2x RTX 5060 Ti

6 records across 3 models (quants: Q4_K_M, Q5_K_M).

hardwaremodelquantbackendtok/spp tok/sWttft sctx/tgdateclassflagssource
2x RTX 5060 TiLlama 3.1 8B Q5_K_Mllama.cpp (CUDA)50–––81922026-03-31community–source
2x RTX 5060 TiLlama 3.1 8B Q5_K_Mllama.cpp (CUDA)8.4–––81922026-03-31community–source
2x RTX 5060 TiQwen 2.5 14B Q5_K_Mllama.cpp (CUDA)28–––81922026-03-31community–source
2x RTX 5060 TiQwen 2.5 14B Q5_K_Mllama.cpp (CUDA)5.3–––81922026-03-31community–source
2x RTX 5060 TiQwen 2.5 32B Q4_K_Mllama.cpp (CUDA)14.3–––81922026-03-31community–source
2x RTX 5060 TiQwen 2.5 32B Q4_K_Mllama.cpp (CUDA)5.5–––81922026-03-31community–source

All rows cite their source. Provenance: community 40018 sourced 1085

← all hardware