Thursday, 2 February 2023

Present and Future of Artificial Intelligence

 Mazhar Ali Dootio

Artificial Intelligence (AI) is an emerging field of technology that is rapidly transforming the way we live, work and interact with the world around us. It involves the development of algorithms and computer systems that are capable of performing tasks that would normally require human intelligence, such as speech recognition, image classification, and decision making. It is a rapidly growing field that has been making tremendous progress in recent years. It also includes the development of algorithms and computer systems that are capable of performing tasks that would normally require human intelligence.

The present state of AI is already making a big impact in various industries. For instance, in healthcare, AI is being used to improve patient outcomes through more accurate diagnoses and treatments. AI algorithms are being used to analyze medical images and identify potential health issues, such as cancer, before they become more serious. In finance, AI is being used to detect and prevent fraud, as well as to make more informed investment decisions by analyzing large amounts of data and identifying patterns.

In the near future, AI is expected to continue its rapid growth and to become even more integrated into our daily lives. It is likely that AI will be used to automate many tasks that are currently performed by humans, such as customer service, data analysis, and even some forms of creative work. This could result in significant benefits for society, including increased efficiency and productivity, but also raise ethical and social questions about job displacement and the role of AI in decision making.

The future of AI holds even more exciting possibilities. In the next decade, AI is expected to become even more integrated into our daily lives, performing tasks that are currently performed by humans, such as customer service and data analysis. For example, chatbots powered by AI can respond to customer inquiries 24/7, reducing the need for human customer service representatives. AI-powered virtual assistants are also becoming more common, helping us manage our daily tasks, such as scheduling appointments and ordering groceries.

Another important area where AI is expected to make a big impact is in robotics. AI will play a crucial role in enabling robots to perform a wider range of tasks and to interact more naturally with humans. This could result in significant improvements in fields such as manufacturing, logistics, and healthcare, as well as the development of new technologies and applications that we can't yet imagine. For example, AI-powered robots are already being used in hospitals to assist with tasks such as transporting equipment and supplies.

AI is also expected to have a significant impact on education. AI algorithms can personalize learning and provide students with more effective and engaging learning experiences. For example, AI-powered learning platforms can analyze a student's learning style and pace and adjust the material accordingly, providing a more personalized education experience.

In conclusion, the present and future of AI is both exciting and challenging. While AI has the potential to bring significant benefits to society, it also raises important ethical and social questions that must be carefully considered. Nevertheless, there is no doubt that AI will play a major role in shaping the future, and it is important that we continue to invest in research and development in this field to ensure that the benefits of AI are realized in the most positive and responsible way possible.

Tuesday, 31 January 2023

Background and Significance of ChatGPT in the digital age


Mazhar Ali Dootio

Technology is constantly evolving and it is likely that we see significant advances in fields such as artificial intelligence, renewable energy, biotechnology, and robotics etc. These developments could have major implications for industries, economies, and societies around the world.

AI refers to the development of computer systems that are able to perform tasks that normally require human intelligence, such as learning, problem-solving, and decision-making. AI has the potential to revolutionize many industries by automating tasks and processes, improving efficiency, and providing new insights and solutions.

There are several applications of AI for various purposes however, the one key application of AI is in the field of natural language processing (NLP) in modern age is ChatGPT-3, which involves the development of computer systems that can understand, interpret, and respond to human language. ChatGPT is a state of the art NLP system that is developed by OPenAI, a San Francisco-based company, that is able to generate human-like text based on a given prompt. ChatGPT was made available for public testing on 30th November 2022. This technology has the potential to transform industries such as customer service, education, and content creation, by providing more accurate and efficient language processing capabilities. The chatGPT-3 is a chatbot that uses the GPT-3 language model to generate responses to user inputs. It is designed to be able to engage in open-ended conversations with users on a variety of topics, and can provide information, answer questions, and even tell jokes or stories.

GPT-3 (short for "Generative Pre-training Transformer 3") is a language processing artificial intelligence model. It is one of the largest and most powerful language models that has been developed to date, with the ability to process and understand natural language inputs and generate human-like responses.

GPT-3 is a machine learning model that has been trained on a massive dataset of text from the internet, allowing it to learn the patterns and structures of human language. It uses this knowledge to generate coherent and reasonable responses to user inputs, whether that be answering a question, completing a prompt, or engaging in a conversation.

