luisrguerra
commited on
Update index.html
Browse files- index.html +18 -3
index.html
CHANGED
@@ -46,21 +46,27 @@
|
|
46 |
<div id="tableBenchMark"></div>
|
47 |
<h4>Best models for solving math problems:</h4>
|
48 |
<ul>
|
|
|
|
|
49 |
<li>gpt-4-0125-preview (turbo)</li>
|
50 |
<li>gpt-4-1106-preview (turbo)</li>
|
51 |
<li>gpt-4-0613</li>
|
52 |
<li>gpt-4-0314</li>
|
53 |
<li>Gemini Ultra 1.0</li>
|
54 |
<li>Gemini Pro 1.5</li>
|
|
|
55 |
<li>Claude 3 Opus</li>
|
56 |
<li>Claude 3 Sonnet</li>
|
57 |
</ul>
|
58 |
<h4>Best models for large text:</h4>
|
59 |
<ul>
|
|
|
|
|
60 |
<li>gpt-4-0125-preview (turbo)</li>
|
61 |
<li>gpt-4-1106-preview (turbo)</li>
|
62 |
-
<li>Gemini Ultra</li>
|
63 |
<li>Gemini Pro 1.5</li>
|
|
|
64 |
<li>Claude 3 Opus</li>
|
65 |
<li>Claude 3 Sonnet</li>
|
66 |
<li>Claude 3 Haiku</li>
|
@@ -69,6 +75,7 @@
|
|
69 |
</ul>
|
70 |
<h4>Models with the best cost benefit:</h4>
|
71 |
<ul>
|
|
|
72 |
<li>Gemini Pro 1.5</li>
|
73 |
<li>gpt-3.5-turbo-0125</li>
|
74 |
<li>gpt-3.5-turbo-0613</li>
|
@@ -98,10 +105,11 @@
|
|
98 |
<ul>
|
99 |
<li>Mixtral 8x7B Instruct</li>
|
100 |
<li>Mistral 7B</li>
|
|
|
101 |
<li>Yi 34B</li>
|
102 |
<li>Grok 1</li>
|
103 |
<li>DBRX Instruct</li>
|
104 |
-
<li>Llama
|
105 |
<li>Gemma 2-7B</li>
|
106 |
</ul>
|
107 |
<h4>Can be trained in online service:</h4>
|
@@ -112,6 +120,7 @@
|
|
112 |
</ul>
|
113 |
<h4>Can be trained locally:</h4>
|
114 |
<ul>
|
|
|
115 |
<li>Mixtral 8x7B Instruct</li>
|
116 |
<li>Yi 34B</li>
|
117 |
</ul>
|
@@ -137,8 +146,9 @@
|
|
137 |
</ul>
|
138 |
<h4>Models with the same level of GPT-4 but lower than GPT-4 Turbo:</h4>
|
139 |
<ul>
|
140 |
-
<li>Gemini Ultra</li>
|
141 |
<li>Gemini Pro 1.5</li>
|
|
|
142 |
<li>Gemini Pro (Bard/Online)</li>
|
143 |
<li>Claude 3 Sonnet</li>
|
144 |
</ul>
|
@@ -155,6 +165,8 @@
|
|
155 |
</ul>
|
156 |
<h4>Versions of models already surpassed by fine-tune, new versions or new architectures:</h4>
|
157 |
<ul>
|
|
|
|
|
158 |
<li>Gemini Pro 1.0</li>
|
159 |
<li>Grok 1</li>
|
160 |
<li>Phi-2</li>
|
@@ -167,6 +179,7 @@
|
|
167 |
<li>gpt-4-0314</li>
|
168 |
<li>Claude 2-2.1</li>
|
169 |
<li>Claude Instant 1-1.2</li>
|
|
|
170 |
<li>Falcon 180B</li>
|
171 |
<li>Llama 1 and Llama 2</li>
|
172 |
<li>Guanaco 65B</li>
|
@@ -174,6 +187,8 @@
|
|
174 |
<li>Dolly V2</li>
|
175 |
<li>Alpaca</li>
|
176 |
<li>CodeLlama-34b-Instruct-hf</li>
|
|
|
|
|
177 |
<li>Mistral-7B-v0.1</li>
|
178 |
<li>MythoMax-L2</li>
|
179 |
<li>Zephyr 7B Alpha and Beta</li>
|
|
|
46 |
<div id="tableBenchMark"></div>
|
47 |
<h4>Best models for solving math problems:</h4>
|
48 |
<ul>
|
49 |
