Frequently Asked Questions
Frequently Asked Questions
Frequently Asked Questions
General
What is Giza?
Giza is a verifiable machine learning platform and protocol leveraging cryptographic primitives to bridge AI and web3. Giza allows the conversion of ML models into verifiable programs, as well as decentralized hosting, proof generation, model ownership and monetization.
What is Giza?
Giza is a verifiable machine learning platform and protocol leveraging cryptographic primitives to bridge AI and web3. Giza allows the conversion of ML models into verifiable programs, as well as decentralized hosting, proof generation, model ownership and monetization.
What is Giza?
Giza is a verifiable machine learning platform and protocol leveraging cryptographic primitives to bridge AI and web3. Giza allows the conversion of ML models into verifiable programs, as well as decentralized hosting, proof generation, model ownership and monetization.
How does Giza enable scaling for ML models?
Giza utilizes a widely adopted ML interoperability framework called ONNX (Open Neural Network Exchange) as a gateway to converting models into various provable languages. Verifiable models which are served on Giza can thus be executed off-chain in computationally cheap environments and be verified on-chain, enabling ML inferences to be utilised trustlessly by smart contracts.
How does Giza enable scaling for ML models?
Giza utilizes a widely adopted ML interoperability framework called ONNX (Open Neural Network Exchange) as a gateway to converting models into various provable languages. Verifiable models which are served on Giza can thus be executed off-chain in computationally cheap environments and be verified on-chain, enabling ML inferences to be utilised trustlessly by smart contracts.
How does Giza enable scaling for ML models?
Giza utilizes a widely adopted ML interoperability framework called ONNX (Open Neural Network Exchange) as a gateway to converting models into various provable languages. Verifiable models which are served on Giza can thus be executed off-chain in computationally cheap environments and be verified on-chain, enabling ML inferences to be utilised trustlessly by smart contracts.
Which ML frameworks are supported by Giza?
Leveraging ONNX, Giza supports a wide range of ML frameworks such as TensorFlow, PyTorch, Scikit-Learn, and many more.
Which ML frameworks are supported by Giza?
Leveraging ONNX, Giza supports a wide range of ML frameworks such as TensorFlow, PyTorch, Scikit-Learn, and many more.
Which ML frameworks are supported by Giza?
Leveraging ONNX, Giza supports a wide range of ML frameworks such as TensorFlow, PyTorch, Scikit-Learn, and many more.
What are the benefits of building with Giza?
Multi-layer Interoperability: Giza allows deploying ML models developed in various frameworks (thanks to the usage of ONNX) into highly performant provable languages based on the desired use-case (performance, privacy, etc). Abstraction: Giza abstracts away all complexities associated with cryptographic verifiability from its users, allowing ML developers to utilise the frameworks of their choice with ease. Scaling: The platform handles the underlying infrastructure and architecture, ensuring high availability, fault tolerance, and performance with zero downtime. Transparency: Monitoring and management of deployed models are simplified, as all the data and traces can be stored on the blockchain.
What are the benefits of building with Giza?
Multi-layer Interoperability: Giza allows deploying ML models developed in various frameworks (thanks to the usage of ONNX) into highly performant provable languages based on the desired use-case (performance, privacy, etc). Abstraction: Giza abstracts away all complexities associated with cryptographic verifiability from its users, allowing ML developers to utilise the frameworks of their choice with ease. Scaling: The platform handles the underlying infrastructure and architecture, ensuring high availability, fault tolerance, and performance with zero downtime. Transparency: Monitoring and management of deployed models are simplified, as all the data and traces can be stored on the blockchain.
What are the benefits of building with Giza?
Multi-layer Interoperability: Giza allows deploying ML models developed in various frameworks (thanks to the usage of ONNX) into highly performant provable languages based on the desired use-case (performance, privacy, etc). Abstraction: Giza abstracts away all complexities associated with cryptographic verifiability from its users, allowing ML developers to utilise the frameworks of their choice with ease. Scaling: The platform handles the underlying infrastructure and architecture, ensuring high availability, fault tolerance, and performance with zero downtime. Transparency: Monitoring and management of deployed models are simplified, as all the data and traces can be stored on the blockchain.
