Harshit Yadav
Member since 2025
Member since 2025
In this Google DeepMind course you will focus on the training process for machine learning models. You will learn how to spot and mitigate issues when training a model, such as overfitting and underfitting. In practical coding labs, you will implement and evaluate the multilayer perceptron for simple classification tasks. This will provide insights into the mechanics of training a neural network model and the backpropagation algorithm. Research case studies will demonstrate how neural networks power real-world models. Additionally, you will consider the broader social impacts of innovation by looking beyond immediate benefits to anticipate potential risks, safety concerns, and further-reaching societal consequences.
In this Google DeepMind course, you will complete a capstone project that brings together the technical knowledge, ethical awareness, and creative problem-solving skills that you have developed throughout the Google DeepMind: AI Research Foundations curriculum. You will apply what you have learned to a problem of your choosing. You will first plan your project to ensure that you are clear on your project aims and intended impact. You will then collect and prepare data for your model. Finally, you will implement a full fine-tuning workflow using Gemma 3 with low-rank adaptation (LoRA) and evaluate its performance. You will receive guidance, but the emphasis will be on adaptation and experimentation - skills that are essential for applied AI research.
Train more powerful models with a single GPU. In this course, you will learn how hardware can speed up model training and the key considerations when training models on a GPU. First, you will learn how to estimate the number of computations and the amount of computer memory required to train large neural networks. You will then discover techniques for reducing the computing and memory requirements when training a model. Techniques which you will apply for fine-tuning a Gemma model with 4 billion parameters. Finally, you will consider the potential environmental impacts of machine learning, with a focus on where questions of energy, water, and e-waste intersect with justice and equity.
Unleash the power of language models with fine-tuning. In this course, you will learn how to adjust a pre-trained model to a specific task. You will start with full-parameter fine-tuning using a small language model. To tune larger models like Gemma, you will learn parameter-efficient techniques with a focus on LoRA. Finally, you will be briefly introduced to reinforcement learning as an alternative to supervised fine-tuning (SFT). You will also explore how AI is imagined and made sense of in cultural contexts. You will consider why responsible AI is not just about technical safety but also about building governance systems that reflect community values and protect the public interest.
In this Google DeepMind course you will discover the mechanisms of the transformer architecture. You will investigate how transformer language models process prompts to make context-sensitive next-token predictions. Through practical activities you will explore the attention mechanism, visualize attention weights, and encounter advanced concepts like masked attention and multi-head attention. You will also learn other techniques that are necessary to build neural networks that are well-suited to be used as language models. Finally, through activities on values, stakeholder mapping and community engagement, you will practice concrete tools for ensuring AI projects are developed with communities, not just for them.
In this Google DeepMind course you will learn how to prepare text data for language models to process. You will investigate the tools and techniques used to prepare, structure, and represent text data for language models, with a focus on tokenization and embeddings. You will be encouraged to think critically about the decisions behind data preparation, and what biases within the data may be introduced into models. You will analyze trade-offs, learn how to work with vectors and matrices, how meaning is represented in language models. Finally, you will practice designing a dataset ethically using the Data Cards process, ensuring transparency, accountability, and respect for community values in AI development.
In this Google DeepMind course, you will learn the fundamentals of language models and gain a high-level understanding of the machine learning development pipeline. You will consider the strengths and limitations of traditional n-gram models and advanced transformer models. Practical coding labs will enable you to develop insights into how machine learning models work and how they can be used to generate text and identify patterns in language. Through real-world case studies, you will build an understanding around how research engineers operate. Drawing on these insights you will identify problems that you wish to tackle in your own community and consider how to leverage the power of machine learning responsibly to address these problems within a global and local context.
本課程旨在說明負責任 AI 技術的概念和 AI 開發原則,同時介紹各項技術,在實務上找出公平性和偏誤,減少 AI/機器學習做法上的偏誤。我們也將探討實用方法和工具,透過 Google Cloud 產品和開放原始碼工具,導入負責任 AI 技術的最佳做法。
本課程針對評估生成式和預測式 AI 模型,向機器學習從業人員介紹相關的基礎工具、技術和最佳做法。模型評估是機器學習的重要領域,確保這類系統能在正式環境中提供可靠、準確且成效優異的結果。 學員將深入瞭解多種評估指標與方法,以及適用於不同模型類型和工作的應用方式。此外,也會特別介紹生成式 AI 模型帶來的獨特難題,並提供有效的應對策略。透過 Google Cloud Vertex AI 平台,學員將瞭解在模型挑選、最佳化和持續監控方面,該如何導入穩健的評估程序。
本課程旨在提供必要的知識和工具,協助您探索機器學習運作團隊在部署及管理生成式 AI 模型時面臨的獨特挑戰,並瞭解 Vertex AI 如何幫 AI 團隊簡化機器學習運作程序,打造成效非凡的生成式 AI 專案。