<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Sai Nath on Medium]]></title>
        <description><![CDATA[Stories by Sai Nath on Medium]]></description>
        <link>https://medium.com/@zszazi?source=rss-13a47fbd9d4c------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*nHN0-O3jUcbV75qz.</url>
            <title>Stories by Sai Nath on Medium</title>
            <link>https://medium.com/@zszazi?source=rss-13a47fbd9d4c------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 09:25:00 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@zszazi/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Vox Helios — Audio Question Answering System]]></title>
            <link>https://medium.com/@zszazi/vox-helios-audio-question-answering-system-4d17a2c48f22?source=rss-13a47fbd9d4c------2</link>
            <guid isPermaLink="false">https://medium.com/p/4d17a2c48f22</guid>
            <category><![CDATA[vqa]]></category>
            <dc:creator><![CDATA[Sai Nath]]></dc:creator>
            <pubDate>Sat, 09 Dec 2023 02:03:17 GMT</pubDate>
            <atom:updated>2023-12-10T06:41:56.524Z</atom:updated>
            <content:encoded><![CDATA[<h3>Vox Helios — Audio Question Answering System</h3><p>Ever wondered how good it would be to get your questions answered by a genie as you speak ? Enter Vox Helios — one such genie</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*kNzoWBXjJq9Mt2dUsi7v8w.jpeg" /><figcaption>Vox Helios — AI Generated Image</figcaption></figure><p><strong>Introduction</strong></p><p>Vox Helios, coined from “Vox” for audio and “Helios” symbolizing the god of sight, is an Audio Question Answering (AQA) system that we created utilizing existing open-source models. The system comprises two main components: an Audio Transcription module and a Visual Question Answering module.</p><p>Audio Question Answering models represent a type of multimodal model employing deep learning and natural language processing to establish contextual comprehension of inputs. These models utilize deep learning algorithms to transcribe audio signals, interpreting and analyzing them to respond appropriately to posed questions based on the input image. Their applications extend to speech recognition models and understanding semantic relationships. The future trajectory of this field aims to emulate the human ability to listen to questions and provide relevant answers. An illustrative example involves a question about an artist’s creation, where the model might misinterpret the query due to external audio noise. Addressing this challenge involves developing strategies to handle such noise and connecting it to the correct answer. Further advancements in these models include handling diverse languages and accents in audio inputs, as well as extracting meaningful information from the signals.</p><p>The second aspect involves the Visual Question Answering (VQA) task, which focuses on developing models capable of answering questions about visual content, such as images or videos. The objective is to build systems that comprehend both the visual information and the textual context of a question, leveraging the capabilities of computer vision and natural language processing. In a typical VQA scenario, the model receives an image and a corresponding question in natural language, requiring it to analyze the visual content and understand the question’s semantics to generate an accurate textual response. Through Vox Helios, we propose a system that integrates both tasks, creating an end-to-end pipeline for an Audio Question Answering platform.</p><p><strong>Motivation</strong></p><p>The motivation behind the Vox Helios project lies in the desire to create a comprehensive audio question-answering model. By enhancing a multi-modal VQA system, the team envisions making technology more accessible to users with different abilities. The fusion of text, audio, and images aims to provide users with a seamless and intuitive way to interact with machines. Leveraging deep learning concepts, the project integrates audio into the VQA model, allowing users to generate questions verbally along with relevant images.