GPT-3 has a wide range of potential applications, including language translation, content generation, and chatbots like ChatGPT. It is an advanced and highly sophisticated AI model that continues to push the boundaries of what is possible with natural language processing and machine learning. GPT-3 is the third iteration of the Generative Pre-training Transformer (GPT) language model developed by OpenAI. The "3" in the name indicates that it is the third version of this model that has been released. The GPT language model was first introduced in 2018 with the release of GPT-1. This was followed by the release of GPT-2 in 2019, which was a significantly larger and more powerful version of the model. GPT-3 is the latest version of the model, and it is even larger and more powerful than its predecessor, GPT-2.

GPT-3 language model has a wide range of potential applications. Some of the ways in which GPT-3 could be used include:

           Language translation: GPT-3 could be used to translate text from one language to another, potentially even handling more complex tasks like translating idiomatic expressions or local slang.

           Content generation: GPT-3 could be used to generate content for websites, social media, or other platforms. It could be used to write articles, create social media posts, or even generate entire websites.

           Chatbots: GPT-3 could be used to create chatbots that are able to engage in open-ended conversations with users on a variety of topics. Cha tbots powered by GPT-3 could be used to provide customer service, answer frequently asked questions, or even just provide a fun and engaging way for users to pass the time.

           Text summarization: GPT-3 could be used to summarize long articles or documents, providing a concise summary of the main points and ideas.

           Question answering: GPT-3 could be used to answer questions on a wide range of topics, potentially even handling more complex and open-ended questions.

These are just a few examples of the many potential applications of GPT-3. As a highly sophisticated and powerful language processing AI model, it has the potential to be used in a wide variety of contexts and industries.

Instead, GPT-3 is trained on a large dataset of texts in a specific language, and it uses this training to generate text in that language that is similar to the training data. GPT-3 is available in several different versions, each of which is trained on a different dataset and optimized for a specific language or set of languages. Currently, GPT-3 is available in English, French, Spanish. German, Chinese (Simplified), Portuguese (Brazilian) and Dutch languages, however, it may translate in various languages of the world:

Each version of GPT-3 is trained on a different dataset and optimized for a specific language or set of languages, so the quality of the generated text will vary depending on the version you use. To use ChatGPT, you can start by typing a message or asking a question. ChatGPT will then generate a response based on the input you provided, using the GPT-3 language model to understand the context and meaning of your message and generate a relevant response. You can continue the conversation by responding to ChatGPT's message, and the chatbot will continue to generate responses in a way that tries to mimic natural human conversation.

The use of AI and chatGPT also has the potential to improve the quality of life for individuals. For example, AI-powered personal assistants such as Apple's Siri and Amazon's Alexa can assist with tasks such as setting reminders, answering questions, and providing information, freeing up time and cognitive resources for more important tasks. ChatGPT can also be used to generate personalized content and recommendations, improving the user experience and enabling more informed decision-making.

It is possible that GPT-3 and other advanced language generation models could potentially reduce the need for some language-related tasks that are currently performed by Google and other search engines. For example, GPT-3 can generate natural language text in response to prompts, which could potentially be used to generate content for websites or to answer questions in chatbots and other applications.

However, it is important to note that GPT-3 is just one tool among many that can be used for language-related tasks, and it is not designed to replace search engines like Google. Search engines like Google are designed to index and search the vast amount of information available on the internet, while GPT-3 is a language generation model that is capable of generating text based on the data it has been trained on.

So, while GPT-3 and other language generation models may potentially reduce the need for some language-related tasks, they are not likely to significantly reduce the significance or usage of search engines like Google.

It is also difficult to predict exactly what the effect of GPT-3 and other advanced language generation models will be on society and universities. However, these models have the potential to significantly impact a wide range of fields and industries, including education, journalism, marketing, and customer service, to name just a few. Therefore, in the field of education, GPT-3 and other language generation models could potentially be used to create personalized learning experiences for students, generate customized study materials, or even grade papers and assignments. In journalism, these models could potentially be used to generate news articles or to assist with fact-checking and research. In marketing, language generation models could be used to create personalized content for customers or to assist with customer service inquiries.

