{"id":376739,"date":"2021-12-05T13:00:12","date_gmt":"2021-12-05T10:00:12","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/"},"modified":"2021-12-05T13:00:12","modified_gmt":"2021-12-05T10:00:12","slug":"building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/","title":{"rendered":"#Building apps with GPT-3? Here&#8217;s how to balance cost and performance"},"content":{"rendered":"<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_85 counter-hierarchy ez-toc-counter ez-toc-custom ez-toc-container-direction\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">Table of Contents<\/p>\n<label for=\"ez-toc-cssicon-toggle-item-6a3717aed333a\" class=\"ez-toc-cssicon-toggle-label\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #dd3333;color:#dd3333\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #dd3333;color:#dd3333\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/label><input type=\"checkbox\"  id=\"ez-toc-cssicon-toggle-item-6a3717aed333a\" checked aria-label=\"Toggle\" \/><nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/#Models_and_tokens\" >Models and tokens<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/#Which_model_should_you_use\" >Which model should you use?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/#Balancing_costs_and_quality\" >Balancing costs and quality<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/#Finetuning_GPT-3\" >Finetuning GPT-3<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/buradabiliyorum.com\/en\/building-apps-with-gpt-3-heres-how-to-balance-cost-and-performance\/#GPT-3_alternatives\" >GPT-3 alternatives<\/a><\/li><\/ul><\/nav><\/div>\n<p>&#8220;<strong>#Building <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/download-scripts-themes-apps\/\" data-internallinksmanager029f6b8e52c=\"9\" title=\"Download Scripts &amp; Themes &amp; Apps\" target=\"_blank\" rel=\"noopener\">app<\/a>s with GPT-3? Here&#8217;s how to balance cost and performance<\/strong>&#8221;<\/p>\n<div>Last week, OpenAI removed the waitlist for the application programming interface to GPT-3, its flagship language model. Now, any developer who meets the conditions for using the OpenAI API can apply and start integrating GPT-3 into their applications.<\/p>\n<p>Since the beta release of GPT-3, developers have built hundreds of applications on top of the language model. But <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/gpt-3-startups-businesses\">building successful GPT-3 products<\/a> presents unique challenges. You must find a way to leverage the power of OpenAI\u2019s advanced deep learning models to provide the best value to your users while keeping your operations <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/gpt-3-economy-business-model\">scalable and cost-efficient<\/a>.<\/p>\n<p>Fortunately, OpenAI provides a variety of options that can help you make the best use of your money when using GPT-3. Here\u2019s what the people who have been developing applications with GPT-3 have to say about best practices.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Models_and_tokens\"><\/span>Models and tokens<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<figure class=\"post-image post-mediaBleed aligncenter\"><img decoding=\"async\" loading=\"lazy\" alt=\"OpenAI provides GPT-3 in different sizes, prices, and performance levels.\" width=\"3190\" height=\"626\" class=\"js-lazy\" src=\"https:\/\/i1.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-model-sizes-and-prices.jpg?ssl=1\"\/><figcaption><a rel=\"nofollow noopener\" target=\"_blank\" href=\"#\" data-url=\"https:\/\/twitter.com\/intent\/tweet?url=https%3A%2F%2Feditorial.thenextweb.com%2Fneural%2F2021%2F12%2F05%2Fbuilding-apps-gpt-3-what-devs-need-know-cost-performance-syndication%2F&amp;via=thenextweb&amp;related=thenextweb&amp;text=Check out this picture on: OpenAI provides GPT-3 in different sizes, prices, and performance levels.\" data-title=\"Share OpenAI provides GPT-3 in different sizes, prices, and performance levels. on Twitter\" data-width=\"685\" data-height=\"500\" class=\"post-image-share popitup\" title=\"Share OpenAI provides GPT-3 in different sizes, prices, and performance levels. on Twitter\"><i class=\"icon icon--inline icon--twitter--dark\"\/><\/a>OpenAI provides GPT-3 in different sizes, prices, and performance levels.