6E880"",""pattern"":{""shape"":""""}},""name"":""chat-bison-32k (PaLM 2 32K)"",""offsetgroup"":""chat-bison-32k (PaLM 2 32K)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0001975],""xaxis"":""x"",""y"":[""chat-bison-32k (PaLM 2 32K)""],""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"":""Execution cost ($)""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Model""},""categoryorder"":""array"",""categoryarray"":[""chat-bison-32k (PaLM 2 32K)"",""chat-bison (PaLM 2)"",""gemini-pro"",""Mixtral-8x7B-Instruct-v0.1"",""llama-2-70b-chat"",""gpt-3.5-turbo"",""gpt-4-turbo"",""gpt-4""]},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Costs of execution of 20 test queries per model""},""barmode"":""relative""}}","{""y"": ""model"", ""x"": ""model_query_costs"", ""color"": ""model"", ""orientation"": ""h"", ""title"": ""Costs of execution of 20 test queries per model"", ""labels"": {""model"": ""Model"", ""model_query_costs"": ""Execution cost ($)""}}",",model_query_costs,model
130
- 2,0.0249,gpt-4
131
- 1,0.0229,gpt-4-turbo
132
- 0,0.00083,gpt-3.5-turbo
133
- 3,0.000711,llama-2-70b-chat
134
- 4,0.000711,Mixtral-8x7B-Instruct-v0.1
135
- 8,0.0001975,gemini-pro
136
- 9,0.0001975,chat-bison (PaLM 2)
137
- 10,0.0001975,chat-bison-32k (PaLM 2 32K)
 
