Datasetpublic

witcheer/windows-rtx-4060ti-8gb-moe-offload-bench-2026-05

RTX 4060 Ti 8GB — Multi-Model Benchmark (2026-05) practitioner benchmarks on consumer hardware (8GB VRAM, 32GB RAM). 10 models tested, covering MoE expert offload, hybrid SSM architectures, dense models, MLA, dense partial GPU offload, and the 1B speed ceiling. all runs on the same physical rig, same methodology. current leaderboard (decode tok/s at sweet spot) model active params GGUF size sweet spot tok/s quality (6 tests) architecture Llama 3.2 1B… See the full description on the dataset page: https://huggingface.co/datasets/witcheer/windows-rtx-4060ti-8gb-moe-offload-bench-2026-05.

sourceHugging Facecc-by-4.0updated 4mo agoView on Hugging Face

3likes57downloads

30 commits on main

6bbcf9b4mo ago

append 5 Gemma 4-E4B rows to combined JSONL (62 total rows)

witcheer

a9731754mo ago

add Gemma 4-E4B benchmark data (5 rows, 2026-05-21)

witcheer

4c0695f4mo ago

add gpt-oss-20b source file to source/

witcheer

df685d44mo ago

fix: remove stray root JSONL that broke dataset viewer

witcheer

7be629a4mo ago

fix: merge gpt-oss-20b rows into main file, remove stray root JSONL

witcheer

b9297fc4mo ago

Upload README.md with huggingface_hub

witcheer

bd9d1724mo ago

Upload bench-gpt-oss-20b-2026-05-20.jsonl with huggingface_hub

witcheer

89762334mo ago

fix data files table: source/ not data/

witcheer

be8be6e4mo ago

remove file from data/ (conflicts with dataset viewer, moved to source/)

witcheer

35f42144mo ago

move Qwen3.6 27B source file to source/ directory

witcheer

87794d84mo ago

merge Qwen3.6 27B dense rows into main dataset (47 -> 52 rows)

witcheer

c579ca54mo ago

update README: add Qwen3.6 27B dense (9th model, campaign complete)

witcheer

ec547ad4mo ago

add Qwen3.6 27B dense benchmark (MoE vs dense control experiment)

witcheer

44b8dd74mo ago

Upload README.md with huggingface_hub

witcheer

4036b8a4mo ago