<\/figcaption><noscript><img decoding=\"async\" loading=\"lazy\" src=\"https:\/\/i1.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-model-sizes-and-prices.jpg?ssl=1\" alt=\"OpenAI provides GPT-3 in different sizes, prices, and performance levels.\" width=\"3190\" height=\"626\" class=\"\" srcset=\"\"\/><\/noscript><\/figure>\n<p>OpenAI offers <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/openai.com\/api\/pricing\/\">four versions of GPT-3<\/a>: Ada, Babbage, Curie, and Davinci. Ada is the fastest, least expensive, and lowest-performing model. Davinci is the slowest, most expensive, and highest performing. Babbage and Curie are in-between the two extremes.<\/p>\n<p>OpenAI\u2019s website doesn\u2019t provide architectural details on each of the models, but <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/arxiv.org\/abs\/2005.14165\">the original GPT-3 paper<\/a> includes a list of different versions of the language model. The main difference between the models is the number of parameters and layers, going from 12 layers and 125 million parameters to 96 layers and 175 billion parameters. Adding layers and parameters improves the model\u2019s learning capacity but also increases the processing time and costs.<\/p>\n<p><img decoding=\"async\" alt=\"GPT-3 model size\" class=\"js-lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-model-size.jpg?resize=696%2C195&amp;ssl=1\"\/><\/p>\n<p><noscript><img decoding=\"async\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-model-size.jpg?resize=696%2C195&amp;ssl=1\" alt=\"GPT-3 model size\" class=\"\" srcset=\"\"\/><\/noscript><\/p>\n<p>OpenAI calculates the pricing of its models based on tokens. According to OpenAI, \u201cone token <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/general\/\" data-internallinksmanager029f6b8e52c=\"3\" title=\"General\" target=\"_blank\" rel=\"noopener\">general<\/a>ly corresponds to ~4 characters of text for common English text. This translates to roughly \u00be of a word (so 100 tokens ~= 75 words).\u201d<\/p>\n<p>Here\u2019s an example from <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/beta.openai.com\/tokenizer\">OpenAI\u2019s Tokenizer tool<\/a>:<\/p>\n<figure class=\"post-image post-mediaBleed aligncenter\"><img decoding=\"async\" loading=\"lazy\" alt=\"GPT-3 tokenizer example\" width=\"696\" height=\"565\" class=\"js-lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-tokenizer-example-1024x831.jpg?resize=696%2C565&amp;ssl=1\"\/><figcaption><a rel=\"nofollow noopener\" target=\"_blank\" href=\"#\" data-url=\"https:\/\/twitter.com\/intent\/tweet?url=https%3A%2F%2Feditorial.thenextweb.com%2Fneural%2F2021%2F12%2F05%2Fbuilding-apps-gpt-3-what-devs-need-know-cost-performance-syndication%2F&amp;via=thenextweb&amp;related=thenextweb&amp;text=Check out this picture on: Example of tokenized text\" data-title=\"Share Example of tokenized text on Twitter\" data-width=\"685\" data-height=\"500\" class=\"post-image-share popitup\" title=\"Share Example of tokenized text on Twitter\"><i class=\"icon icon--inline icon--twitter--dark\"\/><\/a>Example of tokenized text<\/figcaption><noscript><img decoding=\"async\" loading=\"lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-tokenizer-example-1024x831.jpg?resize=696%2C565&amp;ssl=1\" alt=\"GPT-3 tokenizer example\" width=\"696\" height=\"565\" class=\"\" srcset=\"\"\/><\/noscript><\/figure>\n<p>In general, if you use good English (avoid jargon, use simple words with few syllables, etc.), you\u2019ll get better token-to-word ratios. In the example below, aside from \u201cGPT-3,\u201d every other word counts as one token.<\/p>\n<p><img decoding=\"async\" alt=\"GPT-3 simple english tokenization\" class=\"js-lazy\" src=\"https:\/\/i2.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-simple-english-tokenization.jpg?resize=696%2C540&amp;ssl=1\"\/><\/p>\n<p><noscript><img decoding=\"async\" src=\"https:\/\/i2.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-simple-english-tokenization.jpg?resize=696%2C540&amp;ssl=1\" alt=\"GPT-3 simple english tokenization\" class=\"\" srcset=\"\"\/><\/noscript><\/p>\n<p>One of the benefits of GPT-3 is its <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/what-is-one-shot-learning\">few-shot learning capabilities<\/a>. If you\u2019re not satisfied with the model\u2019s response to a prompt, you can guide it by giving it a longer prompt that includes correct examples. These examples will work like real-time training and improve GPT-3\u2019s results without the need to readjust its parameters.