{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigw3ctkfhfofwihgw5tt3bke4n5ipvshj6ddtw4lfdx2vjgavkhem",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mh36zf3atid2"
  },
  "path": "/t/wave-coherence-scoring-phase-aware-alternative-to-cosine-similarity/173375#post_9",
  "publishedAt": "2026-03-15T01:16:44.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "kerr-server",
    "Wave Coherence",
    "Kerr Engine",
    "Kerr Server"
  ],
  "textContent": "Kerr-ODE: Full Stack Now Live — Train, Serve, Chat\n\nThree updates since the last post:\n\n**1. Inference Server Released** kerr-server — 640 lines of Rust, OpenAI-compatible API. Load a checkpoint, serve on localhost, connect any chat UI. SSE streaming, bearer token auth, temperature/top-k/top-p sampling.\n\nTested with LM Studio 0.4.6 via the openai-compat-endpoint plugin. Selected “gpt-4.1” from the model dropdown (the server doesn’t care what name the client sends — it serves whatever checkpoint is loaded). Typed “hello.” A 354K parameter Shakespeare-trained Kerr-ODE model responded with character-level Shakespeare fragments through the GPT-4.1 label. Not exactly GPT-4.1’s output, but the pipeline works end to end.\n\nThe full pipeline: train a model with the engine → serve it with kerr-server → chat through LM Studio or any OpenAI-compatible client. No custom clients needed.\n\n**2. Engine Performance: 13s → 1.72s at 768-dim** Three new WGSL compute shaders (`matvec_batch`, `layer_norm_batch`, `kerr_step_batch`) batch all forward pass operations across positions in single dispatches. Forward pass went from 8 seconds to 500ms — 16x speedup. Total iteration from 13s to 1.72s. 17 shaders total, ~7,200 lines of Rust. GPU at 38%, 49°C, 1.2GB VRAM training a 12M parameter model on an RTX 4070 Ti.\n\n**3. BPE Tokenizer Support** The engine and server now accept any HuggingFace `tokenizer.json` via `--bpe` flag. Qwen, Llama, GPT-2 — borrow their vocabulary, train your own Kerr-ODE model from scratch. No more character-level only.\n\n**The stack:**\n\n  * Wave Coherence — research framework, 68 defensive patents (MIT)\n\n  * Kerr Engine — training, 17 WGSL shaders, 3x faster than PyTorch at 128-dim (Apache 2.0)\n\n  * Kerr Server — inference, OpenAI-compatible API (Apache 2.0)\n\n\n\n\nAll open source. All documented. Contributor targets listed in both READMEs.",
  "title": "Wave-Coherence Scoring: Phase-Aware Alternative to Cosine Similarity"
}