Can ML models on Giza be used in Ethereum L1?
Yes, ML models can run off-chain and the results of inferences can be sent to Ethereum L1 through the verification of the associated ZK proof of the inference. The only requirement is the deployment of a verifier smart contract on the desired L1. Current architecture allows for full interoperability with Ethereum using a STARK verifier written in Solidity. In the future, the verifier will be deployed into any EVM chain to provide ML inferences.
Can ML models on Giza be used in Ethereum L1?
Yes, ML models can run off-chain and the results of inferences can be sent to Ethereum L1 through the verification of the associated ZK proof of the inference. The only requirement is the deployment of a verifier smart contract on the desired L1. Current architecture allows for full interoperability with Ethereum using a STARK verifier written in Solidity. In the future, the verifier will be deployed into any EVM chain to provide ML inferences.
Can ML models on Giza be used in Ethereum L1?
Yes, ML models can run off-chain and the results of inferences can be sent to Ethereum L1 through the verification of the associated ZK proof of the inference. The only requirement is the deployment of a verifier smart contract on the desired L1. Current architecture allows for full interoperability with Ethereum using a STARK verifier written in Solidity. In the future, the verifier will be deployed into any EVM chain to provide ML inferences.
What is Orion?
Orion is a framework for Validity ML, which focuses on the verification of ML inference sources and traces. It provides a transparent, verifiable, and open-source framework for inference, initially using STARKs.
What is Orion?
Orion is a framework for Validity ML, which focuses on the verification of ML inference sources and traces. It provides a transparent, verifiable, and open-source framework for inference, initially using STARKs.
What is Orion?
Orion is a framework for Validity ML, which focuses on the verification of ML inference sources and traces. It provides a transparent, verifiable, and open-source framework for inference, initially using STARKs.
How does Orion enhance ML inference?
By supporting multiple backends for ZK inference, Orion abstracts cryptographic primitives from the user with a unified API, granting seamless interoperability. In its first version, Orion offers a new ONNX runtime built in Cairo 1.0, enabling verifiable ML model inferences. It provides APIs for operators, numbers, and performance optimization, ensuring reliable and efficient computation. In the future Orion framework will extend to other cryptographic frameworks which can add value and utility to verifiable ML.
How does Orion enhance ML inference?
By supporting multiple backends for ZK inference, Orion abstracts cryptographic primitives from the user with a unified API, granting seamless interoperability. In its first version, Orion offers a new ONNX runtime built in Cairo 1.0, enabling verifiable ML model inferences. It provides APIs for operators, numbers, and performance optimization, ensuring reliable and efficient computation. In the future Orion framework will extend to other cryptographic frameworks which can add value and utility to verifiable ML.
How does Orion enhance ML inference?
By supporting multiple backends for ZK inference, Orion abstracts cryptographic primitives from the user with a unified API, granting seamless interoperability. In its first version, Orion offers a new ONNX runtime built in Cairo 1.0, enabling verifiable ML model inferences. It provides APIs for operators, numbers, and performance optimization, ensuring reliable and efficient computation. In the future Orion framework will extend to other cryptographic frameworks which can add value and utility to verifiable ML.
Is Orion an open-source project?
Yes, Orion is fully open-source, reflecting a commitment to transparency, collaboration, and shared ownership. The project welcomes contributions from the community and offers support to all contributors.
Is Orion an open-source project?
Yes, Orion is fully open-source, reflecting a commitment to transparency, collaboration, and shared ownership. The project welcomes contributions from the community and offers support to all contributors.
Is Orion an open-source project?
Yes, Orion is fully open-source, reflecting a commitment to transparency, collaboration, and shared ownership. The project welcomes contributions from the community and offers support to all contributors.
For Protocols
Start integrating AI Actions without compromising your protocol security and standards.
For Protocols
Start integrating AI Actions without compromising your protocol security and standards.
For Protocols
Start integrating AI Actions without compromising your protocol security and standards.