138
  "
139
  model_sizes,./html/plots/model_sizes.html,"Figure({
140
  'data': [{'alignmentgroup': 'True',
@@ -337,9 +353,9 @@ model_sizes,./html/plots/model_sizes.html,"Figure({
337
  9,7.0,RedPajama-INCITE Chat (7B)
338
  4,1.1,TinyLlama/TinyLlama-1.1B-Chat-v1.0
339
  "
340
- model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
341
  'data': [{'alignmentgroup': 'True',
342
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
343
  'legendgroup': 'gpt-4-turbo',
344
  'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
345
  'name': 'gpt-4-turbo',
@@ -353,7 +369,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
353
  'y': array(['gpt-4-turbo'], dtype=object),
354
  'yaxis': 'y'},
355
  {'alignmentgroup': 'True',
356
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
357
  'legendgroup': 'Mistral (7B) Instruct v0.2 (Together AI)',
358
  'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
359
  'name': 'Mistral (7B) Instruct v0.2 (Together AI)',
@@ -367,7 +383,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
367
  'y': array(['Mistral (7B) Instruct v0.2 (Together AI)'], dtype=object),
368
  'yaxis': 'y'},
369
  {'alignmentgroup': 'True',
370
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
371
  'legendgroup': 'Snorkel Mistral PairRM DPO (7B)',
372
  'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
373
  'name': 'Snorkel Mistral PairRM DPO (7B)',
@@ -381,7 +397,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
381
  'y': array(['Snorkel Mistral PairRM DPO (7B)'], dtype=object),
382
  'yaxis': 'y'},
383
  {'alignmentgroup': 'True',
384
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
385
  'legendgroup': 'Qwen 1.5 Chat (7B)',
386
  'marker': {'color': '#ab63fa', 'pattern': {'shape': ''}},
387
  'name': 'Qwen 1.5 Chat (7B)',
@@ -395,7 +411,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
395
  'y': array(['Qwen 1.5 Chat (7B)'], dtype=object),
396
  'yaxis': 'y'},
397
  {'alignmentgroup': 'True',
398
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
399
  'legendgroup': 'gpt-4',
400
  'marker': {'color': '#FFA15A', 'pattern': {'shape': ''}},
401
  'name': 'gpt-4',
@@ -409,7 +425,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
409
  'y': array(['gpt-4'], dtype=object),
410
  'yaxis': 'y'},
411
  {'alignmentgroup': 'True',
412
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
413
  'legendgroup': 'chat-bison (PaLM 2)',
414
  'marker': {'color': '#19d3f3', 'pattern': {'shape': ''}},
415
  'name': 'chat-bison (PaLM 2)',
@@ -423,7 +439,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
423
  'y': array(['chat-bison (PaLM 2)'], dtype=object),
424
  'yaxis': 'y'},
425
  {'alignmentgroup': 'True',
426
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
427
  'legendgroup': 'OpenHermes-2.5-Mistral (7B)',
428
  'marker': {'color': '#FF6692', 'pattern': {'shape': ''}},
429
  'name': 'OpenHermes-2.5-Mistral (7B)',
@@ -437,7 +453,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
437
  'y': array(['OpenHermes-2.5-Mistral (7B)'], dtype=object),
438
  'yaxis': 'y'},
439
  {'alignmentgroup': 'True',
440
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
441
  'legendgroup': 'gpt-3.5-turbo',
442
  'marker': {'color': '#B6E880', 'pattern': {'shape': ''}},
443
  'name': 'gpt-3.5-turbo',
@@ -451,7 +467,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
451
  'y': array(['gpt-3.5-turbo'], dtype=object),
452
  'yaxis': 'y'},
453
  {'alignmentgroup': 'True',
454
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
455
  'legendgroup': 'WizardLM v1.2 (13B)',
456
  'marker': {'color': '#FF97FF', 'pattern': {'shape': ''}},
457
  'name': 'WizardLM v1.2 (13B)',
@@ -465,7 +481,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
465
  'y': array(['WizardLM v1.2 (13B)'], dtype=object),
466
  'yaxis': 'y'},
467
  {'alignmentgroup': 'True',
468
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
469
  'legendgroup': 'LLaMA-2 Chat (7B)',
470
  'marker': {'color': '#FECB52', 'pattern': {'shape': ''}},
471
  'name': 'LLaMA-2 Chat (7B)',
@@ -479,7 +495,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
479
  'y': array(['LLaMA-2 Chat (7B)'], dtype=object),
480
  'yaxis': 'y'},
481
  {'alignmentgroup': 'True',
482
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
483
  'legendgroup': 'Vicuna v1.5 (7B)',
484
  'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
485
  'name': 'Vicuna v1.5 (7B)',
@@ -493,7 +509,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
493
  'y': array(['Vicuna v1.5 (7B)'], dtype=object),
494
  'yaxis': 'y'},
495
  {'alignmentgroup': 'True',
496
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
497
  'legendgroup': '01-ai Yi Chat (34B)',
498
  'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
499
  'name': '01-ai Yi Chat (34B)',
@@ -507,7 +523,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
507
  'y': array(['01-ai Yi Chat (34B)'], dtype=object),
508
  'yaxis': 'y'},
509
  {'alignmentgroup': 'True',
510
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
511
  'legendgroup': 'Upstage SOLAR Instruct v1 (11B)',
512
  'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
513
  'name': 'Upstage SOLAR Instruct v1 (11B)',
@@ -521,7 +537,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
521
  'y': array(['Upstage SOLAR Instruct v1 (11B)'], dtype=object),
522
  'yaxis': 'y'},
523
  {'alignmentgroup': 'True',
524
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
525
  'legendgroup': 'Chronos Hermes (13B)',
526
  'marker': {'color': '#ab63fa', 'pattern': {'shape': ''}},
527
  'name': 'Chronos Hermes (13B)',
@@ -535,7 +551,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
535
  'y': array(['Chronos Hermes (13B)'], dtype=object),
536
  'yaxis': 'y'},
537
  {'alignmentgroup': 'True',
538
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
539
  'legendgroup': 'Falcon Instruct (7B)',
540
  'marker': {'color': '#FFA15A', 'pattern': {'shape': ''}},
541
  'name': 'Falcon Instruct (7B)',
@@ -549,7 +565,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
549
  'y': array(['Falcon Instruct (7B)'], dtype=object),
550
  'yaxis': 'y'},
551
  {'alignmentgroup': 'True',
552
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
553
  'legendgroup': 'RedPajama-INCITE Chat (7B)',
554
  'marker': {'color': '#19d3f3', 'pattern': {'shape': ''}},
555
  'name': 'RedPajama-INCITE Chat (7B)',
@@ -563,7 +579,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
563
  'y': array(['RedPajama-INCITE Chat (7B)'], dtype=object),
564
  'yaxis': 'y'},
565
  {'alignmentgroup': 'True',
566
- 'hovertemplate': 'Model=%{y}<br>Model input size (tokens)=%{x}<extra></extra>',
567
  'legendgroup': 'llama-2-70b-chat',
568
  'marker': {'color': '#FF6692', 'pattern': {'shape': ''}},
569
  'name': 'llama-2-70b-chat',
@@ -579,8 +595,8 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
579
  'layout': {'barmode': 'relative',
580
  'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
581
  'template': '...',
582
- 'title': {'text': 'Model input sizes in tokens'},
583
- 'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Model input size (tokens)'}},
584
  'yaxis': {'anchor': 'x',
585
  'categoryarray': [llama-2-70b-chat, RedPajama-INCITE Chat
586
  (7B), Falcon Instruct (7B), Chronos
@@ -596,7 +612,7 @@ model_input_sizes,./html/plots/model_input_sizes.html,"Figure({
596
  'categoryorder': 'array',
597
  'domain': [0.0, 1.0],
598
  'title': {'text': 'Model'}}}
599
- })",Model input sizes in tokens,,"{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4-turbo"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""gpt-4-turbo"",""offsetgroup"":""gpt-4-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[128000],""xaxis"":""x"",""y"":[""gpt-4-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mistral (7B) Instruct v0.2 (Together AI)"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""Mistral (7B) Instruct v0.2 (Together AI)"",""offsetgroup"":""Mistral (7B) Instruct v0.2 (Together AI)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Mistral (7B) Instruct v0.2 (Together AI)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Snorkel Mistral PairRM DPO (7B)"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""Snorkel Mistral PairRM DPO (7B)"",""offsetgroup"":""Snorkel Mistral PairRM DPO (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Snorkel Mistral PairRM DPO (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Qwen 1.5 Chat (7B)"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""Qwen 1.5 Chat (7B)"",""offsetgroup"":""Qwen 1.5 Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Qwen 1.5 Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""gpt-4"",""offsetgroup"":""gpt-4"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32000],""xaxis"":""x"",""y"":[""gpt-4""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison (PaLM 2)"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""chat-bison (PaLM 2)"",""offsetgroup"":""chat-bison (PaLM 2)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[8196],""xaxis"":""x"",""y"":[""chat-bison (PaLM 2)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""OpenHermes-2.5-Mistral (7B)"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""OpenHermes-2.5-Mistral (7B)"",""offsetgroup"":""OpenHermes-2.5-Mistral (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[8192],""xaxis"":""x"",""y"":[""OpenHermes-2.5-Mistral (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-3.5-turbo"",""marker"":{""color"":""#B6E880"",""pattern"":{""shape"":""""}},""name"":""gpt-3.5-turbo"",""offsetgroup"":""gpt-3.5-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""gpt-3.5-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""WizardLM v1.2 (13B)"",""marker"":{""color"":""#FF97FF"",""pattern"":{""shape"":""""}},""name"":""WizardLM v1.2 (13B)"",""offsetgroup"":""WizardLM v1.2 (13B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""WizardLM v1.2 (13B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""LLaMA-2 Chat (7B)"",""marker"":{""color"":""#FECB52"",""pattern"":{""shape"":""""}},""name"":""LLaMA-2 Chat (7B)"",""offsetgroup"":""LLaMA-2 Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""LLaMA-2 Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Vicuna v1.5 (7B)"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""Vicuna v1.5 (7B)"",""offsetgroup"":""Vicuna v1.5 (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""Vicuna v1.5 (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""01-ai Yi Chat (34B)"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""01-ai Yi Chat (34B)"",""offsetgroup"":""01-ai Yi Chat (34B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""01-ai Yi Chat (34B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Upstage SOLAR Instruct v1 (11B)"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""Upstage SOLAR Instruct v1 (11B)"",""offsetgroup"":""Upstage SOLAR Instruct v1 (11B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""Upstage SOLAR Instruct v1 (11B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Chronos Hermes (13B)"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""Chronos Hermes (13B)"",""offsetgroup"":""Chronos Hermes (13B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""Chronos Hermes (13B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Falcon Instruct (7B)"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""Falcon Instruct (7B)"",""offsetgroup"":""Falcon Instruct (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""Falcon Instruct (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""RedPajama-INCITE Chat (7B)"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""RedPajama-INCITE Chat (7B)"",""offsetgroup"":""RedPajama-INCITE Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""RedPajama-INCITE Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel input size (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""llama-2-70b-chat"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""llama-2-70b-chat"",""offsetgroup"":""llama-2-70b-chat"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""llama-2-70b-chat""],""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 input size (tokens)""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Model""},""categoryorder"":""array"",""categoryarray"":[""llama-2-70b-chat"",""RedPajama-INCITE Chat (7B)"",""Falcon Instruct (7B)"",""Chronos Hermes (13B)"",""Upstage SOLAR Instruct v1 (11B)"",""01-ai Yi Chat (34B)"",""Vicuna v1.5 (7B)"",""LLaMA-2 Chat (7B)"",""WizardLM v1.2 (13B)"",""gpt-3.5-turbo"",""OpenHermes-2.5-Mistral (7B)"",""chat-bison (PaLM 2)"",""gpt-4"",""Qwen 1.5 Chat (7B)"",""Snorkel Mistral PairRM DPO (7B)"",""Mistral (7B) Instruct v0.2 (Together AI)"",""gpt-4-turbo""]},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Model input sizes in tokens""},""barmode"":""relative""}}","{""x"": ""model_input_sizes"", ""y"": ""model"", ""color"": ""model"", ""orientation"": ""h"", ""title"": ""Model input sizes in tokens"", ""labels"": {""model"": ""Model"", ""model_input_sizes"": ""Model input size (tokens)""}}",",model_input_sizes,model
600
  1,128000,gpt-4-turbo
601
  9,32768,Mistral (7B) Instruct v0.2 (Together AI)
602
  11,32768,Snorkel Mistral PairRM DPO (7B)
 
10
  'showlegend': True,
11
  'textposition': 'auto',
12
  'type': 'bar',
13
+ 'x': array([0.1446]),
14
  'xaxis': 'x',
15
  'y': array(['gpt-4'], dtype=object),
16
  'yaxis': 'y'},
 
24
  'showlegend': True,
25
  'textposition': 'auto',
26
  'type': 'bar',
27
+ 'x': array([0.056]),
28
  'xaxis': 'x',
29
  'y': array(['gpt-4-turbo'], dtype=object),
30
  'yaxis': 'y'},
 
38
  'showlegend': True,
39
  'textposition': 'auto',
40
  'type': 'bar',
41
+ 'x': array([0.00442]),
42
  'xaxis': 'x',
43
  'y': array(['gpt-3.5-turbo'], dtype=object),
44
  'yaxis': 'y'},
 
52
  'showlegend': True,
53
  'textposition': 'auto',
54
  'type': 'bar',
55
+ 'x': array([0.002808]),
56
  'xaxis': 'x',
57
  'y': array(['llama-2-70b-chat'], dtype=object),
58
  'yaxis': 'y'},
 