</p><p><strong>Methodology</strong></p><p>The proposed methodology harnesses cutting-edge technologies to tackle challenges in audio transcription and image-based question-answering (VQA). OpenAI Whisper, known for its exceptional prowess in converting audio to text, is employed to transcribe diverse audio inputs. Concurrently, Salesforce BLIP, a robust VQA model, is leveraged to address queries related to images. In order to enhance the precision of the VQA model, a fine-tuning process is implemented using noisy data generated through the language model (LLM). This unique approach allows the model to adapt to real-world scenarios where input data may lack pristine quality. Additionally, Whisper is utilized to programmatically process noisy audio data generated by the language model. The entire workflow is seamlessly orchestrated: Whisper transcribes audio to text, the VQA model undergoes fine-tuning with noisy data, and the integrated system is encapsulated in a Gradio app, ensuring user-friendly access to advanced capabilities in audio transcription and image-based question-answering. Hugging Face transformers were employed to construct a pipeline that encompasses Audio → Text → VQA → Result.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wk4gKv0UVZRljV-UoeDMKg.png" /><figcaption>Fig : Vox Helios Pipeline diagram</figcaption></figure><p><strong>Dataset</strong></p><p>Creating data for Audio Question Answering (AQA) presented a challenge due to the absence of readily available reference datasets. To address this, we divided the creation process into two segments: one for audio samples and another for Visual Question Answering (VQA) samples.</p><p><em>Audio Dataset:</em><br>We manually generated our audio dataset by recording 100 custom samples using microphones:<br>- 70 samples without noise<br>- 30 samples with programmatically inserted noise at various intervals.</p><p><em>Concept of Noise in Audio Samples</em><br>Noise, in this context, involves intentionally introducing distortions or interference into audio samples. This deliberate strategy enhances the model’s training by exposing it to realistic scenarios where audio is not perfectly clean. Various types of noise were added, such as varying voices, modulating pitch, and ensuring unclear and concise voices.</p><p><em>VQA Dataset:</em><br>For fine-tuning BLIP-VQA, we selected images of popular footballers for the training set:<br>- 10 images (e.g., Messi, Mbappe)<br>- 10 questions with added noise for each image<br>- 1 correct answer for each question.</p><p><em>Noise in Textual Data:</em><br>In textual data, noise refers to unintended variations, errors, or inconsistencies. Types of noise include grammatical errors, spelling mistakes, phonetic mismatches, semantic variations, and intentional variations.</p><p><strong><em>Examples:</em></strong><br>1. <em>Grammatical Errors</em><br> — Incorrect: “Wher from Harry Cain?”<br> — Correct: “Where is Harry Kane from?”</p><p>2. Spelling Mistakes<br> — Incorrect: “Which klubb does Harry Kane play forr?”<br> — Correct: “Which club does Harry Kane play for?”</p><p>These examples illustrate the different types of noise in textual data. Understanding and handling such noise is crucial in natural language processing tasks to account for language variability. Intentional noise was introduced in both audio and textual datasets to enhance model robustness.</p><p><strong>Fine Tuning</strong></p><p>We utilized Hugging Face transformers for fine-tuning on our previously described custom VQA dataset. The fine-tuning process was executed within the Hugging Face spaces, and the specified parameters for fine-tuning are outlined below:</p><p>1. Learning Rate: 5e-05<br>2. Train Batch Size: 4<br>3. Optimizer with betas and epsilon: (0.9, 0.999) and 1e-08<br>4. Number of Epochs: 20</p><p>The decision to opt for a low train batch size was driven by the limited number of datapoints available for fine-tuning. The fine-tuning metrics, including a tensorboard for streamlined inference, are presented. We identified a model with a smoothed training loss of 1.4023 as the optimal choice. The entire fine-tuning process was completed in approximately 16 minutes.</p><p><strong>Results</strong></p><p>Our findings revealed that by employing a fine-tuning approach, we successfully addressed issues related to posing noisy questions. The model demonstrated the ability to provide accurate responses even in the presence of spelling errors, a characteristic anticipated in real-world scenarios.