+
<li>gpt-4o-2024-05-13</li>
|
50 |
+
<li>gpt-4-Turbo-2024-04-09</li>
|
51 |
<li>gpt-4-0125-preview (turbo)</li>
|
52 |
<li>gpt-4-1106-preview (turbo)</li>
|
53 |
<li>gpt-4-0613</li>
|
54 |
<li>gpt-4-0314</li>
|
55 |
<li>Gemini Ultra 1.0</li>
|
56 |
<li>Gemini Pro 1.5</li>
|
57 |
+
<li>Gemini Advanced</li>
|
58 |
<li>Claude 3 Opus</li>
|
59 |
<li>Claude 3 Sonnet</li>
|
60 |
</ul>
|
61 |
<h4>Best models for large text:</h4>
|
62 |
<ul>
|
63 |
+
<li>gpt-4o-2024-05-13</li>
|
64 |
+
<li>gpt-4-Turbo-2024-04-09</li>
|
65 |
<li>gpt-4-0125-preview (turbo)</li>
|
66 |
<li>gpt-4-1106-preview (turbo)</li>
|
67 |
+
<li>Gemini Ultra 1.0</li>
|
68 |
<li>Gemini Pro 1.5</li>
|
69 |
+
<li>Gemini Advanced</li>
|
70 |
<li>Claude 3 Opus</li>
|
71 |
<li>Claude 3 Sonnet</li>
|
72 |
<li>Claude 3 Haiku</li>
|
|
|
75 |
</ul>
|
76 |
<h4>Models with the best cost benefit:</h4>
|
77 |
<ul>
|
78 |
+
<li>gpt-4o-2024-05-13</li>
|
79 |
<li>Gemini Pro 1.5</li>
|
80 |
<li>gpt-3.5-turbo-0125</li>
|
81 |
<li>gpt-3.5-turbo-0613</li>
|
|
|
105 |
<ul>
|
106 |
<li>Mixtral 8x7B Instruct</li>
|
107 |
<li>Mistral 7B</li>
|
108 |
+
<li>Phi-3</li>
|
109 |
<li>Yi 34B</li>
|
110 |
<li>Grok 1</li>
|
111 |
<li>DBRX Instruct</li>
|
112 |
+
<li>Llama 3 8-70B</li>
|
113 |
<li>Gemma 2-7B</li>
|
114 |
</ul>
|
115 |
<h4>Can be trained in online service:</h4>
|
|
|
120 |
</ul>
|
121 |
<h4>Can be trained locally:</h4>
|
122 |
<ul>
|
123 |
+
<li>Llama 3 8-70B</li>
|
124 |
<li>Mixtral 8x7B Instruct</li>
|
125 |
<li>Yi 34B</li>
|
126 |
</ul>
|
|
|
146 |
</ul>
|
147 |
<h4>Models with the same level of GPT-4 but lower than GPT-4 Turbo:</h4>
|
148 |
<ul>
|
149 |
+
<li>Gemini Ultra 1.0</li>
|
150 |
<li>Gemini Pro 1.5</li>
|
151 |
+
<li>Gemini Advanced</li>
|
152 |
<li>Gemini Pro (Bard/Online)</li>
|
153 |
<li>Claude 3 Sonnet</li>
|
154 |
</ul>
|
|
|
165 |
</ul>
|
166 |
<h4>Versions of models already surpassed by fine-tune, new versions or new architectures:</h4>
|
167 |
<ul>
|
168 |
+
<li>gpt-4-0613</li>
|
169 |
+
<li>gpt-4-0314</li>
|
170 |
<li>Gemini Pro 1.0</li>
|
171 |
<li>Grok 1</li>
|
172 |
<li>Phi-2</li>
|
|
|
179 |
<li>gpt-4-0314</li>
|
180 |
<li>Claude 2-2.1</li>
|
181 |
<li>Claude Instant 1-1.2</li>
|
182 |
+
<li>Qwen 1.0</li>
|
183 |
<li>Falcon 180B</li>
|
184 |
<li>Llama 1 and Llama 2</li>
|
185 |
<li>Guanaco 65B</li>
|
|
|
187 |
<li>Dolly V2</li>
|
188 |
<li>Alpaca</li>
|
189 |
<li>CodeLlama-34b-Instruct-hf</li>
|
190 |
+
<li>SOLAR-10.7B-Instruct-v1.0</li>
|
191 |
+
<li>Mistral-7B-v0.2</li>
|
192 |
<li>Mistral-7B-v0.1</li>
|
193 |
<li>MythoMax-L2</li>
|
194 |
<li>Zephyr 7B Alpha and Beta</li>
|