It is also possible that the widespread adoption of language generation models could lead to job displacement in some fields, as these models may be able to perform certain tasks more efficiently than humans. However, it is also possible that the use of these models could create new job opportunities in fields such as data science and machine learning. Generally, the impact of GPT-3 and other language generation models on society and universities is likely to be complex and multifaceted. It will be important to carefully consider the potential benefits and drawbacks of these technologies as they continue to develop and be adopted in a wide range of fields and industries.

GPT (Generative Pretrained Transformer) is a type of language model developed by OpenAI that is designed to generate human-like text. ChatGPT is a variant of GPT specifically optimized for conversation. It has been trained on a large dataset of human conversations and can generate responses to prompts in a way that is similar to how a human might respond.

Every technology created by humans has some limitations. Like all language models, ChatGPT has certain limitations. Here are a few:

1.      It is not a real person and does not have personal experiences or knowledge outside of what it has been trained on.

2.      It is not able to browse the internet or access new information, so it can only provide responses based on the information it has been provided.

3.      It is not able to understand and respond to every possible type of input. It may struggle to understand or provide a response to prompts that are very long, complex, or use uncommon words or phrases.

4.      It may generate responses that are inappropriate, offensive, or not relevant to the prompt.

5.      It is not able to engage in genuine conversation or understand the emotions or intentions of the person it is communicating with.

Overall, the significances of AI and chatGPT in the modern age is numerous and varied. These technologies have the potential to revolutionize industries, improve efficiency, and enhance the quality of life for individuals. As such, they will likely play an increasingly important role in the future.


Monday, 15 August 2022

تحقيق، محقق ۽ تحقيقي رويا


ڊاڪٽر مظھر علي ڏوتيو

دنيا ۾ تبديلين جو اهم سبب تحقيق ئي رهي آهي. تنهنڪري ڪنهن به سماج لاءِ تحقيق ۽ محقق اهم هوندا آهن. تحقيق محقق جي سڃاڻپ هوندي آهي تنهنڪري محقق کي تحقيق ڪرڻ دوران تحقيق جي هر پهلو جو خيال رکڻ گهرجي ته جيئن مقالو تحقيقي قانون ۽ قاعدن مطابق هجي ۽ نتيجا به تحقيق جي مقصدن مطابق هجن. بھتر نتيجن کي حاصل ڪرڻ لاءِ بھتر طريقيڪار يا ميٿوڊولوجي جي ڊزائين اھم ھوندي آھي. هڪ ڳاله محققن کي ياد رکڻ گهرجي ته غير معياري تحقيق ڪنهن نه ڪنهن سطح تي شرمندگي جو باعث ضرور بڻبي آهي. ڪجھ ماڻهو مقالن جي گهربل تعداد کي پورو ڪرڻ لاءِ يا پنهنجي تحقيقي مقالن جو تعداد وڌائڻ لاءِ مڙئي سڙئي مقالا لکي ڪنهن نه ڪنهن تحقيقي جرنل ۾ ڇپرائيندا آهن پر انهن مقالن جو معيار ڪو خاص نه هوندو آهي يا وري ڪي ماڻهو شروع ۾ غير معياري تحقيق ڪندا آهن جيڪا ڪنهن جرنل يا ڪانفرنس ۾ شايع ٿيندي اهي پر اڳتي هلي جڏهن انهن جي تحقيق معياري ٿي ويندي آهي ۽ دنيا جا تحقيق ڪندڙ ماڻهو انهن جي تحقيق کي پڙهندا آهن ۽ حوالا ڏيندا آهن ته ان وقت اهي محقق پنهنجي پراڻي تحقيق کي لڪائڻ جي ڪوشش ڪندا آهن ته جيئن هو ڪنهن قسم جي شرمندگي کان بچي سگهن
اسان وٽ هڪ ٻيو به رجحان آهي. ڪي تحقيق ڪندڙ پنهنجي شعبي کان سواءِ ٻين شعبن ۾ ٿيندڙ تحقيقي ڪم ۾ شامل ٿيڻ جي ڪوشش ڪندا آهن جنهن سان ڪا معياري تحقيق سامهون نه ايندي آهي. اهو رجحان ايترو ته وڌندو وڃي پيو جو ماڻهو پنهنجا نالا ٻين شعبن جي تحقيقي مقالن ۾ دوستي ياري جي رشتي سان به شامل ڪرائين ٿا، جيڪو تمام غلط ڪم اهي. مثال طور هڪ ماڻهو جي تحقيقي ڊگري ٻولي ۾ هجي ۽ هن جا مقالا وري ميڊيسن يا معاشيات ۾ ڇپجن ته اها هن جي پنهنجي فيلڊ ۽ ٻي فيلڊ سان زيادتي آهي. ااهو ئي سبب اهي جو اسان جو سماج، اسان جا تعليمي ۽ تحقيقي ادارا وغيره اڃا ڪا واضح ترقي نه ڪري سگهيا آهن. اسان وٽ اڪثر محقق تحقيقي حد بندين، قانون ۽ قاعدن جو خيال ڪون ڪن پيا. ھن موضوع تي منھنجو تفصيلي وڊيو موجود آھي جيڪو توھان سان ونڊجي ٿو. وڊيو کي ڏسڻ لاءِ ھيٺين لنڪ کي کوليو. اوھان ڪمينٽ باڪس ۾ پنھنجي راءِ جو اظھار ڪريو