<\/p>\n<p>It is worth noting that OpenAI charges you for the total tokens in your input prompt plus the output tokens GPT-3 returns. Therefore, long prompts with few-shot learning examples will increase the cost of using GPT-3.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Which_model_should_you_use\"><\/span>Which model should you use?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>With a 75x cost difference between the cheapest and most expensive GPT-3 models, it is important to know which option best suits your application.<\/p>\n<p>Matt Shumer, the co-founder and CEO of OthersideAI, has used GPT-3 to develop AI-powered writing tools. <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/hyperwriteai.com\/\">HyperWrite<\/a>, OthersideAI\u2019s main product, uses GPT-3 for text generation, autocomplete, and rephrasing.<\/p>\n<p>When choosing between different GPT-3 models, Shumer starts by considering the complexity of the intended use case, he told TechTalks.<\/p>\n<p>\u201cIf it\u2019s something simple, like binary classification, I might start with Ada or Babbage. If it\u2019s something very complex, like conditional generation where high-quality output and reliability is necessary, I start with Davinci,\u201d he said.<\/p>\n<p>When unsure of complexity, Shumer starts by trying the biggest model, Davinci. Then, he works his way down toward the smaller models.<\/p>\n<p>\u201cWhen I get it working with Davinci, I try to modify the prompt to use Curie. This typically means adding more examples, refining the structure, or both. If it works on Curie, I move to Babbage, then Ada,\u201d he said.<\/p>\n<p>For some applications, he uses a multi-step system that includes a mix of different models.<\/p>\n<p>\u201cFor example, if it\u2019s a generative task that requires some classification as a precursor step, I might use Babbage for the classification, then Curie or Davinci for the generative step,\u201d he said. \u201cAfter using it for a while, you get a feel for what might be useful for different use cases.\u201d<\/p>\n<figure class=\"post-image post-mediaBleed aligncenter\"><img decoding=\"async\" loading=\"lazy\" alt=\"openai gpt-3 playground\" width=\"696\" height=\"540\" class=\"js-lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/openai-gpt-3-playground.jpg?resize=696%2C540&amp;ssl=1\"\/><figcaption><a rel=\"nofollow noopener\" target=\"_blank\" href=\"#\" data-url=\"https:\/\/twitter.com\/intent\/tweet?url=https%3A%2F%2Feditorial.thenextweb.com%2Fneural%2F2021%2F12%2F05%2Fbuilding-apps-gpt-3-what-devs-need-know-cost-performance-syndication%2F&amp;via=thenextweb&amp;related=thenextweb&amp;text=Check out this picture on: OpenAI\u2019s Playground lets you directly try prompts on different GPT-3 models\" data-title=\"Share OpenAI\u2019s Playground lets you directly try prompts on different GPT-3 models on Twitter\" data-width=\"685\" data-height=\"500\" class=\"post-image-share popitup\" title=\"Share OpenAI\u2019s Playground lets you directly try prompts on different GPT-3 models on Twitter\"><i class=\"icon icon--inline icon--twitter--dark\"\/><\/a>OpenAI\u2019s Playground lets you directly try prompts on different GPT-3 models<\/figcaption><noscript><img decoding=\"async\" loading=\"lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/openai-gpt-3-playground.jpg?resize=696%2C540&amp;ssl=1\" alt=\"openai gpt-3 playground\" width=\"696\" height=\"540\" class=\"\" srcset=\"\"\/><\/noscript><\/figure>\n<p>Paul Bellow, author and developer of <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.litrpgadventures.com\/\">LitRPG Adventures<\/a>, used Davinci for his GPT-3-powered RPG content generator.<\/p>\n<p>\u201cI wanted to generate the highest quality output possible\u2014for later fine-tuning,\u201d Bellow told TechTalks. \u201cDavinci is the slowest and most expensive, but the tradeoff is higher quality output which was important to me at this stage of development. I\u2019ve spent a premium, but I now have over 10,000 generations that I can use for future fine-tuning. Datasets have value.