Post by joitarani333 on Apr 30, 2024 5:01:09 GMT
Comparison results for MTS AI InstructruK The following graph shows the results of comparing various Russian language models with gpt.turbo on our MTS AI InstructruK benchmark. We did not arrive at this result right away our first version of MTS AI Chet defeater Chatty in only of cases. We also calculate which categories and which model wins. MTS AI Chat copes best with the categories of information extraction from text and opine and in the classification task we are inferior to GigaChatPro. The best answers in the closeeqa creative writing and brainstorming categories were from YandexGPT.
In other categories GigaChatPro and GigaChatPro share the championship. It is noticeable how the Restaurant Email List GigaChat models are somewhat similar in each category while YandexGPT is shifte to the right performing poorly on tasks of extracting information from text and summarizing. Lets look at specific examples of questions and answers for MTS AI ChatB and ChatGPT The case when MTS AI ChatB won over ChatGPT detecting evidence from context ChatGPT answer. the answer is already in the context of the question. ChatGPTs answer. The models answer is incorrect because the answer is already.
In the context of the question. Reply MTS AI ChatB. Reply MTS AI ChatB. A case where both models responde well writing simple code via a text request ChatGPT response is correct the code works. ChatGPTs answer is correct the code works. The answer from MTS AI ChatB is also correct the code is working. The answer from MTS AI ChatB is also correct the code is working. Case where ChatGPT answere better than MTS AI ChatB composing acronyms ChatGPT answer. All acronyms are spelle correctly. ChatGPTs answer. All acronyms are spelle correctly. Reply MTS AI ChatB. All acronyms are spelle incorrectly.
In other categories GigaChatPro and GigaChatPro share the championship. It is noticeable how the Restaurant Email List GigaChat models are somewhat similar in each category while YandexGPT is shifte to the right performing poorly on tasks of extracting information from text and summarizing. Lets look at specific examples of questions and answers for MTS AI ChatB and ChatGPT The case when MTS AI ChatB won over ChatGPT detecting evidence from context ChatGPT answer. the answer is already in the context of the question. ChatGPTs answer. The models answer is incorrect because the answer is already.
In the context of the question. Reply MTS AI ChatB. Reply MTS AI ChatB. A case where both models responde well writing simple code via a text request ChatGPT response is correct the code works. ChatGPTs answer is correct the code works. The answer from MTS AI ChatB is also correct the code is working. The answer from MTS AI ChatB is also correct the code is working. Case where ChatGPT answere better than MTS AI ChatB composing acronyms ChatGPT answer. All acronyms are spelle correctly. ChatGPTs answer. All acronyms are spelle correctly. Reply MTS AI ChatB. All acronyms are spelle incorrectly.