66
  'showlegend': True,
67
  'textposition': 'auto',
68
  'type': 'bar',
69
+ 'x': array([0.00207]),
70
  'xaxis': 'x',
71
  'y': array(['Mixtral-8x7B-Instruct-v0.1'], dtype=object),
72
  'yaxis': 'y'},
 
80
  'showlegend': True,
81
  'textposition': 'auto',
82
  'type': 'bar',
83
+ 'x': array([0.001195]),
84
  'xaxis': 'x',
85
  'y': array(['gemini-pro'], dtype=object),
86
  'yaxis': 'y'},
 
94
  'showlegend': True,
95
  'textposition': 'auto',
96
  'type': 'bar',
97
+ 'x': array([0.001075]),
98
  'xaxis': 'x',
99
  'y': array(['chat-bison (PaLM 2)'], dtype=object),
100
  'yaxis': 'y'},
 
108
  'showlegend': True,
109
  'textposition': 'auto',
110
  'type': 'bar',
111
+ 'x': array([0.001025]),
112
  'xaxis': 'x',
113
  'y': array(['chat-bison-32k (PaLM 2 32K)'], dtype=object),
114
+ 'yaxis': 'y'},
115
+ {'alignmentgroup': 'True',
116
+ 'hovertemplate': 'Model=%{y}<br>Execution cost ($)=%{x}<extra></extra>',
117
+ 'legendgroup': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
118
+ 'marker': {'color': '#FF97FF', 'pattern': {'shape': ''}},
119
+ 'name': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
120
+ 'offsetgroup': 'TinyLlama/TinyLlama-1.1B-Chat-v1.0',
121
+ 'orientation': 'h',
122
+ 'showlegend': True,
123
+ 'textposition': 'auto',
124
+ 'type': 'bar',
125
+ 'x': array([0.00062769]),
126
+ 'xaxis': 'x',
127
+ 'y': array(['TinyLlama/TinyLlama-1.1B-Chat-v1.0'], dtype=object),
128
  'yaxis': 'y'}],
129
  'layout': {'barmode': 'relative',
130
  'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
 
132
  'title': {'text': 'Costs of execution of 20 test queries per model'},
133
  'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Execution cost ($)'}},
134
  'yaxis': {'anchor': 'x',
135
+ 'categoryarray': [TinyLlama/TinyLlama-1.1B-Chat-v1.0,
136
+ chat-bison-32k (PaLM 2 32K), chat-bison
137
  (PaLM 2), gemini-pro,
138
  Mixtral-8x7B-Instruct-v0.1,
139
  llama-2-70b-chat, gpt-3.5-turbo,
 
141
  'categoryorder': 'array',
142
  'domain': [0.0, 1.0],
143
  'title': {'text': 'Model'}}}
144
+ })",Costs of execution of 20 test queries per model,,"{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""gpt-4"",""offsetgroup"":""gpt-4"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.1446],""xaxis"":""x"",""y"":[""gpt-4""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4-turbo"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""gpt-4-turbo"",""offsetgroup"":""gpt-4-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.056],""xaxis"":""x"",""y"":[""gpt-4-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-3.5-turbo"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""gpt-3.5-turbo"",""offsetgroup"":""gpt-3.5-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.00442],""xaxis"":""x"",""y"":[""gpt-3.5-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""llama-2-70b-chat"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""llama-2-70b-chat"",""offsetgroup"":""llama-2-70b-chat"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.002808],""xaxis"":""x"",""y"":[""llama-2-70b-chat""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mixtral-8x7B-Instruct-v0.1"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""Mixtral-8x7B-Instruct-v0.1"",""offsetgroup"":""Mixtral-8x7B-Instruct-v0.1"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0020700000000000002],""xaxis"":""x"",""y"":[""Mixtral-8x7B-Instruct-v0.1""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gemini-pro"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""gemini-pro"",""offsetgroup"":""gemini-pro"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0011949999999999999],""xaxis"":""x"",""y"":[""gemini-pro""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison (PaLM 2)"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""chat-bison (PaLM 2)"",""offsetgroup"":""chat-bison (PaLM 2)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.001075],""xaxis"":""x"",""y"":[""chat-bison (PaLM 2)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison-32k (PaLM 2 32K)"",""marker"":{""color"":""#B6E880"",""pattern"":{""shape"":""""}},""name"":""chat-bison-32k (PaLM 2 32K)"",""offsetgroup"":""chat-bison-32k (PaLM 2 32K)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0010249999999999999],""xaxis"":""x"",""y"":[""chat-bison-32k (PaLM 2 32K)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eExecution cost ($)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""marker"":{""color"":""#FF97FF"",""pattern"":{""shape"":""""}},""name"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""offsetgroup"":""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[0.0006276866594950359],""xaxis"":""x"",""y"":[""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0""],""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"":""Execution cost ($)""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Model""},""categoryorder"":""array"",""categoryarray"":[""TinyLlama\u002fTinyLlama-1.1B-Chat-v1.0"",""chat-bison-32k (PaLM 2 32K)"",""chat-bison (PaLM 2)"",""gemini-pro"",""Mixtral-8x7B-Instruct-v0.1"",""llama-2-70b-chat"",""gpt-3.5-turbo"",""gpt-4-turbo"",""gpt-4""]},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Costs of execution of 20 test queries per model""},""barmode"":""relative""}}","{""y"": ""model"", ""x"": ""model_query_costs"", ""color"": ""model"", ""orientation"": ""h"", ""title"": ""Costs of execution of 20 test queries per model"", ""labels"": {""model"": ""Model"", ""model_query_costs"": ""Execution cost ($)""}}",",model_query_costs,model
145
+ 2,0.1446,gpt-4
146
+ 1,0.056,gpt-4-turbo
147
+ 0,0.00442,gpt-3.5-turbo
148
+ 3,0.002808,llama-2-70b-chat
149
+ 4,0.0020700000000000002,Mixtral-8x7B-Instruct-v0.1
150
+ 8,0.0011949999999999999,gemini-pro
151
+ 9,0.001075,chat-bison (PaLM 2)
152
+ 10,0.0010249999999999999,chat-bison-32k (PaLM 2 32K)
153
+ 7,0.0006276866594950359,TinyLlama/TinyLlama-1.1B-Chat-v1.0
154
  "
155
  model_sizes,./html/plots/model_sizes.html,"Figure({
156
  'data': [{'alignmentgroup': 'True',
 
353
  9,7.0,RedPajama-INCITE Chat (7B)
354
  4,1.1,TinyLlama/TinyLlama-1.1B-Chat-v1.0
355
  "
356
+ model_context_lengths,./html/plots/model_context_lengths.html,"Figure({
357
  'data': [{'alignmentgroup': 'True',
358
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
359
  'legendgroup': 'gpt-4-turbo',
360
  'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
361
  'name': 'gpt-4-turbo',
 
369
  'y': array(['gpt-4-turbo'], dtype=object),
370
  'yaxis': 'y'},
371
  {'alignmentgroup': 'True',
372
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
373
  'legendgroup': 'Mistral (7B) Instruct v0.2 (Together AI)',
374
  'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
375
  'name': 'Mistral (7B) Instruct v0.2 (Together AI)',
 
383
  'y': array(['Mistral (7B) Instruct v0.2 (Together AI)'], dtype=object),
384
  'yaxis': 'y'},
385
  {'alignmentgroup': 'True',
386
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
387
  'legendgroup': 'Snorkel Mistral PairRM DPO (7B)',
388
  'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
389
  'name': 'Snorkel Mistral PairRM DPO (7B)',
 