\u201d (More on fine-tuning later.)<\/p>\n<p>Bellow says that the best way to find out if another model is going to work for a task is to run some tests on Playground, a tool you can use to directly try prompts on different GPT-3 models (note that OpenAI bills you for using Playground).<\/p>\n<p>\u201cA lot of the time, a well-thought-out prompt can get good content out of the Curie model. It all just depends on the use-case,\u201d Bellow said.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Balancing_costs_and_quality\"><\/span>Balancing costs and quality<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>When choosing a model for your application, you\u2019ll have to weigh the balance between the cost and value. Choosing a high-performing model might provide better quality output, but the improved results might not justify the price difference.<\/p>\n<p>\u201cYou have to build a business model around your product that supports the engines you\u2019re using,\u201d Shumer said. \u201cIf you want high-quality outputs for your users, it\u2019ll be worth it to use Davinci\u2014you can pass off the costs to your users. If you\u2019re looking to build a large-scale free product, and your users are okay with mediocre results, you can use a smaller engine. It all depends on your product goals.\u201d<\/p>\n<p>OthersideAI has developed a solution that uses a mix of different GPT-3 models to enable different use cases, Shumer said. Paid users enjoy the power of large GPT-3 models, while free-tier users get access to the smaller models.<\/p>\n<p>For LitRPG Adventures, quality is prime, which is why Bellow initially stuck to the Davinci model. He used the base Davinci model with one- or two-shot prompts, which increased the costs but made sure GPT-3 provided quality output.<\/p>\n<p>\u201cOpenAI API Davinci model is a bit expensive at this time, but I see the cost going down eventually,\u201d he said. \u201cWhat provides flexibility right now is the ability to fine-tune the Curie and lower models, or Davinci with permission. This will bring my costs per generation down quite a bit while hopefully maintaining high quality.\u201d<\/p>\n<p>He has been able to develop a business model that maintains a profit margin while using Davinci.<\/p>\n<p>\u201cWhile not a huge money-maker, the LitRPG Adventures project is paying for itself and just about ready to scale up,\u201d he said.<\/p>\n<h2><span class=\"ez-toc-section\" id=\"Finetuning_GPT-3\"><\/span>Finetuning GPT-3<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>OpenAI\u2019s scientists initially introduced GPT-3 as a task-agnostic language model. According to their initial tests, GPT-3 rivaled state-of-the-art models on specific tasks without the need for further training. But they also mentioned fine-tuning as a \u201cpromising direction of future work.\u201d<\/p>\n<p>In the months that followed the beta release of GPT-3, OpenAI and Microsoft fine-tuned the model for a number of different tasks, including <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/microsoft-gpt-3-and-the-future-of-openai\">database query<\/a> and <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/openai-codex-ai-programming\">source-code generation<\/a>.<\/p>\n<p><a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/what-is-transfer-learning\">Like other deep learning architectures<\/a>, fine-tuning has several benefits for GPT-3. OpenAI API allows customers to create fine-tuned versions of its GPT-3 for a premium. You can create your own training dataset, upload it to OpenAI\u2019s servers, and use it to create a finetuned model of GPT-3. OpenAI will host your model and make it available to you through its API.<\/p>\n<p>Fine-tuning will enable you to tackle problems that are impossible to solve with the basic models.<\/p>\n<p>\u201cThe vanilla models are highly capable and are usable for many tasks. However, some tasks (i.e., multi-step generation) are too complex for a vanilla model, even Davinci, to complete with high accuracy,\u201d Shumer said. \u201cIn cases like this, you have two options: 1) create a prompt chain that feeds outputs from one prompt into another prompt, or 2) fine-tune a model. I typically first try to create a prompt chain, and if that doesn\u2019t work, I then move to fine-tuning.