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*jUuYnnW9CNStzljI3A_xgw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*nyFYbkVTjNKK_b5_Bek4wQ.png" /></figure><p>We presented our project as a web application, which was built using Gradio framework.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*0gCEAbZBm2Sp6P62_859lw.png" /><figcaption>Gradio Web app for Vox Helios</figcaption></figure><p><strong>Conclusion</strong></p><p>Our goal was to lay the groundwork for advancing Audio Question Answering using open-source models. The successful performance of our fine-tuned model in real-world scenarios, particularly with noisy input, was gratifying. Looking ahead, we aim to develop a framework for generating noisy data to train deep learning models across various modal inputs like text, audio, and images.</p><p>The project suggests potential avenues for future improvements. First, implementing a feedback loop in the system could enhance adaptability by collecting user feedback on audio transcription and image-based question-answering. This iterative fine-tuning process would create a dynamic system responsive to user needs.</p><p>Second, developing a dedicated mobile app would enhance user experience by providing convenient access to audio transcription and VQA capabilities on handheld devices. A user-friendly interface and real-time processing would improve overall accessibility.</p><p>Extending the system to support multilingual audio question-answering would broaden its global appeal. Integrating advanced language models for queries in multiple languages would make the system versatile and inclusive, meeting diverse linguistic needs.</p><p>We express our gratitude to Dr. Abhijit Mishra — Course Advisor for providing guidance and support throughout the project.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4d17a2c48f22" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fastai Training metrics as Notification on mobile]]></title>
            <link>https://medium.com/tensorpad/fastai-training-metrics-as-notification-on-mobile-1b2d419535e7?source=rss-13a47fbd9d4c------2</link>
            <guid isPermaLink="false">https://medium.com/p/1b2d419535e7</guid>
            <category><![CDATA[transfer-learning]]></category>
            <category><![CDATA[fastai]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[training-notifications]]></category>
            <dc:creator><![CDATA[Sai Nath]]></dc:creator>
            <pubDate>Mon, 05 Aug 2019 11:56:06 GMT</pubDate>
            <atom:updated>2019-09-17T18:49:32.271Z</atom:updated>
            <content:encoded><![CDATA[<h3>FastAI Training metrics as Notification on mobile</h3><h4>Get the model updates on mobile as notifications — this would work on any of the Cloud providers or even Google Colab</h4><p>Ever had to train a machine learning model for so long that you had to sit and stare at the screen every time an epoch got over to have a look at the model metrics?</p><p>Forget it! using the combined power of Fastai callbacks and pushover.net you can get training notifications right on your mobile phone</p><p>This is possible because of the rich callbacks system in FastAI library</p><p>Lets split our tutorial into three sections</p><ul><li>pushover.net configuration on mobile</li><li>Pneumonia detection model training with Fastai</li></ul><p>Github repo link → <a href="https://github.com/zszazi/pneumonia-detection-fastai">pneumonia-detection-fastai</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/354/1*0VsGHw2goOFpeis6bqZJhA.png" /><figcaption><strong>Training notifs on mobile</strong></figcaption></figure><h4>STEP 1 -&gt; Setting up a workspace on your favourite cloud provider</h4><p>Open the terminal from Jupyter lab and clone the repo git clone https://github.com/zszazi/pneumonia-detection-fastai.git to your workspace</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PF9iXCpzzqBfFq5fxwZqiQ.png" /></figure><ul><li>now open the Jupyter notebook by name pneumonia_prediction_using_fastau_with_mobile_notification.ipynb</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iYKdmB_Ciy_DbDcc7olq7w.png" /></figure><h4>STEP 2 -&gt; Pushover.net configuration on mobile</h4><p>Head over to <a href="https://pushover.net/">https://pushover.net</a> and set up an account and download the mobile app too, You will only be able to receive notifications on mobile if you have their app installed. They have extensive <a href="https://pushover.net/api">docs</a> and <a href="https://pushover.net/faq">faq</a> which cover on setting up, to receive notifications on your pushover mobile app.