397
  'y': array(['Snorkel Mistral PairRM DPO (7B)'], dtype=object),
398
  'yaxis': 'y'},
399
  {'alignmentgroup': 'True',
400
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
401
  'legendgroup': 'Qwen 1.5 Chat (7B)',
402
  'marker': {'color': '#ab63fa', 'pattern': {'shape': ''}},
403
  'name': 'Qwen 1.5 Chat (7B)',
 
411
  'y': array(['Qwen 1.5 Chat (7B)'], dtype=object),
412
  'yaxis': 'y'},
413
  {'alignmentgroup': 'True',
414
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
415
  'legendgroup': 'gpt-4',
416
  'marker': {'color': '#FFA15A', 'pattern': {'shape': ''}},
417
  'name': 'gpt-4',
 
425
  'y': array(['gpt-4'], dtype=object),
426
  'yaxis': 'y'},
427
  {'alignmentgroup': 'True',
428
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
429
  'legendgroup': 'chat-bison (PaLM 2)',
430
  'marker': {'color': '#19d3f3', 'pattern': {'shape': ''}},
431
  'name': 'chat-bison (PaLM 2)',
 
439
  'y': array(['chat-bison (PaLM 2)'], dtype=object),
440
  'yaxis': 'y'},
441
  {'alignmentgroup': 'True',
442
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
443
  'legendgroup': 'OpenHermes-2.5-Mistral (7B)',
444
  'marker': {'color': '#FF6692', 'pattern': {'shape': ''}},
445
  'name': 'OpenHermes-2.5-Mistral (7B)',
 
453
  'y': array(['OpenHermes-2.5-Mistral (7B)'], dtype=object),
454
  'yaxis': 'y'},
455
  {'alignmentgroup': 'True',
456
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
457
  'legendgroup': 'gpt-3.5-turbo',
458
  'marker': {'color': '#B6E880', 'pattern': {'shape': ''}},
459
  'name': 'gpt-3.5-turbo',
 
467
  'y': array(['gpt-3.5-turbo'], dtype=object),
468
  'yaxis': 'y'},
469
  {'alignmentgroup': 'True',
470
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
471
  'legendgroup': 'WizardLM v1.2 (13B)',
472
  'marker': {'color': '#FF97FF', 'pattern': {'shape': ''}},
473
  'name': 'WizardLM v1.2 (13B)',
 
481
  'y': array(['WizardLM v1.2 (13B)'], dtype=object),
482
  'yaxis': 'y'},
483
  {'alignmentgroup': 'True',
484
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
485
  'legendgroup': 'LLaMA-2 Chat (7B)',
486
  'marker': {'color': '#FECB52', 'pattern': {'shape': ''}},
487
  'name': 'LLaMA-2 Chat (7B)',
 
495
  'y': array(['LLaMA-2 Chat (7B)'], dtype=object),
496
  'yaxis': 'y'},
497
  {'alignmentgroup': 'True',
498
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
499
  'legendgroup': 'Vicuna v1.5 (7B)',
500
  'marker': {'color': '#636efa', 'pattern': {'shape': ''}},
501
  'name': 'Vicuna v1.5 (7B)',
 
509
  'y': array(['Vicuna v1.5 (7B)'], dtype=object),
510
  'yaxis': 'y'},
511
  {'alignmentgroup': 'True',
512
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
513
  'legendgroup': '01-ai Yi Chat (34B)',
514
  'marker': {'color': '#EF553B', 'pattern': {'shape': ''}},
515
  'name': '01-ai Yi Chat (34B)',
 
523
  'y': array(['01-ai Yi Chat (34B)'], dtype=object),
524
  'yaxis': 'y'},
525
  {'alignmentgroup': 'True',
526
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
527
  'legendgroup': 'Upstage SOLAR Instruct v1 (11B)',
528
  'marker': {'color': '#00cc96', 'pattern': {'shape': ''}},
529
  'name': 'Upstage SOLAR Instruct v1 (11B)',
 
537
  'y': array(['Upstage SOLAR Instruct v1 (11B)'], dtype=object),
538
  'yaxis': 'y'},
539
  {'alignmentgroup': 'True',
540
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
541
  'legendgroup': 'Chronos Hermes (13B)',
542
  'marker': {'color': '#ab63fa', 'pattern': {'shape': ''}},
543
  'name': 'Chronos Hermes (13B)',
 
551
  'y': array(['Chronos Hermes (13B)'], dtype=object),
552
  'yaxis': 'y'},
553
  {'alignmentgroup': 'True',
554
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
555
  'legendgroup': 'Falcon Instruct (7B)',
556
  'marker': {'color': '#FFA15A', 'pattern': {'shape': ''}},
557
  'name': 'Falcon Instruct (7B)',
 
565
  'y': array(['Falcon Instruct (7B)'], dtype=object),
566
  'yaxis': 'y'},
567
  {'alignmentgroup': 'True',
568
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
569
  'legendgroup': 'RedPajama-INCITE Chat (7B)',
570
  'marker': {'color': '#19d3f3', 'pattern': {'shape': ''}},
571
  'name': 'RedPajama-INCITE Chat (7B)',
 
579
  'y': array(['RedPajama-INCITE Chat (7B)'], dtype=object),
580
  'yaxis': 'y'},
581
  {'alignmentgroup': 'True',
582
+ 'hovertemplate': 'Model=%{y}<br>Model context length (tokens)=%{x}<extra></extra>',
583
  'legendgroup': 'llama-2-70b-chat',
584
  'marker': {'color': '#FF6692', 'pattern': {'shape': ''}},
585
  'name': 'llama-2-70b-chat',
 
595
  'layout': {'barmode': 'relative',
596
  'legend': {'title': {'text': 'Model'}, 'tracegroupgap': 0},
597
  'template': '...',
598
+ 'title': {'text': 'Model context lengths in tokens'},
599
+ 'xaxis': {'anchor': 'y', 'domain': [0.0, 1.0], 'title': {'text': 'Model context length (tokens)'}},
600
  'yaxis': {'anchor': 'x',
601
  'categoryarray': [llama-2-70b-chat, RedPajama-INCITE Chat
602
  (7B), Falcon Instruct (7B), Chronos
 
