Skip to content

armgit / janas

Public Imported from GitHub
0

37 tok/s from Qwen3-Next-80B (48 GB) on a 32 GB laptop, 24 without drafts vs 2.8 for llama.cpp; reading 62 vs 16 tok/s. Faster than llama.cpp on all 13 models compared in Q4_K_M. LLMs bigger than your RAM, pure C, CPU or GPU.

NameLast commit date
..
fb
.gitkeep
audio_clips.h
bench_attn.c
bench_gemm.c
bench_kernels.c
bench_long.c
bench_matvec.c
bench_moe_layer.c
bench_power.c
fb_check.c
fb_run.h
gpu_check.c
gpu_fast_check.c
gpu_ffn_check.c
jns_check.c
llm_eval.c
llm_gen.c
llm_keep.c
llm_session.c
llm_slide.c
mcp_fake_server.c
mtp_eval.c
node_check.c
node_run.h
npu_check.c
test_aligned.c
test_attention.c
test_audio.c
test_deflate.c
test_deltanet.c
test_expert_cache.c
test_fb.c
test_flights_parse.c
test_geo.c
test_get_catalog.c
test_gguf.c
test_git_parse.c
test_github_parse.c
test_gpu_output_copy.c
test_grammar.c
test_image_data.h
test_image.c
test_import_regions.c
test_jmath.c
test_jns.c
test_layout.c
test_locate.c
test_maps_parse.c
test_mcp.c
test_mxfp4.c
test_node.c
test_npu.c
test_nsm.c
test_pool.c
test_prices_parse.c
test_quakes_parse.c
test_quant.c
test_sample.c
test_sandbox.c
test_server_files.c
test_server_fim.c
test_server_origin.c
test_server_request.c
test_server_store.c
test_sha256.c
test_system_parse.c
test_template.c
test_tools.c
test_tuner.c
test_weather_parse.c
test_wiki_parse.c
test_words.c
tok_check.c
vocab_check.c