Saturday, 13 August 2022

Significance of usage of AR and VR technologies for tourist places

Dr. Mazhar Ali Dootio 

Let's think and work...!!!

The technology combines virtual objects with physical objects to present original objects in a real time environment. Many historical monument buildings, archeological sites and tourist places in Sindh with important historical features have been designated as cultural reserves/heritages and tourist places. Due to unavailability of augmented reality (AR) technology based applications and information media, visitors do not know details about the historical places / monuments / archeological sites, and tourist places therefore, by implementing/using Augmented Reality (AR) technology, visitors can access information freely in a real time environment.
The AR technology presents the information and also displays 3D live images of monument buildings / archeological sites and tourist places using android based smart devices. AR technology is an effort to combine digital or virtual elements with physical objects to provide real time information about an object (that may be even monuments etc.), it runs interactively, and there is integration between objects in three dimensions.
The research and development projects based on AR technology may be given to various universities of Sindh by the Sindh government tourism department to develop AR technology based applications that may provide/create live information about these historical sites / monuments building and tourist places.
Implementation of AR technology on historical monuments/archeological sites and tourist places provides a new alternative that presents an information medium for historical monuments and archeological sites and tourist places.
There are genius minds available in the universities which may develop such types of AR technology based projects/applications and may present the live information about the culture and heritage to world

Monday, 8 August 2022

ڪمپيوٽر مشين لرننگ جو ماڊل برٽ ڇا آھي؟

 ڇا آھي؟BERTمشين لرننگ جو مآڊل  ڪمپيوٽر

ڊاڪٽر مظھر علي ڏوتيو

  جو.  Bidirectional Encoder Representationsبرٽ مخفف آھي

 برٽ مشين لرننگ جو  ھڪ ماڊل آھي جيڪو  نيچرل لئنگوئجز  پروسيسنگ  لاءِ تيار ڪيو ويو آھي. ھي ماڊل 2018ع ۾ گوگل جي آرٽيفيشل انٽيليجنس (ھٿراڌو ڏاھپ) ٻولين واري ليب ۾ محققن طرفان ٺاھيو ويو ھو.  ھي ماڊل ٻولي جي مختلف پھلوئن تي ڪم ڪري ٿو مثال طور: احساسن جو تجزيو، جملن کي خودڪار طريقي سان پيدا ڪرڻ، نيمڊ انٽٽي رڪگنيشن  وغيرھ . ڪمپيوٽر لاءِ ڪنھن بھ ٻولي کي سمجھڻ ھميشھ ھڪ ڏکيو عمل رھيو آھي پر ھٿراڌو ڏاھپ ۽ نيچرل لئنگوئيج پروسيسنگ جي ذريعي ڪمپيوٽرن ٻولين کي سمجھڻ ۽ انھن کي مختلف ڪمن جي استعمال لاءِ ھميشھ بھتر ڪم ڪيو آھي. ٻولين کي سمجھڻ لاءِ ڪمپيوٽر مشين جا ھٿراڌو ڏاھپ جا اھم اوزار نيچرل لئنگوئيج پروسيس ۽ مشين لرننگ رھيا آھن جن ذريعي متن ۽ ڳالھايل لفظن مان مطلب ڪڍڻ، انھن جي گرائمر کي سمجھڻ ۽  انھن مان جذبن کي ڪڍڻ آسان ڪم بڻجي ويا آھن.