612
  'categoryorder': 'array',
613
  'domain': [0.0, 1.0],
614
  'title': {'text': 'Model'}}}
615
+ })",Model context lengths in tokens,,"{""data"":[{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4-turbo"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""gpt-4-turbo"",""offsetgroup"":""gpt-4-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[128000],""xaxis"":""x"",""y"":[""gpt-4-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Mistral (7B) Instruct v0.2 (Together AI)"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""Mistral (7B) Instruct v0.2 (Together AI)"",""offsetgroup"":""Mistral (7B) Instruct v0.2 (Together AI)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Mistral (7B) Instruct v0.2 (Together AI)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Snorkel Mistral PairRM DPO (7B)"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""Snorkel Mistral PairRM DPO (7B)"",""offsetgroup"":""Snorkel Mistral PairRM DPO (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Snorkel Mistral PairRM DPO (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Qwen 1.5 Chat (7B)"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""Qwen 1.5 Chat (7B)"",""offsetgroup"":""Qwen 1.5 Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32768],""xaxis"":""x"",""y"":[""Qwen 1.5 Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-4"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""gpt-4"",""offsetgroup"":""gpt-4"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[32000],""xaxis"":""x"",""y"":[""gpt-4""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""chat-bison (PaLM 2)"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""chat-bison (PaLM 2)"",""offsetgroup"":""chat-bison (PaLM 2)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[8196],""xaxis"":""x"",""y"":[""chat-bison (PaLM 2)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""OpenHermes-2.5-Mistral (7B)"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""OpenHermes-2.5-Mistral (7B)"",""offsetgroup"":""OpenHermes-2.5-Mistral (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[8192],""xaxis"":""x"",""y"":[""OpenHermes-2.5-Mistral (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""gpt-3.5-turbo"",""marker"":{""color"":""#B6E880"",""pattern"":{""shape"":""""}},""name"":""gpt-3.5-turbo"",""offsetgroup"":""gpt-3.5-turbo"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""gpt-3.5-turbo""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""WizardLM v1.2 (13B)"",""marker"":{""color"":""#FF97FF"",""pattern"":{""shape"":""""}},""name"":""WizardLM v1.2 (13B)"",""offsetgroup"":""WizardLM v1.2 (13B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""WizardLM v1.2 (13B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""LLaMA-2 Chat (7B)"",""marker"":{""color"":""#FECB52"",""pattern"":{""shape"":""""}},""name"":""LLaMA-2 Chat (7B)"",""offsetgroup"":""LLaMA-2 Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""LLaMA-2 Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Vicuna v1.5 (7B)"",""marker"":{""color"":""#636efa"",""pattern"":{""shape"":""""}},""name"":""Vicuna v1.5 (7B)"",""offsetgroup"":""Vicuna v1.5 (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""Vicuna v1.5 (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""01-ai Yi Chat (34B)"",""marker"":{""color"":""#EF553B"",""pattern"":{""shape"":""""}},""name"":""01-ai Yi Chat (34B)"",""offsetgroup"":""01-ai Yi Chat (34B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""01-ai Yi Chat (34B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Upstage SOLAR Instruct v1 (11B)"",""marker"":{""color"":""#00cc96"",""pattern"":{""shape"":""""}},""name"":""Upstage SOLAR Instruct v1 (11B)"",""offsetgroup"":""Upstage SOLAR Instruct v1 (11B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[4096],""xaxis"":""x"",""y"":[""Upstage SOLAR Instruct v1 (11B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Chronos Hermes (13B)"",""marker"":{""color"":""#ab63fa"",""pattern"":{""shape"":""""}},""name"":""Chronos Hermes (13B)"",""offsetgroup"":""Chronos Hermes (13B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""Chronos Hermes (13B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""Falcon Instruct (7B)"",""marker"":{""color"":""#FFA15A"",""pattern"":{""shape"":""""}},""name"":""Falcon Instruct (7B)"",""offsetgroup"":""Falcon Instruct (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""Falcon Instruct (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""RedPajama-INCITE Chat (7B)"",""marker"":{""color"":""#19d3f3"",""pattern"":{""shape"":""""}},""name"":""RedPajama-INCITE Chat (7B)"",""offsetgroup"":""RedPajama-INCITE Chat (7B)"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""RedPajama-INCITE Chat (7B)""],""yaxis"":""y"",""type"":""bar""},{""alignmentgroup"":""True"",""hovertemplate"":""Model=%{y}\u003cbr\u003eModel context length (tokens)=%{x}\u003cextra\u003e\u003c\u002fextra\u003e"",""legendgroup"":""llama-2-70b-chat"",""marker"":{""color"":""#FF6692"",""pattern"":{""shape"":""""}},""name"":""llama-2-70b-chat"",""offsetgroup"":""llama-2-70b-chat"",""orientation"":""h"",""showlegend"":true,""textposition"":""auto"",""x"":[2048],""xaxis"":""x"",""y"":[""llama-2-70b-chat""],""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 context length (tokens)""}},""yaxis"":{""anchor"":""x"",""domain"":[0.0,1.0],""title"":{""text"":""Model""},""categoryorder"":""array"",""categoryarray"":[""llama-2-70b-chat"",""RedPajama-INCITE Chat (7B)"",""Falcon Instruct (7B)"",""Chronos Hermes (13B)"",""Upstage SOLAR Instruct v1 (11B)"",""01-ai Yi Chat (34B)"",""Vicuna v1.5 (7B)"",""LLaMA-2 Chat (7B)"",""WizardLM v1.2 (13B)"",""gpt-3.5-turbo"",""OpenHermes-2.5-Mistral (7B)"",""chat-bison (PaLM 2)"",""gpt-4"",""Qwen 1.5 Chat (7B)"",""Snorkel Mistral PairRM DPO (7B)"",""Mistral (7B) Instruct v0.2 (Together AI)"",""gpt-4-turbo""]},""legend"":{""title"":{""text"":""Model""},""tracegroupgap"":0},""title"":{""text"":""Model context lengths in tokens""},""barmode"":""relative""}}","{""x"": ""model_context_lengths"", ""y"": ""model"", ""color"": ""model"", ""orientation"": ""h"", ""title"": ""Model context lengths in tokens"", ""labels"": {""model"": ""Model"", ""model_context_lengths"": ""Model context length (tokens)""}}",",model_context_lengths,model
616
  1,128000,gpt-4-turbo
617
  9,32768,Mistral (7B) Instruct v0.2 (Together AI)
618
  11,32768,Snorkel Mistral PairRM DPO (7B)
data/time_of_day_plots.csv CHANGED
The diff for this file is too large to render. See raw diff
 
pipeline/models.py CHANGED
@@ -15,7 +15,7 @@ class Model(object):
15
  cost_per_million_tokens: int = None
16
  cost_per_million_input_tokens: int = None
17
  cost_per_million_output_tokens: int = None
18
- input_size: int = None
19
  selected: bool = False
20
 
21
  def __post_init__(self):
@@ -48,7 +48,7 @@ MODELS = [
48
  cost_per_million_input_tokens=1,
49
  cost_per_million_output_tokens=2,
50
  # https://learn.microsoft.com/en-us/answers/questions/1356487/what-is-the-exact-maximum-input-tokens-of-azure-gp
51
- input_size=4096,
52
  ),
53
  Model(
54
  "gpt-4-turbo",
@@ -59,7 +59,7 @@ MODELS = [
59
  cost_per_million_input_tokens=10,
60
  cost_per_million_output_tokens=30,
61
  # https://writesonic.com/blog/gpt-4-turbo-vs-gpt-4
62
- input_size=128_000,
63
  ),
64
  Model(
65
  "gpt-4",
@@ -69,7 +69,7 @@ MODELS = [
69
  supports_functions=True,
70
  cost_per_million_input_tokens=30,
71
  cost_per_million_output_tokens=60,
72
- input_size=32_000,
73
  ),
74
  # source: https://www.together.ai/pricing
75
  Model(
@@ -80,7 +80,7 @@ MODELS = [
80
  cost_per_million_tokens=0.9,
81
  size_billion_parameters=70,
82
  # https://github.com/facebookresearch/llama/issues/148
83
- input_size=2048,
84
  ),
85
  Model(
86
  "Mixtral-8x7B-Instruct-v0.1",
@@ -106,6 +106,7 @@ MODELS = [
106
  "Hugging Face Inference Endpoint",
107
  hourly_cost=1.30,
108
  size_billion_parameters=7,
 
