{"id":11537,"date":"2024-05-28T11:49:08","date_gmt":"2024-05-28T03:49:08","guid":{"rendered":"https:\/\/www.1ai.net\/?p=11537"},"modified":"2024-05-28T11:49:20","modified_gmt":"2024-05-28T03:49:20","slug":"%e4%bd%bf%e7%94%a8fastchat%ef%bc%8c%e5%b0%b1%e8%83%bd%e5%85%8d%e8%b4%b9%e4%bd%bf%e7%94%a8gpt-4-turbo-%e5%92%8c-claude3","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/11537.html","title":{"rendered":"Use FastChat to use gpt-4-turbo and Claude3 for free"},"content":{"rendered":"

This time let\u2019s talk about Github 33.5k FastChat<\/a> project<\/strong>, which is an open platform for training, serving, and evaluating chatbots based on large language models.<\/p>\n

It provides support for Chatbot Arena, whichCollected 70+ large models for usersfree<\/a>use<\/strong>, and are ranked based on user votes.<\/p>\n

\n

Large model ranking address: https:\/\/chat.lmsys.org\/?leaderboard<\/p>\n<\/blockquote>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
\n

Ranking<\/p>\n<\/td>\n

\n

Model Name<\/p>\n<\/td>\n

\n

Average score<\/p>\n<\/td>\n

\n

Developers<\/p>\n<\/td>\n<\/tr>\n

\n

1<\/p>\n<\/td>\n

\n

GPT-4-Turbo-2024-04-09<\/p>\n<\/td>\n

\n

1260<\/p>\n<\/td>\n

\n

OpenAI<\/p>\n<\/td>\n<\/tr>\n

\n

1<\/p>\n<\/td>\n

\n

Claude 3 Opus<\/p>\n<\/td>\n

\n

1255<\/p>\n<\/td>\n

\n

Anthropic<\/p>\n<\/td>\n<\/tr>\n

\n

1<\/p>\n<\/td>\n

\n

GPT-4-1106-preview<\/p>\n<\/td>\n

\n

1254<\/p>\n<\/td>\n

\n

OpenAI<\/p>\n<\/td>\n<\/tr>\n

\n

2<\/p>\n<\/td>\n

\n

GPT-4-0125-preview<\/p>\n<\/td>\n

\n

1250<\/p>\n<\/td>\n

\n

OpenAI<\/p>\n<\/td>\n<\/tr>\n

\n

5<\/p>\n<\/td>\n

\n

Bard (Gemini Pro)<\/p>\n<\/td>\n

\n

1209<\/p>\n<\/td>\n

\n

Google<\/p>\n<\/td>\n<\/tr>\n

\n

5<\/p>\n<\/td>\n

\n

Claude 3 Sonnet<\/p>\n<\/td>\n

\n

1203<\/p>\n<\/td>\n

\n

Anthropic<\/p>\n<\/td>\n<\/tr>\n

\n

7<\/p>\n<\/td>\n

\n

Command R+<\/p>\n<\/td>\n

\n

1193<\/p>\n<\/td>\n

\n

Cohere<\/p>\n<\/td>\n<\/tr>\n

\n

7<\/p>\n<\/td>\n

\n

GPT-4-0314<\/p>\n<\/td>\n

\n

1189<\/p>\n<\/td>\n

\n

OpenAI<\/p>\n<\/td>\n<\/tr>\n

\n

9<\/p>\n<\/td>\n

\n

Claude 3 Haiku<\/p>\n<\/td>\n

\n

1182<\/p>\n<\/td>\n

\n

Anthropic<\/p>\n<\/td>\n<\/tr>\n

\n

10<\/p>\n<\/td>\n

\n

GPT-4-0613<\/p>\n<\/td>\n

\n

1164<\/p>\n<\/td>\n

\n

OpenAI<\/p>\n<\/td>\n<\/tr>\n

\n

10<\/p>\n<\/td>\n

\n

Mistral-Large-2402<\/p>\n<\/td>\n

\n

1158<\/p>\n<\/td>\n

\n

Mistral<\/p>\n<\/td>\n<\/tr>\n

\n

11<\/p>\n<\/td>\n

\n

Qwen1.5-72B-Chat<\/p>\n<\/td>\n

\n

1154<\/p>\n<\/td>\n

\n

Alibaba<\/p>\n<\/td>\n<\/tr>\n

\n

12<\/p>\n<\/td>\n

\n

Claude-1<\/p>\n<\/td>\n

\n

1150<\/p>\n<\/td>\n

\n

Anthropic<\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

In Chatbot Arena, we canFree access to 30+ large models including Claude-3, ChatGPT4 and Gemini Pro<\/strong>, and provides users withFour types of conversations<\/strong>, each way allows for an ongoing conversation.<\/p>\n

Chatbot Arena: https:\/\/chat.lmsys.org\/<\/strong><\/p>\n

Next, I will introduce these four methods to you respectively:<\/p>\n

Two big models PK<\/h1>\n

There are two ways to play two big models PK:<\/p>\n

    \n
  • Anonymous large model PK<\/strong> : Select the "Arena(battle)<\/strong>", it will provide two anonymous models (such as ChatGPT, Claude, Llama) to users, and users will vote for their favorite one after using it!<\/li>\n<\/ul>\n
    \"get-684\"<\/div>\n

    For example, let\u2019s ask the following question:<\/p>\n

    \"get-686\"<\/div>\n

    As you can see, the two models will answer together, and then we can choose A is better or B is better based on the answer. After the selection, the names of the two models will be told to us.<\/p>\n

    If you are not satisfied with the answer, you can select "Regenerate" to regenerate the answer, or directly start a new "New Round".<\/p>\n

      \n
    • Choose any two large models to PK<\/strong> : Select theArena (side-by-side)<\/strong>", under this tab we can select any two large models to ask questions.<\/li>\n<\/ul>\n
      \"get-685\"<\/div>\n
      \"get-688\"<\/div>\n

      In addition to PK chat, direct chat with a single model is also supported.<\/p>\n

      Single Model Chat<\/h1>\n

      Select theDirect Chat<\/strong>"You can select a single model to have a conversation just like you usually do with ChatGPT.<\/p>\n

      \"get-687\"<\/div>\n

      There are many models, you can choose:<\/p>\n

      \"get-689\"<\/div>\n

      Parsing pictures<\/h1>\n

      Select theVision Direct Chat<\/strong>", you can have a picture conversation, each conversation can only use one image to chat. And the image size must be less than 15 MB.<\/p>\n

      \"get-690\"<\/div>\n

      All the big models on this website are free to use, so we can use them as much as we want.<\/p>\n

      If you are interested, go and try it.<\/p>\n

       <\/p>","protected":false},"excerpt":{"rendered":"

      Let's talk about the 33.5k FastChat project on Github, an open platform for training, serving, and evaluating chatbots based on large language models. It powers Chatbot Arena, a collection of 70+ large models that are free to use and ranked based on user votes. Ranking of large models is available at: https:\/\/chat.lmsys.org\/?leaderboard Rank Model Name Average Score Developer 1 GPT-4-Turbo-2024-04-09 1260 OpenAI 1 Claude 3 Opus 1255 Anthropic 1 GP<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[144],"tags":[1494,2812,385,2443,206],"collection":[],"class_list":["post-11537","post","type-post","status-publish","format-standard","hentry","category-baike","tag-claude3","tag-fastchat","tag-github","tag-gpt-4-turbo","tag-free"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/11537","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=11537"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/11537\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=11537"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=11537"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=11537"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=11537"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}