برٽ ماڊل  کي ڪنھن بھ ٻولي  يا ان ۾ لکيل مواد يا سوالن ۽ جوابن وغيرھ کي سمجھڻ لاءِ ڪمپيوٽر مشين برٽ ماڊل کي مختلف ڪمن يا ٽاسڪس لاءِ استعمال ڪري ٿي. مثال طور:

ڪنھن بھ فلم يا ڪتاب جي تجزيي ڪرڻ وقت جيڪي رايا پيش ڪيا وڃن ٿا سي ڪيترا مثبت يا منفي آھي.   (جذبن جو تجزيو ڪرڻ)

چيٽ بوٽس کي ماڻھن طرفان پڇيل سوالن جي جواب ڏيڻ ۾ مدد ڪري ٿو   (سوالن جا جواب ڏيڻ)

ڪو بھ مواد لکڻ وقت يا جي ميل ذريعي ايميل ڪرڻ وقت اڳين جملي جي خودڪار نظام ذريعي پاڻ  اڳڪٿي ڪري تھ ھن قسم جو جملو استعمال ڪري سگھو ٿا. (متن يا جملي جي پيشن گوئي)

خودڪار نظام ذريعي ڪنھن بھ قسم جو آرٽيڪل لکڻ، جنھن لاءِ پھريان ڪجه جملا پاڻ لکڻ  کانپوءِ ڪمپيوٽر پنھنجو پاڻ آرٽيڪل لکي وٺي. (پنھنجو پاڻ جملا پيدا ڪرڻ يا لکڻ جنھن کي انگريزي ۾ ٽيڪسٽ جنريشن چئجي ٿو)

وڏن دستاويزن يا معاھدن خاص ڪري قانوني دستاويزن يا معاھدن جو خلاصو پيش ڪرڻ. (مواد جو خلاصو يا سمرائيزيشن)

اھڙا ڪيترائي لفظ موجود ھوندا آھن جن جون ڪيتريون ئي معنائون ٿي سگھن ٿيون. مثال لفظ وار، جنھنجو ٻھ معنائون تھ آساني سان ٿي سگھن ٿيون. وار معني مٿي يا جسم تي موجود وار، وار معني ڪھاڙي يا لٺ وغيرھ سان حملو.  (پوليسيمي رزوليوشن )


اھي ڪجھ خصوصيتيون آھن جيڪي مشين لرننگ جي ماڊل برٽ کي اھم بڻائي رھيو آھن. نيچرل لئنگوئيج پروسيسنگ گوگل جي ھر ڪم جي پٺيان ڪم ڪري رھي آھي. مثال طور ترجميڪاري وارو نظام، آواز وارو نظام وغيرھ. تنھنڪري ڪنھن بھ ٻولي کي عالمي سطح تي آڻڻ لاءِ ان تي مشين لرننگ يا ڊيپ لرننگ جي مختلف ماڊلز کي استعمال ڪيو وڃي ٿو پر انھن سمورن ماڊلن مان برٽ ھڪ اھم ماڊل آھي. جنھنکي ڪمپيوٽر سائنس جا شاگرد يا ٻولي تي ڪم ڪندڙ محقق استعمال ڪري سگھن ٿا ۽ پنھنجي ٻولين تي موجودھ دور جي گھرجن مطابق ڪم ڪري سگھن ٿا.