109
  ),
110
  Model(
111
  "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
@@ -134,7 +135,7 @@ MODELS = [
134
  cost_per_million_input_tokens=0.25,
135
  cost_per_million_output_tokens=0.5,
136
  # https://ai.google.dev/models/palm
137
- input_size=8196,
138
  ),
139
  Model(
140
  "chat-bison-32k (PaLM 2 32K)",
@@ -151,7 +152,7 @@ MODELS = [
151
  "together_ai/zero-one-ai/Yi-34B-Chat",
152
  None,
153
  "Together AI",
154
- input_size=4096,
155
  # selected=True,
156
  ),
157
  Model(
@@ -159,7 +160,7 @@ MODELS = [
159
  "together_ai/Austism/chronos-hermes-13b",
160
  None,
161
  "Together AI",
162
- input_size=2048,
163
  # selected=True,
164
  ),
165
  Model(
@@ -167,50 +168,58 @@ MODELS = [
167
  "together_ai/deepseek-ai/deepseek-coder-33b-instruct",
168
  None,
169
  "Together AI",
170
- input_size=16384,
171
  ),
172
  Model(
173
  "Platypus2 Instruct (70B)",
174
  "together_ai/garage-bAInd/Platypus2-70B-instruct",
175
  None,
176
  "Together AI",
177
- input_size=4096,
178
  ),
179
  Model(
180
  "MythoMax-L2 (13B)",
181
  "together_ai/Gryphe/MythoMax-L2-13b",
182
  None,
183
  "Together AI",
184
- input_size=4096,
185
  ),
186
  Model(
187
  "Vicuna v1.5 (13B)",
188
  "together_ai/lmsys/vicuna-13b-v1.5",
189
  None,
190
  "Together AI",
191
- input_size=4096,
 
 
 
 
 
 
 
 
 
192
  ),
193
- Model("Vicuna v1.5 (7B)", "together_ai/lmsys/vicuna-7b-v1.5", None, "Together AI", input_size=4096, size_billion_parameters=7, selected=True),
194
  Model(
195
  "Code Llama Instruct (13B)",
196
  "together_ai/codellama/CodeLlama-13b-Instruct-hf",
197
  None,
198
  "Together AI",
199
- input_size=16384,
200
  ),
201
  Model(
202
  "Code Llama Instruct (34B)",
203
  "together_ai/codellama/CodeLlama-34b-Instruct-hf",
204
  None,
205
  "Together AI",
206
- input_size=16384,
207
  ),
208
  Model(
209
  "Code Llama Instruct (70B)",
210
  "together_ai/codellama/CodeLlama-70b-Instruct-hf",
211
  None,
212
  "Together AI",
213
- input_size=4096,
214
  ),
215
  Model(
216
  "Code Llama Instruct (7B)",
@@ -218,7 +227,7 @@ MODELS = [
218
  None,
219
  "Together AI",
220
  size_billion_parameters=7,
221
- input_size=16384,
222
  ),
223
  Model(
224
  "LLaMA-2 Chat (13B)",
@@ -226,7 +235,7 @@ MODELS = [
226
  None,
227
  "Together AI",
228
  size_billion_parameters=13,
229
- input_size=4096,
230
  ),
231
  Model(
232
  "LLaMA-2 Chat (70B)",
@@ -234,7 +243,7 @@ MODELS = [
234
  None,
235
  "Together AI",
236
  size_billion_parameters=70,
237
- input_size=4096,
238
  ),
239
  Model(
240
  "LLaMA-2 Chat (7B)",
@@ -242,7 +251,7 @@ MODELS = [
242
  None,
243
  "Together AI",
244
  size_billion_parameters=7,
245
- input_size=4096,
246
  # selected=True,
247
  ),
248
  Model(
@@ -251,7 +260,7 @@ MODELS = [
251
  None,
252
  "Together AI",
253
  size_billion_parameters=7,
254
- input_size=4096,
255
  ),
256
  Model(
257
  "Mistral (7B) Instruct v0.2 (Together AI)",
@@ -259,7 +268,7 @@ MODELS = [
259
  None,
260
  "Together AI",
261
  size_billion_parameters=7,
262
- input_size=32768,
263
  selected=True,
264
  ),
265
  Model(
@@ -267,8 +276,8 @@ MODELS = [
267
  "together_ai/mistralai/Mixtral-8x7B-Instruct-v0.1",
268
  None,
269
  "Together AI",
270
- size_billion_parameters=8*7,
271
- input_size=32768,
272
  ),
273
  Model(
274
  "Nous Capybara v1.9 (7B)",
@@ -276,23 +285,23 @@ MODELS = [
276
  None,
277
  "Together AI",
278
  size_billion_parameters=7,
279
- input_size=8192,
280
  ),
281
  Model(
282
  "Nous Hermes 2 - Mixtral 8x7B-DPO (46.7B)",
283
  "together_ai/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
284
  None,
285
  "Together AI",
286
- size_billion_parameters=8*7,
287
- input_size=32768,
288
  ),
289
  Model(
290
  "Nous Hermes 2 - Mixtral 8x7B-SFT (46.7B)",
291
  "together_ai/NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT",
292
  None,
293
  "Together AI",
294
- size_billion_parameters=8*7,
295
- input_size=32768,
296
  ),
297
  Model(
298
  "Nous Hermes LLaMA-2 (7B)",
@@ -300,7 +309,7 @@ MODELS = [
300
  None,
301
  "Together AI",
302
  size_billion_parameters=7,
303
- input_size=4096,
304
  ),
305
  Model(
306
  "Nous Hermes Llama-2 (13B)",
@@ -308,7 +317,7 @@ MODELS = [
308
  None,
309
  "Together AI",
310
  size_billion_parameters=13,
311
- input_size=4096,
312
  ),
313
  Model(
314
  "Nous Hermes-2 Yi (34B)",
@@ -316,7 +325,7 @@ MODELS = [
316
  None,
317
  "Together AI",
318
  size_billion_parameters=34,
319
- input_size=4096,
320
  ),
321
  Model(
322
  "OpenChat 3.5 (7B)",
@@ -324,7 +333,7 @@ MODELS = [
324
  None,
325
  "Together AI",
326
  size_billion_parameters=7,
327
- input_size=8192,
328
  ),
329
  Model(
330
  "OpenOrca Mistral (7B) 8K",
@@ -332,7 +341,7 @@ MODELS = [
332
  None,
333
  "Together AI",
334
  size_billion_parameters=7,
335
- input_size=8192,
336
  ),
337
  Model(
338
  "Qwen-Chat (7B)",
@@ -340,7 +349,7 @@ MODELS = [
340
  None,
341
  "Together AI",
342
  size_billion_parameters=7,
343
- input_size=8192,
344
  ),
345
  Model(
346
  "Qwen 1.5 Chat (0.5B)",
@@ -348,14 +357,14 @@ MODELS = [
348
  None,
349
  "Together AI",
350
  size_billion_parameters=0.5,
351
- input_size=32768,
352
  ),
353
  Model(
354
  "Qwen 1.5 Chat (1.8B)",
355
  "together_ai/Qwen/Qwen1.5-1.8B-Chat",
356
  None,
357
  "Together AI",
358
- input_size=32768,
359
  size_billion_parameters=1.8,
360
  ),
361
  Model(
@@ -364,10 +373,15 @@ MODELS = [
364
  None,
365
  "Together AI",
366
  size_billion_parameters=4,
367
- input_size=32768,
368
  ),
369
  Model(
370
- "Qwen 1.5 Chat (7B)", "together_ai/Qwen/Qwen1.5-7B-Chat", None, "Together AI", input_size=32768, size_billion_parameters=7
 
 
 
 
 