\u201d<\/p>\n<p>If done properly, fine-tuning can also reduce the costs of using GPT-3. If you\u2019ll be using GPT-3 for a specific application, a fine-tuned small model can produce results that are as good as those provided by a large vanilla model. Fine-tuned models also reduce the size of prompts, which further slashes your token usage.<\/p>\n<p>\u201cOne other case where I tend to fine-tune is when I can get something working with a vanilla model, but the prompt ends up being so long that it is costly to serve to users. In cases like these, I fine-tune, as it actually can reduce the overall serving costs,\u201d Shumer said.<\/p>\n<p>But fine-tuning isn\u2019t without challenges. Without a quality training dataset, finetuning can have adverse effects.<\/p>\n<p>\u201cClean your dataset as much as you can. Garbage in, garbage out is one of my big mantras now when it comes to prompt engineering,\u201d Bellow said.<\/p>\n<p>If you manage to gather a sizeable dataset of quality examples, however, fine-tuning can do wonders. After starting LitRPG with the Davinci model, Bellow gathered and cleaned a dataset of around 4,000 samples in a 7-megabyte JSON file. While he is still experimenting, the initial results show that he can move from Davinci to Curie without a noticeable change in quality, which reduces the costs of GPT-3 queries by 90 percent.<\/p>\n<p>Another consideration is the time it takes to fine-tune GPT-3, which grows with the size of the model and the training dataset.<\/p>\n<p>\u201cIt can take as little as five minutes to fine-tune a smaller model on a few hundred examples,\u201d Shumer said. \u201cI\u2019ve also seen cases where it takes upwards of five hours to train a larger model on thousands of examples.\u201d<\/p>\n<p>There\u2019s also an inverse correlation between the size of the model and the amount of data you need to fine-tune GPT-3, according to Shumer\u2019s experiments. Larger models require less data for fine-tuning.<\/p>\n<p>\u201cFor many tasks, you can think of increasing base model size as a way to reduce how much data you\u2019ll need to fine-tune a quality model,\u201d Shumer said. \u201cA Curie fine-tuned on 100 examples may have similar results to a Babbage fine-tuned on 2,000 examples. The larger models can do remarkable things with very little data.\u201d<\/p>\n<h2><span class=\"ez-toc-section\" id=\"GPT-3_alternatives\"><\/span>GPT-3 alternatives<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p><img decoding=\"async\" alt=\"GPT-3 alternative GPT-J\" class=\"js-lazy\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-alternative-GPT-J.jpg?resize=696%2C392&amp;ssl=1\"\/><\/p>\n<p><noscript><img decoding=\"async\" src=\"https:\/\/i0.wp.com\/bdtechtalks.com\/wp-content\/uploads\/2021\/11\/GPT-3-alternative-GPT-J.jpg?resize=696%2C392&amp;ssl=1\" alt=\"GPT-3 alternative GPT-J\" class=\"\" srcset=\"\"\/><\/noscript><\/p>\n<p>OpenAI received a lot of criticism for deciding <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/openai-gpt-3-commercial-ai\">not to release GPT-3 as an open-source model<\/a>. Subsequently, other developers released GPT-3 alternatives and made them available to the public. One very popular project is <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/6b.eleuther.ai\/\">GPT-J by EleutherAI<\/a>. Like other open-source projects, GPT-J requires technical effort on the part of application developers to set up and run. It also doesn\u2019t benefit from the ease of use and scalability that comes with hosting and fine-tuning your models on Microsoft\u2019s Azure cloud.<\/p>\n<p>But open-source models are nonetheless useful and are worth considering if you have the in-house talent to set them up and they meet your application\u2019s requirements.<\/p>\n<p>\u201cGPT-J isn\u2019t the same as full-scale GPT-3\u2014but it is useful if you know how to work with it. It\u2019s exponentially harder to get a complex prompt working on GPT-J, as compared with Davinci, but it is possible for most use-cases,\u201d Shumer said. \u201cYou won\u2019t get the same super high-quality output, but you can likely get to something passable with some time and effort. Plus, these models can be cheaper to run, which is a big plus, considering the cost of Davinci. We have successfully used models like these at Otherside.