community pool / model answer

Ornstein3.6-35B-A3B-GGUF

chat · DJLougen/Ornstein3.6-35B-A3B-GGUF

100.2 tok/s · model median · 2 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Radeon RX 7900 XTX 24GB4-bit62.4 tok/sreported-- GB640llama.cpp
RTX 3090 24GB ×24-bit138 tok/sreported-- GB8,192llama.cpp

Context tested

The table above is the tested context for each exact rig and quantization cell. Empty fields remain empty rather than being filled with an estimate.