برٽ ماڊل ڪم ڪيئن ڪري ٿو؟

برٽ ماڊل ذريعي ڪمپيوٽر مشين کي ڪنھن بھ ٻوليءَ جي متن يا جملن جي ڀنڊار يا ڪارپس ذريعي سکيا ڏني وڃي ٿي. ياد رھي جملن جي ڀنڊار جو تعداد وڏو ھئڻ گھرجي، جنھنجو مطلب ٿيو تھ  جنھن بھ موضوع تي اوھان ڪمپيوٽر مشين کي سکيا ڏيڻ چاھيو ٿا تنھن موضوع جا گهٽ ۾ گهٽ ٽي چار ڪتابن جو ڪارپس يا جملن جو ڀنڊار ٺاھي ماڊل کي سکيا ڏيو. اھي ڪتاب ماڊل جي داخلا جي سرشتي ذريعي اِنپٽ ڪجن يا وجھجن. ان عمل سان ماڊل مختلف قسم جا لفظ، جملا ۽ خيال ملي ويندا، جنھن ڪري ھو اڳتي صحي اڳڪٿي ڪندو. جيڪڏھن سکيا صحي نھ ڪئي ويندي تھ نتيجا ۽ اڳڪٿيون بھ صحي نھ ٿينديون.

ماسڪ ٿيل ٻولي جو ماڊل ( Masked Language Model).


MLM  برٽ ماڊل کي قابل بڻائي ٿو تھ ھو ڪنھن بھ قسم جي متن کي ٻھ طرفي سکيا Bidirectional learning  ذريعي ان جو تجزيو ڪري يا ڪنھن بھ مسئلي کي حل ڪري مثال: ڪنھن بھ جملي مان ڪنھن بھ لفظ کي لڪائي ان کي ٻھ طرفي سکيا ذريعي حل ڪري سگھو ٿو. اھڙي قسم جو ڪم اڳ ۾ ڪڏھن بھ ڪون ٿيو ھو پر برٽ BERT ماڊل ذريعي اھو ڪم آساني سان ٿي سگھي ٿو.

مثال:  پئرس فرانس جو [ Mask] آھي. ھي جملو برٽ ماڊل ۾ سکيا دوران وڌو ويو. ھڪ مخصوص پروسيسنگ کان پوءِ برٽ ان جو جواب ڏنو

گادي جو ھنڌ    0.997 سيڪڙو

دل   0.001 سيڪڙو

مرڪز  0.000 سڪيڙو

سڀ کان ڪارائتو جواب گادي جو ھنڌ آھي. جڏھن تھ دل ۽ مرڪز گھٽ سيڪڙو ڏيکارين ٿا.


اڳين جملي جي اڳڪٿي ڪرڻ( Next Sentence Prediction (NSP)).


ھي ماڊل برٽ ماڊل جي مدد ڪري ٿو تھ جيئن جملن جي جوڙ کي صحي طريقي سان سمجھي سگھجي. ھن ماڊل زريعي خبر پوي ٿي تھ ڇا پھرين جملي کان پوءِ آيل ٻيو جملو پھرين جملي سان مشاھبت رکي ٿو. جيڪڏھن رکي ٿو تھ برٽ ماڊل NSP ذريعي ٻڌائيندو تھ پويون جملو اڳين جملي سان مشاھبت رکي پر جيڪڏھن پويون جملو اڳين جملي کان مختلف آھي تھ ماڊل ٻڌائيندو  تھ پويون جملو اڳين جملي کان مختلف آھي.  مثال:


اسلم خريداري ڪرڻ ويو. ھن ھڪ قميص ورتي (پويون جمل اڳين جملي سان مشاھبت رکي ٿو تنھنڪري ھي صحي آھي. ھنن ٻنھي جملن جو جوڙ بھ صحي آھي)

مريم ڪافي ٺاھي. آئيسڪريم کپائي.  ( پويون جملو پھرين جملي کان مختلف آھي. اھڙي طرح پھرين جملي سان مشاھبت بھ نٿو رکي تنھنڪري جملن جو جوڙ صحي نھ آھي)


برٽ ماڊل کي سکيا ڏيڻ وقت 50 سيڪڙو صحي جملا ڏنا ويندا آھن ۽ 50 سيڪڙو بغير ترتيب ڏنل جملا ڏنا ويندا آھن. اھڙي طرح ترتيب ڏنل ۽ بغير ترتيب ڏنل جملن کي پاڻ ۾ ملايو ويندو آھي تھ جيئن  برٽ مآڊل بھتر قسم جي اڳڪٿي ڪري سگھي ۽ اڳڪٿي جي درستگي بھ وڌيڪ ھجي.


