src/worker/curated-model-list.json

[
  {
    "id": "mlc-ai/Llama-3.1-8B-Instruct-q4f32_1-MLC",
    "name": "Llama 3.1 8B Instruct q4f32",
    "model_type": "llama",
    "architectures": ["llama"],
    "classification": "gen",
    "confidence": "high",
    "size_hint": "58Mb",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": true,
    "isTransformersJsReady": true
  },
  {
    "id": "/models/google/gemma-2b/resolve/main",
    "name": "Gemma 2B (local /models/google/gemma-2b/resolve/main)",
    "model_type": "gemma",
    "architectures": ["gemma"],
    "classification": "gen",
    "confidence": "medium",
    "size_hint": "~2GB",
    "fetchStatus": "pending-local",
    "hasTokenizer": true,
    "hasOnnxModel": false,
    "isTransformersJsReady": false,
    "requiresAuth": false,
    "info": { "local_url": "/models/google/gemma-2b/resolve/main", "notes": "Place downloaded HF-style repo under ./models/google/gemma-2b so files resolve at /models/google/gemma-2b/resolve/main/..." }
  },
  {
    "id": "Xenova/llama2.c-stories15M",
    "name": "Llama2.c Stories 15M",
    "model_type": "llama",
    "architectures": ["llama"],
    "classification": "gen",
    "confidence": "high",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": true,
    "isTransformersJsReady": true,
    "info": {
      "display_name": "Llama2.c Stories 15M (Xenova)",
      "params": "15M",
      "params_count": 15000000,
      "architecture": "LLaMA-family (tiny conversion for stories use via Xenova toolchain)",
      "context_window": 2048,
      "quantization": "likely float16/uint8/packed formats supported by Xenova runtime",
      "hf_url": "https://huggingface.co/Xenova/llama2.c-stories15M",
      "recommended_runtime": "transformers.js / Xenova runtime (wasm/webgpu)",
      "is_mobile_capable": true,
      "verified": false,
      "assumed": true,
      "notes": "Original curated metadata preserved; tokenizer/ONNX flags assumed for Xenova-converted tiny model."
    }
  },
  {
    "id": "Xenova/phi-3-mini-4k-instruct",
    "name": "phi-3-mini-4k-instruct",
    "model_type": "phi3",
    "architectures": ["phi3"],
    "classification": "gen",
    "confidence": "high",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": true,
    "isTransformersJsReady": true,
    "info": {
      "display_name": "Phi-3 Mini 4k Instruct (Xenova)",
      "params": "mini (estimate)",
      "params_count": null,
      "architecture": "Phi family (Phi-3 mini conversion)",
      "context_window": 4096,
      "hf_url": "https://huggingface.co/Xenova/phi-3-mini-4k-instruct",
      "recommended_runtime": "transformers.js / Xenova runtime (wasm/webgpu)",
      "is_mobile_capable": true,
      "verified": false,
      "assumed": true,
      "notes": "Parameter count and asset layout should be verified against HF API."
    }
  },
  {
    "id": "Xenova/distilgpt2",
    "name": "distilgpt2",
    "model_type": "gpt2",
    "architectures": ["gpt2"],
    "classification": "gen",
    "confidence": "high",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": true,
    "isTransformersJsReady": true,
    "info": {
      "display_name": "DistilGPT-2 (Xenova mirror)",
      "params": "≈82M",
      "params_count": 82000000,
      "architecture": "GPT-2 distilled",
      "context_window": 1024,
      "hf_url": "https://huggingface.co/Xenova/distilgpt2",
      "recommended_runtime": "transformers.js",
      "is_mobile_capable": true,
      "verified": false,
      "assumed": true
    }
  },
  {
    "id": "Xenova/gpt2",
    "name": "gpt2",
    "model_type": "gpt2",
    "architectures": ["gpt2"],
    "classification": "gen",
    "confidence": "high",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": true,
    "size_hint": "500Mb",
    "isTransformersJsReady": true,
    "info": {
      "display_name": "GPT-2 (Xenova mirror)",
      "params": "117M (base)",
      "params_count": 117000000,
      "architecture": "GPT-2",
      "context_window": 1024,
      "hf_url": "https://huggingface.co/Xenova/gpt2",
      "recommended_runtime": "transformers.js",
      "is_mobile_capable": true,
      "verified": false,
      "assumed": true
    }
  },
  {
    "id": "Xenova/qwen-2.5-small-instruct",
    "name": "qwen-2.5-small-instruct",
    "model_type": "qwen",
    "architectures": ["qwen"],
    "classification": "gen",
    "confidence": "medium",
    "fetchStatus": "ok",
    "hasTokenizer": true,
    "hasOnnxModel": false,
    "isTransformersJsReady": false,
    "requiresAuth": true,
    "info": {
      "display_name": "Qwen-2.5 Small Instruct (Xenova mirror)",
      "params": "≈2.5B (estimate)",
      "params_count": null,
      "architecture": "Qwen family",
      "context_window": 8192,
      "hf_url": "https://huggingface.co/Xenova/qwen-2.5-small-instruct",
      "recommended_runtime": "Xenova runtime / transformers.js (quantized)",
      "is_mobile_capable": false,
      "verified": false,
      "assumed": true,
      "notes": "Marked conservatively as not transformers.js-ready / no ONNX in HF assets."
    }
  }
]