371
  # selected=True
372
  ),
373
  Model(
@@ -376,21 +390,21 @@ MODELS = [
376
  None,
377
  "Together AI",
378
  size_billion_parameters=14,
379
- input_size=32768,
380
  ),
381
  Model(
382
  "Qwen 1.5 Chat (72B)",
383
  "together_ai/Qwen/Qwen1.5-72B-Chat",
384
  None,
385
  "Together AI",
386
- input_size=4096,
387
  ),
388
  Model(
389
  "Snorkel Mistral PairRM DPO (7B)",
390
  "together_ai/snorkelai/Snorkel-Mistral-PairRM-DPO",
391
  None,
392
  "Together AI",
393
- input_size=32768,
394
  # selected=True,
395
  ),
396
  Model(
@@ -398,21 +412,21 @@ MODELS = [
398
  "together_ai/togethercomputer/alpaca-7b",
399
  None,
400
  "Together AI",
401
- input_size=2048,
402
  ),
403
  Model(
404
  "OpenHermes-2-Mistral (7B)",
405
  "teknium/OpenHermes-2-Mistral-7B",
406
  None,
407
  "Together AI",
408
- input_size=8192,
409
  ),
410
  Model(
411
  "OpenHermes-2.5-Mistral (7B)",
412
  "together_ai/teknium/OpenHermes-2p5-Mistral-7B",
413
  None,
414
  "Together AI",
415
- input_size=8192,
416
  # selected=True,
417
  ),
418
  Model(
@@ -420,14 +434,14 @@ MODELS = [
420
  "together_ai/togethercomputer/falcon-40b-instruct",
421
  None,
422
  "Together AI",
423
- input_size=2048,
424
  ),
425
  Model(
426
  "Falcon Instruct (7B)",
427
  "together_ai/togethercomputer/falcon-7b-instruct",
428
  None,
429
  "Together AI",
430
- input_size=2048,
431
  # selected=True,
432
  ),
433
  Model(
@@ -435,7 +449,7 @@ MODELS = [
435
  "together_ai/togethercomputer/Llama-2-7B-32K-Instruct",
436
  None,
437
  "Together AI",
438
- input_size=32768,
439
  ),
440
  Model(
441
  "RedPajama-INCITE Chat (3B)",
@@ -443,14 +457,14 @@ MODELS = [
443
  None,
444
  "Together AI",
445
  size_billion_parameters=3,
446
- input_size=2048,
447
  ),
448
  Model(
449
  "RedPajama-INCITE Chat (7B)",
450
  "together_ai/togethercomputer/RedPajama-INCITE-7B-Chat",
451
  None,
452
  "Together AI",
453
- input_size=2048,
454
  size_billion_parameters=7,
455
  # selected=True,
456
  ),
@@ -459,7 +473,7 @@ MODELS = [
459
  "together_ai/togethercomputer/StripedHyena-Nous-7B",
460
  None,
461
  "Together AI",
462
- input_size=32768,
463
  size_billion_parameters=7,
464
  ),
465
  Model(
@@ -467,7 +481,7 @@ MODELS = [
467
  "together_ai/Undi95/ReMM-SLERP-L2-13B",
468
  None,
469
  "Together AI",
470
- input_size=4096,
471
  size_billion_parameters=13,
472
  ),
473
  Model(
@@ -475,15 +489,15 @@ MODELS = [
475
  "together_ai/Undi95/Toppy-M-7B",
476
  None,
477
  "Together AI",
478
- input_size=4096,
479
- size_billion_parameters=7
480
  ),
481
  Model(
482
  "WizardLM v1.2 (13B)",
483
  "together_ai/WizardLM/WizardLM-13B-V1.2",
484
  None,
485
  "Together AI",
486
- input_size=4096,
487
  size_billion_parameters=13,
488
  # selected=True,
489
  ),
@@ -492,7 +506,7 @@ MODELS = [
492
  "together_ai/upstage/SOLAR-10.7B-Instruct-v1.0",
493
  None,
494
  "Together AI",
495
- input_size=4096,
496
  size_billion_parameters=11,
497
  # selected=True,
498
  ),
 
15
  cost_per_million_tokens: int = None
16
  cost_per_million_input_tokens: int = None
17
  cost_per_million_output_tokens: int = None
18
+ context_length: int = None
19
  selected: bool = False
20
 
21
  def __post_init__(self):
 
48
  cost_per_million_input_tokens=1,
49
  cost_per_million_output_tokens=2,
50
  # https://learn.microsoft.com/en-us/answers/questions/1356487/what-is-the-exact-maximum-input-tokens-of-azure-gp
51
+ context_length=4096,
52
  ),
53
  Model(
54
  "gpt-4-turbo",
 
59
  cost_per_million_input_tokens=10,
60
  cost_per_million_output_tokens=30,
61
  # https://writesonic.com/blog/gpt-4-turbo-vs-gpt-4
62
+ context_length=128_000,
63
  ),
64
  Model(
65
  "gpt-4",
 
69
  supports_functions=True,
70
  cost_per_million_input_tokens=30,
71
  cost_per_million_output_tokens=60,
72
+ context_length=32_000,
73
  ),
74
  # source: https://www.together.ai/pricing
75
  Model(
 
80
  cost_per_million_tokens=0.9,
81
  size_billion_parameters=70,
82
  # https://github.com/facebookresearch/llama/issues/148
83
+ context_length=2048,
84
  ),
85
  Model(
86
  "Mixtral-8x7B-Instruct-v0.1",
 
106
  "Hugging Face Inference Endpoint",
107
  hourly_cost=1.30,
108
  size_billion_parameters=7,
109
+ selected=True,
110
  ),
111
  Model(
112
  "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
 
135
  cost_per_million_input_tokens=0.25,
136
  cost_per_million_output_tokens=0.5,
137
  # https://ai.google.dev/models/palm
138
+ context_length=8196,
139
  ),
140
  Model(
141
  "chat-bison-32k (PaLM 2 32K)",
 
152
  "together_ai/zero-one-ai/Yi-34B-Chat",
153
  None,
154
  "Together AI",
155
+ context_length=4096,
156
  # selected=True,
157
  ),
158
  Model(
 
160
  "together_ai/Austism/chronos-hermes-13b",
161
  None,
162
  "Together AI",
163
+ context_length=2048,
164
  # selected=True,
165
  ),
166
  Model(
 
168
  "together_ai/deepseek-ai/deepseek-coder-33b-instruct",
169
  None,
170
  "Together AI",
171
+ context_length=16384,
172
  ),
173
  Model(
174
  "Platypus2 Instruct (70B)",
175
  "together_ai/garage-bAInd/Platypus2-70B-instruct",
176
  None,
177
  "Together AI",
178
+ context_length=4096,
179
  ),
180
  Model(
181
  "MythoMax-L2 (13B)",
182
  "together_ai/Gryphe/MythoMax-L2-13b",
183
  None,
184
  "Together AI",
185
+ context_length=4096,
186
  ),
187
  Model(
188
  "Vicuna v1.5 (13B)",
189
  "together_ai/lmsys/vicuna-13b-v1.5",
190
  None,
191
  "Together AI",
192
+ context_length=4096,
193
+ ),
194
+ Model(
195
+ "Vicuna v1.5 (7B)",
196
+ "together_ai/lmsys/vicuna-7b-v1.5",
197
+ None,
198
+ "Together AI",
199
+ context_length=4096,
200
+ size_billion_parameters=7,
201
+ # selected=True
202
  ),
 
203
  Model(
204
  "Code Llama Instruct (13B)",
205
  "together_ai/codellama/CodeLlama-13b-Instruct-hf",
206
  None,
207
  "Together AI",
208
+ context_length=16384,
209
  ),
210
  Model(
211
  "Code Llama Instruct (34B)",
212
  "together_ai/codellama/CodeLlama-34b-Instruct-hf",
213
  None,
214
  "Together AI",
215
+ context_length=16384,
216
  ),
217
  Model(
218
  "Code Llama Instruct (70B)",
219
  "together_ai/codellama/CodeLlama-70b-Instruct-hf",
220
  None,
221
  "Together AI",
222
+ context_length=4096,
223
  ),
224
  Model(
225
  "Code Llama Instruct (7B)",
 
227
  None,
228
  "Together AI",
229
  size_billion_parameters=7,
230
+ context_length=16384,
231
  ),
232
  Model(
233
  "LLaMA-2 Chat (13B)",
 