</p><blockquote>pushover currently offers a <a href="https://pushover.net/faq#overview-fees">7 day trial</a> on signup accross all platforms</blockquote><p>Copy the API keys; there are two of these found in your account section</p><pre>&quot;user&quot;  : &quot;your key&quot;,<br>&quot;token&quot; : &quot;Your key&quot;,</pre><h4>STEP 3 -&gt; Training the model and getting notifications on mobile</h4><p>To run the cell press <em>Shift+Enter</em></p><ul><li>Make sure to upload the <em>kaggle.json</em> (watch this <a href="https://www.youtube.com/watch?v=eEgZtNOCJhk">Youtube video</a>)</li><li>We can use the upload button in Jupyter workspace to upload the kaggle.jsonfile</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/510/1*xtiTBx1fKlGxiql9BjDy6g.png" /></figure><ul><li>On running the cell containing learn.fit_one_cycle() you should have got the push notifications on your mobile, and this would be sent for each epoch</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/928/1*KjOdvgBoDiMi9yHphSYo3A.png" /></figure><h4>🔔 Training Notifications on mobile 🔔</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*w375vvWj50ZL2-AK-KLq6w.png" /></figure><p><strong>What’s more?</strong> You can even send your <strong>loss plots</strong> as 🔔 to mobile.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*SMk0mS63xNRrRYNoVpbjzQ.jpeg" /><figcaption><strong>LossPlot as notification!</strong></figcaption></figure><p>Here’s a Github gist showing you the central part of mobile callback notification</p><p><em>code</em></p><p>GitHub gist</p><p>FastAI provides you with a rich collection of callbacks (<em>see pic</em>), but we mainly have used the following ones: whose name are self-explanatory :)</p><ul><li>on_train_begin</li><li>on_epoch_begin</li><li>on_train_end</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/672/1*aVBO-anWoBYA0DFW4zZd2Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*qSUoXOYkyHcyWZztxUdzRA.png" /><figcaption>Screenshot from Fastai <a href="https://docs.fast.ai/callbacks.html">doc</a></figcaption></figure><p><strong>Resources</strong></p><ol><li>FastAI <a href="https://forums.fast.ai/">forums</a></li><li>FastAI <a href="https://docs.fast.ai/callbacks.html">Callbacks</a></li><li>Kaggle<a href="https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia/"> dataset</a></li><li>Pushover <a href="https://pushover.net/api">docs</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1b2d419535e7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/tensorpad/fastai-training-metrics-as-notification-on-mobile-1b2d419535e7">Fastai Training metrics as Notification on mobile</a> was originally published in <a href="https://medium.com/tensorpad">Tensorpad</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pythia (Facebook)— Greek god doing Deep learning]]></title>
            <link>https://medium.com/tensorpad/pythia-facebook-greek-god-doing-deep-learning-ff1d2a76c6d9?source=rss-13a47fbd9d4c------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff1d2a76c6d9</guid>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[facebook]]></category>
            <dc:creator><![CDATA[Sai Nath]]></dc:creator>
            <pubDate>Tue, 11 Jun 2019 08:19:53 GMT</pubDate>
            <atom:updated>2019-09-17T18:50:22.670Z</atom:updated>
            <content:encoded><![CDATA[<h3>Pythia (Facebook) — Greek god doing Deep learning</h3><h4>“Artificial Intelligence” in 2019 has been exciting. Can it be more exciting than this? Guess what I found an answer for it and the answer is from Facebook Research</h4><p>Recently Facebook open sourced their new framework called <strong>Pythia</strong></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgfycat.com%2Fifr%2Fjealousilliterategrassspider&amp;url=https%3A%2F%2Fgfycat.com%2Fjealousilliterategrassspider&amp;image=https%3A%2F%2Fthumbs.gfycat.com%2FJealousIlliterateGrassspider-size_restricted.