برٽ ماڊل جي ماپ ۽ آرڪيٽيڪچر يا تعمير( (BERT Model size and architecture


برٽ ماڊل متن جي ماپ يا سائيز مطابق تيار ڪيا وڃن ٿا. ھي اھڙو ماڊل آھن جنھن ۾ ملين جي تعداد ۾ متن يا ڊيٽا ڏئي سگھجي ٿي يا input ڪري سگھجي ٿي. برٽ ماڊل ۾ عمومن ٻن قسم جا آرڪيٽيڪچر موجود آھن. ھڪ بنيادي برٽ ماڊل، جنھن کي  BERT Base چيو وڃي ٿو ۽ ٻيون ڊگهو برٽ ماڊل، جنھن کي BERT Large چيو وڃي ٿو. بنيادي برٽ ماڊل ۾  110M پئراميٽر ۽ ڊگھي برٽ ماڊل ۾ 340M پئراميٽر تائين ڊيٽا وجھي سگھجي ٿي.


ھي بلاگ مشين لرننگ جي ماڊل برٽ جو  ھڪ ننڍڙو تعارف آھي، جيڪو اوھان جي اڳيان پيش ڪيو آھي. برٽ ماڊل جا مختلف قسم ٿين ٿا۽ انھن جو استعمال بھ مختلف طريقن سان ڪري سگھجي ٿو. اوھان جيڪڏھن نيچرل لئنگوئيجز پروسيس جي شعبي ۾ ڪم ڪرڻ چاھيو ٿا تھ مشين لرننگ سان گڏ برٽ ماڊل کي بھ سمجھو تھ جيئن بھتر ڪم ڪري سگھو.

ڪوشش ڪندس تھ برٽ ماڊل جي مختلف قسمن ۽ انھن جي استعمال تي بلاگ سنڌي ٻولي ۾ لکان ۽ اوھان جي آڏو  پيش ڪريان تھ جيئن پنھنجي ٻولي ۾  ڪمپيوٽر سائينس ۽ ھٿراڌو ڏاھپ جي علمن کي پڙھڻ جو لطف وٺجي. پنھنجو گھڻو خيال رکندا.


Tuesday, 19 May 2020

يونيورسٽين ۾ ٿيندڙ تحقيق جو لاڀ ڇو نه ٿو ملي؟ -

يونيورسٽين ۾ ٿيندڙ تحقيق جو لاڀ ڇو نه ٿو ملي؟ -: جيئن جيئن عالمي سماج جي ترقي تيز ٿيندي وڃي پئي تيئن تيئن تعليمي ادارن جي اهميت ۽ گهرج به وڌندي وڃي پئي. اهڙي طرح تعليمي ادارن جي ساخت ۽ تعليمي معيار به تبديل ٿيندا رهن ٿا. خاص ڪري يونيورسٽين جي … وڌيڪ لاء

