A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
C avx2c99cpu-inferencedeep-learningfrom-scratchinference-enginekimi-k3linear-attentionllmllm-inferencemachine-learningmemory-efficientmixture-of-expertsmoemxfp4quantizationsimdsystems-programmingtransformerzero-dependencies
| Name | Last commit message | Last commit date |
|---|---|---|
| .. | ||
| cache | ||
| cfg | ||
| gates | ||
| golden | ||
| ops | ||
| st | ||
| expert_trace.bin 782 KB | ||
| mxfp4.json 2.8 MB | ||
| partial_index_nomap.json 135 B | ||
| partial_index.json 1.7 KB | ||
| README.md 3.5 KB | ||
| ref_k3.json 5.2 KB | ||
| tiny_k3.bin 8.5 MB | ||
| tiny_k3.json 72.7 KB | ||