You need to agree to share your Hugging Face information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Speedrun Deterministic Embedding D=768 (A100 Colab + Circom R1CS + Base CREATE2)

Bit-exact, batch-invariant D = 768 L2 embedding normalization and orthogonal tangent Jacobian engine (K = 48 chunks of C = 16 coordinates at Q = 2^24 fixed-point scaling), engineered for NVIDIA A100-SXM4 GPUs on Google Colab with OpenTelemetry Landauer bit-erasure energy spans and Base Mainnet (chainId = 8453) wQFLOP CREATE2 settlement.

Repository Artifacts

jax_kernels/deterministic_l2_r1cs_d768.py: Multi-chunk (48 x 16 = 768) 128-bit limb-exact JAX @jax.custom_vjp kernel for A100 GPUs.
circuits/deterministic_l2_embedding_d768.circom: Circom 2.1.6 R1CS constraint system with K = 48 chunked accumulators and singleton witness determinism.
daemons/yennefer_a100_colab_otel_daemon.py: Yennefer daemon reading native A100 hardware energy counters (nvmlDeviceGetTotalEnergyConsumption), emitting OpenTelemetry Landauer spans (genesis.yennefer.landauer.a100), signing EIP-712 EnergyTelemetry, and settling wQFLOP on Base.
script/DeployWQFLOPAttestorD768.s.sol: Foundry deterministic CREATE2 deployment script for Base Mainnet (Chain ID 8453).

Code Ownership

Maintained by Principal Investigator Igor Holt (igor@kovachenterprises.com).

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support