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=gfycat" width="1920" height="912" frameborder="0" scrolling="no"><a href="https://medium.com/media/8d7cad46b0f83298ffe20320faaa2b1e/href">https://medium.com/media/8d7cad46b0f83298ffe20320faaa2b1e/href</a></iframe><p><strong>ELI5:</strong> Give Pythia an image and shoot questions to it, get your questions answered; sounds interesting isn’t it, but Pythia can do more than this….</p><p>Pythia is the new multimodal research framework for supercharging vision and language tasks built on top of Pytorch</p><p>Is there a reason behind naming it Pythia?</p><blockquote>‘<em>The name ‘</em><a href="https://en.wikipedia.org/wiki/Pythia"><em>Pythia</em></a><em>’ is an homage to the Oracle of Apollo at Delphi, who answered questions in Ancient Greece.’</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*BvN1Crkd4CcsFOakuAYQvA.jpeg" /><figcaption><em>Lycurgus Consulting the Pythia</em></figcaption></figure><p>Ever wondered how great it would be if we had a single framework that could incorporate both NLP and vision tasks easily, this is what Pythia does, it is designed for answering questions related to visual data and automatically generating image captions</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Iz2fIHr5UqyqZ009mFvZFg.png" /><figcaption><a href="https://visualqa.org/">https://visualqa.org</a></figcaption></figure><p>CloudCV has a <a href="https://vqa.cloudcv.org/">Demo of Pythi</a>a up and running which you can try it out</p><p>Pythia incorporates elements of Facebook research’s winning entries in recent AI competitions (the <a href="https://arxiv.org/abs/1807.09956">VQA Challenge</a> 2018 and <a href="http://vizwiz.org/data/#challenge">Vizwiz Challenge 2018</a>) done by the Facebook AI Research (FAIR)’s A-STAR (Agents that See, Talk, Act, and Reason) team</p><p><a href="https://visualqa.org/">Visual Question Answering (VQA)</a> Challenge- Given an image and a natural language question about the image, the task is to provide an accurate natural language answer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*M8cdF8zbQA42VzZHSgfZSw.jpeg" /><figcaption>Pythia got an accuracy of 70.24% on VQA v2.0 dataset</figcaption></figure><p>In addition to multitasking, Pythia also supports distributed training and a variety of datasets, as well as custom losses, metrics, scheduling, and optimizers.</p><p>Now let’s get into <strong>Technical Details</strong></p><ol><li><strong>Bottom-up and Top-down attention</strong></li></ol><p>The critical idea in up-down is the use of an object detector — <a href="https://arxiv.org/abs/1506.01497">Faster RCNN</a> pre-trained on the <a href="https://visualgenome.org/">Visual Genome dataset</a> — to extract image features with bottom-up attention, i.e., visual feed-forward attention. A ResNet-101 was chosen as the backbone network. Visual Genome is a knowledge base to connect structured image concept to language</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eHGTu5K0KRj3h1r0lkGzQg.png" /><figcaption>Visual Genome Dataset</figcaption></figure><p>The question’s text is then used to compute the top-down attention, i.e., task-specific attention, for each object in the image. Multi-modal fusion is done through a simple <a href="https://www.google.co.in/url?sa=t&amp;rct=j&amp;q=&amp;esrc=s&amp;source=web&amp;cd=1&amp;cad=rja&amp;uact=8&amp;ved=2ahUKEwj9vJ6hnNriAhUu63MBHYJnBjUQFjAAegQIBRAB&amp;url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FHadamard_transform&amp;usg=AOvVaw0mGIa2CrO6YzEcIjh22pYy">Hadamard product</a> (also called Schur product) followed by a multi-label classifier using a sigmoid activation function to predict the answer scores. The performance reached 70.34% on VQA 2.0 test-std split with an ensemble of 30 models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/834/1*mpP_nu7Bui7UdTIBDe74Xw.png" /><figcaption><a href="https://www.google.co.in/url?sa=t&amp;rct=j&amp;q=&amp;esrc=s&amp;source=web&amp;cd=1&amp;cad=rja&amp;uact=8&amp;ved=2ahUKEwj9vJ6hnNriAhUu63MBHYJnBjUQFjAAegQIBRAB&amp;url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FHadamard_transform&amp;usg=AOvVaw0mGIa2CrO6YzEcIjh22pYy">Hadamard product</a></figcaption></figure><p>2. <strong>Model Architecture</strong></p><p>A weight normalization followed by ReLU was used instead of gated hyperbolic tangent activation to reduce computations involved. As evident from the application of Pythia, it needed some way of connecting the Text and Image, and this was done with the help of element-wise multiplication to combine the features from text and visual modalities when computing the top-down attention(task-specific) instead of just feature concatenation. How do we represent the Task Question? Use <a href="https://nlp.stanford.edu/pubs/glove.pdf">GloVe</a> (Global Vectors) then pass the word embeddings onto a <a href="https://en.wikipedia.org/wiki/Gated_recurrent_unit">GRU</a> network(Gated Recurrent Unit-a LSTM with a forget gate) and a question attention module to extract attentive text features and discard unnecessary ones.