\u201d<\/p>\n<p>\u201cIn my experience, they operate at about the level of the Curie model from OpenAI,\u201d Bellow said. \u201cI\u2019ve also been looking into <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/cohere.ai\/\">Cohere AI<\/a>, but they\u2019re not giving details on the size of their model, so I imagine it\u2019s around the same as GPT-J, et al. I do think (hope) that there will be even more options soon from other players. Competition between suppliers is good for consumers like me.\u201d<\/p>\n<p><em>This article was originally published by Ben Dickson on<span>\u00a0<\/span><a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/\">TechTalks<\/a>, a publication that examines trends in <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/technology\/\" data-internallinksmanager029f6b8e52c=\"4\" title=\"Technology\" target=\"_blank\" rel=\"noopener\">technology<\/a>, how they affect the way we live and do business, and the problems they solve. But we also discuss the evil side of technology, the darker implications of new tech, and what we need to look out for. You can read the original article<span>\u00a0<\/span><a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/bdtechtalks.com\/news\/gpt-3-application-development-tips\">here<\/a>.<\/em><\/p>\n<\/div>\n<p><script async src=\"\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<blockquote><p><strong><span style=\"color: #ff6600;\">If you liked the article, do not forget to share it with your friends. Follow us on\u00a0<span style=\"color: #ff0000;\"><a style=\"color: #ff0000;\" href=\"https:\/\/news.google.com\/publications\/CAAqBwgKMLG0nwswvr63Aw\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">Google News<\/a><\/span>\u00a0too, click on the star and choose us from your favorites.<\/span><\/strong><\/p><\/blockquote>\n<blockquote>\n<p style=\"text-align: center;\">For forums sites go to <span style=\"color: #ff9900;\"><a style=\"color: #ff9900;\" href=\"https:\/\/forum.buradabiliyorum.com\/\" target=\"_blank\" rel=\"noopener\">Forum.BuradaBiliyorum.Com<\/a><\/span><\/strong>\n<\/p><\/blockquote>\n<blockquote>\n<p style=\"text-align: center;\"><strong>If you want to read more like this article, you can visit our <span style=\"color: #ff9900;\"><a style=\"color: #ff9900;\" href=\"https:\/\/en.buradabiliyorum.com\/technology\/\" target=\"_blank\" rel=\"noopener\">Technology category.<\/a><\/span><\/strong><\/p>\n<\/blockquote>\n<p><span style=\"color: black;\"><a style=\"color: #ff9900;\" href=\"https:\/\/thenextweb.com\/news\/building-apps-gpt-3-what-devs-need-know-cost-performance-syndication\" target=\"_blank\" rel=\"noopener\">Source<\/a><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>&#8220;#Building apps with GPT-3? Here&#8217;s how to balance cost and performance&#8221; Last week, OpenAI removed the waitlist for the application programming interface to GPT-3, its flagship language model. Now, any developer who meets the conditions for using the OpenAI API can apply and start integrating GPT-3 into their applications. Since the beta release of GPT-3,&#8230;<\/p>\n","protected":false},"author":1,"featured_media":376740,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/img-cdn.tnwcdn.com\/image\/neural?filter_last=1&fit=1280,640&url=https:\/\/cdn0.tnwcdn.com\/wp-content\/blogs.dir\/1\/files\/2021\/12\/GPT-3-applications-hed.jpg&signature=432035fb2bfb72ef9430e996b90ce8ab","fifu_image_alt":"","footnotes":""},"categories":[18],"tags":[],"class_list":["post-376739","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology"],"_links":{"self":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/376739","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/comments?post=376739"}],"version-history":[{"count":0,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/376739\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media\/376740"}],"wp:attachment":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media?parent=376739"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/categories?post=376739"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/tags?post=376739"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}