235
  None,
236
  "Together AI",
237
  size_billion_parameters=13,
238
+ context_length=4096,
239
  ),
240
  Model(
241
  "LLaMA-2 Chat (70B)",
 
243
  None,
244
  "Together AI",
245
  size_billion_parameters=70,
246
+ context_length=4096,
247
  ),
248
  Model(
249
  "LLaMA-2 Chat (7B)",
 
251
  None,
252
  "Together AI",
253
  size_billion_parameters=7,
254
+ context_length=4096,
255
  # selected=True,
256
  ),
257
  Model(
 
260
  None,
261
  "Together AI",
262
  size_billion_parameters=7,
263
+ context_length=4096,
264
  ),
265
  Model(
266
  "Mistral (7B) Instruct v0.2 (Together AI)",
 
268
  None,
269
  "Together AI",
270
  size_billion_parameters=7,
271
+ context_length=32768,
272
  selected=True,
273
  ),
274
  Model(
 
276
  "together_ai/mistralai/Mixtral-8x7B-Instruct-v0.1",
277
  None,
278
  "Together AI",
279
+ size_billion_parameters=8 * 7,
280
+ context_length=32768,
281
  ),
282
  Model(
283
  "Nous Capybara v1.9 (7B)",
 
285
  None,
286
  "Together AI",
287
  size_billion_parameters=7,
288
+ context_length=8192,
289
  ),
290
  Model(
291
  "Nous Hermes 2 - Mixtral 8x7B-DPO (46.7B)",
292
  "together_ai/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
293
  None,
294
  "Together AI",
295
+ size_billion_parameters=8 * 7,
296
+ context_length=32768,
297
  ),
298
  Model(
299
  "Nous Hermes 2 - Mixtral 8x7B-SFT (46.7B)",
300
  "together_ai/NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFT",
301
  None,
302
  "Together AI",
303
+ size_billion_parameters=8 * 7,
304
+ context_length=32768,
305
  ),
306
  Model(
307
  "Nous Hermes LLaMA-2 (7B)",
 
309
  None,
310
  "Together AI",
311
  size_billion_parameters=7,
312
+ context_length=4096,
313
  ),
314
  Model(
315
  "Nous Hermes Llama-2 (13B)",
 
317
  None,
318
  "Together AI",
319
  size_billion_parameters=13,
320
+ context_length=4096,
321
  ),
322
  Model(
323
  "Nous Hermes-2 Yi (34B)",
 
325
  None,
326
  "Together AI",
327
  size_billion_parameters=34,
328
+ context_length=4096,
329
  ),
330
  Model(
331
  "OpenChat 3.5 (7B)",
 
333
  None,
334
  "Together AI",
335
  size_billion_parameters=7,
336
+ context_length=8192,
337
  ),
338
  Model(
339
  "OpenOrca Mistral (7B) 8K",
 
341
  None,
342
  "Together AI",
343
  size_billion_parameters=7,
344
+ context_length=8192,
345
  ),
346
  Model(
347
  "Qwen-Chat (7B)",
 
349
  None,
350
  "Together AI",
351
  size_billion_parameters=7,
352
+ context_length=8192,
353
  ),
354
  Model(
355
  "Qwen 1.5 Chat (0.5B)",
 
357
  None,
358
  "Together AI",
359
  size_billion_parameters=0.5,
360
+ context_length=32768,
361
  ),
362
  Model(
363
  "Qwen 1.5 Chat (1.8B)",
364
  "together_ai/Qwen/Qwen1.5-1.8B-Chat",
365
  None,
366
  "Together AI",
367
+ context_length=32768,
368
  size_billion_parameters=1.8,
369
  ),
370
  Model(
 
373
  None,
374
  "Together AI",
375
  size_billion_parameters=4,
376
+ context_length=32768,
377
  ),
378
  Model(
379
+ "Qwen 1.5 Chat (7B)",
380
+ "together_ai/Qwen/Qwen1.5-7B-Chat",
381
+ None,
382
+ "Together AI",
383
+ context_length=32768,
384
+ size_billion_parameters=7
385
  # selected=True
386
  ),
387
  Model(
 
390
  None,
391
  "Together AI",
392
  size_billion_parameters=14,
393
+ context_length=32768,
394
  ),
395
  Model(
396
  "Qwen 1.5 Chat (72B)",
397
  "together_ai/Qwen/Qwen1.5-72B-Chat",
398
  None,
399
  "Together AI",
400
+ context_length=4096,
401
  ),
402
  Model(
403
  "Snorkel Mistral PairRM DPO (7B)",
404
  "together_ai/snorkelai/Snorkel-Mistral-PairRM-DPO",
405
  None,
406
  "Together AI",
407
+ context_length=32768,
408
  # selected=True,
409
  ),
410
  Model(
 
412
  "together_ai/togethercomputer/alpaca-7b",
413
  None,
414
  "Together AI",
415
+ context_length=2048,
416
  ),
417
  Model(
418
  "OpenHermes-2-Mistral (7B)",
419
  "teknium/OpenHermes-2-Mistral-7B",
420
  None,
421
  "Together AI",
422
+ context_length=8192,
423
  ),
424
  Model(
425
  "OpenHermes-2.5-Mistral (7B)",
426
  "together_ai/teknium/OpenHermes-2p5-Mistral-7B",
427
  None,
428
  "Together AI",
429
+ context_length=8192,
430
  # selected=True,
431
  ),
432
  Model(
 
434
  "together_ai/togethercomputer/falcon-40b-instruct",
435
  None,
436
  "Together AI",
437
+ context_length=2048,
438
  ),
439
  Model(
440
  "Falcon Instruct (7B)",
441
  "together_ai/togethercomputer/falcon-7b-instruct",
442
  None,
443
  "Together AI",
444
+ context_length=2048,
445
  # selected=True,
446
  ),
447
  Model(
 
449
  "together_ai/togethercomputer/Llama-2-7B-32K-Instruct",
450
  None,
451
  "Together AI",
452
+ context_length=32768,
453
  ),
454
  Model(
455
  "RedPajama-INCITE Chat (3B)",
 
457
  None,
458
  "Together AI",
459
  size_billion_parameters=3,
460
+ context_length=2048,
461
  ),
462
  Model(
463
  "RedPajama-INCITE Chat (7B)",
464
  "together_ai/togethercomputer/RedPajama-INCITE-7B-Chat",
465
  None,
466
  "Together AI",
467
+ context_length=2048,
468
  size_billion_parameters=7,
469
  # selected=True,
470
  ),
 
473
  "together_ai/togethercomputer/StripedHyena-Nous-7B",
474
  None,
475
  "Together AI",
476
+ context_length=32768,
477
  size_billion_parameters=7,
478
  ),
479
  Model(
 
481
  "together_ai/Undi95/ReMM-SLERP-L2-13B",
482
  None,
483
  "Together AI",
484
+ context_length=4096,
485
  size_billion_parameters=13,
486
  ),
487
  Model(
 
489
  "together_ai/Undi95/Toppy-M-7B",
490
  None,
491
  "Together AI",
492
+ context_length=4096,
493
+ size_billion_parameters=7,
494
  ),
495
  Model(
496
  "WizardLM v1.2 (13B)",
497
  "together_ai/WizardLM/WizardLM-13B-V1.2",
498
  None,
499
  "Together AI",
500
+ context_length=4096,
501
  size_billion_parameters=13,
502
  # selected=True,
503
  ),
 
506
  "together_ai/upstage/SOLAR-10.7B-Instruct-v1.0",
507
  None,
508
  "Together AI",
509
+ context_length=4096,
510
  size_billion_parameters=11,
511
  # selected=True,
512
  ),