Monday, 27 April 2020

ايپل آئي فون-13 ۾ سنڌي ٻولي جو استعمال ۽ ميڊيا انتشار

ڊاڪٽر مظهر علي ڏوتيو
ايپل آئي فون پنهنجي نئين ماڊل آئي او ايس 13 ( iOS 13 ) کي سيپٽيمبر 2019ع ۾ ابتدائي طور پيش ڪيو جنهن کي بعد ۾ 25 اپريل 2020ع تي اپگريڊ ڪري 13.4.2 بيٽا (13.4.5 Beta 2 (17F5044d) ) ورزن جي صورت ۾ دنيا آڏو پيش ڪيو ويو آهي. هي ورزن آئي فون جو زبردست ورزن آهي جنهن ۾ ڪيترن ئي نئين خصوصيتين ۽ صحولتن سان گڏ دنيا جون ڪيتريون ئي ٻوليون به داخل ڪيون ويون آهن جن ۾ هندستان جون 22 ٻوليون به شامل آهن. سنڌي ٻولي به هندستان جي انهن 22 ٻولين ۾ شامل آهي جنهن جون ٻئي لپيون يعني عريبڪ ۽ ديوناگري لپيون ان ماڊل ۾ شامل ڪيون ويون آهن. جڏهن به ڪي سوفٽويئر ٺهندا آهن ته اهي ٺهڻ سان ئي مڪمل نه هوندا آهن پر انهن ۾ ڪيترين ئي غلطين جي گنجائش هوندي آهي. اهو ئي سبب آهي جو ڪمپنيون انهن سوفٽويئر کي مختلف ورزن ۾ پيش ڪنديون آهن. تنهنڪري ڪنهن بگ (Bug) جو اچڻ ڪا نئين ڳاله يا حيرت ۾ وجھندڙ نه آهي. اهو ئي سبب آهي جو آئي فون جي نئين ماڊل ۾ سنڌي ٻولي جي ڪجھ لفظن (Characters string) ۽  اٽلين فليگ اموجي ( Italian Flag Emoji) جي ڪري بگ (Bug) پيدا ٿيو جنهن آئي فون کي ٿوري وقت لاءِ ڪريش (Crash) ڪيو. ايپل آئي فون ۾ هي ڪريش ڪو پهريون دفعو ڪون آيو آهي پر اهو اڳ ۾ به ڪيترا دفعا آيو آهي. ٻه سال اڳ آئي فون آئي او ايس-11 ۾ ٽيلگو ٻولي جي ڪري بگ پيدا ٿيو هو جنهن آئي فون کي ڪريش ڪيو هو پر نئين ورزن سان اهو بگ ختم ڪيو ويو هو. اهو اڪثر ٻڌڻ ۾ ايندو آهي ته ڪنهن ڪريڪٽر اسٽرنگ (Characters string) جي ڪري بگ (bug) پيدا ٿيو جنهن آئي فون کي ڪريش ڪري ڇڏيو جيڪا ڪا پريشاني جي ڳاله نه هوندي آهي، نه ئي وري ڪو اهڙي قسم جو بگ (bug) ڪو ڊيٽا کي نقصان ڏيندو آهي. تنهنڪري سنڌي اکرن جي اسٽرنگ (String) سان جيڪڏهن ڪو بگ (bug) پيدا ٿيو ۽ ان سان وقتي طور آئي فون هينگ (Hang) ٿي يا رڪجي وئي ته ڪا وڏي ڳاله نه آهي جنهن کي ايڏي وڏي حيرانگي سان پيش ڪيو ويو آهي. ڪمپنين جي پاڻ پروفيشنل جنگ ۽ ميڊيا کي خبر نشر ڪرڻ واري عمل ان ننڍي خبر کي تمام وڏو ڪري پيش ڪيو ۽ سنڌي ٻولي کي پنهنجي نشاني تي رکيو. اسان وٽ اردو ميڊيا کي ته شل نه ڪا سنڌ ۽ سنڌي ماڻهن يا سنڌي ٻولي جي حوالي سان ڪا خبر ملي. ان خبر ۾ مرچ مصاله ملائي اهڙو پيش ڪن ٿا جو ماڻهو پريشان ٿي وڃن ٿا. هينئر به اردو ميڊيا ان خبر کي اهڙي طرح پيش ڪيو جو ڄڻ سنڌي ٻولي ايپل آئي فون کي ڪريش ڪري مارڪيٽ مان ختم ڪري ڇڏيو هجي يا ان عمل جي ڪري ڪو ايپل آئي فون جو ويليو گهٽجي ويو هجي. جڏهن ته اٽلين اموجي جي ڪري پڻ ان ورزن ۾ بگ پڻ پيدا ٿيو جنهن جي هتان جي ميڊيا ڪا خبر ڪون ڏني.
سنڌي ماڻهن کي ته خوش ٿيڻ گهرجي ته ايپل پنهنجي آئي فون، ٽيب وغيره ۾ سنڌي ٻولي جي ٻنهي لپين کي شامل ڪيو آهي. هي بگ نئين ورزن ۾ ختم ٿي ويندو ۽ سنڌي ٻولي جون ٻئي لپيون ڪاميابي سان ايپل آئي فون تي ڪم ڪنديون جيڪو سنڌي ٻولي لاءِ اهم قدم هوندو. ايپل آئي فون تي سنڌي ٻولي نه صرف لکت ۾ استعمال ٿئي پئي پر آواز جي صورت ۾ پڻ آئي فون تي ڪم ڪري ٿي ۽ اڳتي هلي بهتر طريقي سان ڪم ڪندي.

Present and Future of Artificial Intelligence

 Mazhar Ali Dootio A rtificial Intelligence (AI) is an emerging field of technology that is rapidly transforming the way we live, work and i...