Another week, another step forward! Our latest dev updates focus on laying the groundwork for what’s next. Let's break it down.
2/
ECC Transpose Adaptation: Since ECC doesn’t yet support tensor transpose, implemented application-side unified transpose in zkmlface, ensuring functional equivalence across all inputs/outputs.
3/
ECC Framework Review: Identified key issues including strong binding between computation graph and witness (forcing re-setup on each witness change) and reshape bugs for non-power-of-two dimensions.
4/
Gemma3 Quantization: Introduced front-end padding for variable-length inputs to enable flexible inference.
Next Steps (Circuit Optimization): Plan to prune gemma3 nodes during circuitization to reduce redundancy and improve proof efficiency.
Stay tuned for more updates.
925
0
本頁面內容由第三方提供。除非另有說明，OKX 不是所引用文章的作者，也不對此類材料主張任何版權。該內容僅供參考，並不代表 OKX 觀點，不作為任何形式的認可，也不應被視為投資建議或購買或出售數字資產的招攬。在使用生成式人工智能提供摘要或其他信息的情況下，此類人工智能生成的內容可能不準確或不一致。請閱讀鏈接文章，瞭解更多詳情和信息。OKX 不對第三方網站上的內容負責。包含穩定幣、NFTs 等在內的數字資產涉及較高程度的風險，其價值可能會產生較大波動。請根據自身財務狀況，仔細考慮交易或持有數字資產是否適合您。