</p><p>3.<strong> Optimizer</strong></p><p>The optimizer used is <a href="http://ruder.io/optimizing-gradient-descent/index.html#adamax">ADAMAX</a> (a version of ADAM). This fantastic post on <a href="http://ruder.io/optimizing-gradient-descent">ruder.io</a> gives an overview of most of the optimizers.</p><h4>Why Pythia matters?</h4><blockquote>Pythia smooths the process of entering the growing subfield of vision and language and frees researchers to focus on faster prototyping and experimentation. Our goal is to accelerate progress by increasing the reproducibility of these models and results. This will make it easier for the community to build on, and benchmark against, successful systems. We hope that removing some of the obstacles will allow researchers to more quickly develop new ways for people and intelligent machines to communicate. This work should also help researchers develop adaptive AI that synthesizes multiple kinds of understanding into a more context-based, multimodal understanding. In addition to this open source release, we plan to continue adding tools, tasks, data sets, and reference models.(source- <a href="https://code.fb.com/ai-research/pythia/">https://code.fb.com/ai-research/pythia/</a>)</blockquote><h3>Features of Pythia</h3><ul><li><strong>Model Zoo: </strong>Reference implementations for state-of-the-art vision and language model</li><li><strong>Multi-Tasking: </strong>Support for multi-tasking which allows training on multiple datasets together</li><li><strong>Datasets: </strong>Includes support for various datasets built-in including VQA, VizWiz, TextVQA, VisualDialog and COCO Captioning.</li><li><strong>Modules: </strong>Provides implementations for many commonly used layers in vision and language domain</li><li><strong>Distributed:</strong> Support for distributed training based on DataParallel as well as DistributedDataParallel.</li><li><strong>Unopinionated:</strong> Unopinionated about the dataset and model implementations built on top of it.</li><li><strong>Customization:</strong> Custom losses, metrics, scheduling, optimizers, TensorBoard; suits all your custom needs.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uFeEK23hL2moKGlIGrfx0Q.jpeg" /><figcaption>Source :Github: <a href="https://github.com/facebookresearch/pythia">Pythia</a></figcaption></figure><p>With Open Source driving the AI community Pythia can be seen as a step getting closer to democratizing AI which earlier was confined to the walls of Research labs and large companies. This easy to use framework can give birth to exciting projects from independent groups and researchers thus accelerating the progress in this field</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*q0LmeeORCp6GCdL6xE2Vgw.png" /><figcaption><a href="https://www.tensorpad.com/?utm_source=medium&amp;utm_campaign=pythia-p1"><strong>Tensorpad</strong></a></figcaption></figure><h4>[Resources]</h4><ol><li>Github: <a href="https://github.com/facebookresearch/pythia">https://github.com/facebookresearch/pythia</a></li><li>FB research: <a href="https://code.fb.com/ai-research/pythia/">https://code.fb.com/ai-research/pythia/</a></li><li>arXiv paper: <a href="https://arxiv.org/abs/1807.09956">https://arxiv.org/abs/1807.09956</a></li><li>Doc: <a href="https://learnpythia.readthedocs.io/en/latest/">https://learnpythia.readthedocs.io/en/latest/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff1d2a76c6d9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/tensorpad/pythia-facebook-greek-god-doing-deep-learning-ff1d2a76c6d9">Pythia (Facebook)— Greek god doing Deep learning</a> was originally published in <a href="https://medium.com/tensorpad">Tensorpad</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>