{
  "id": "orca-2-7b",
  "name": "Orca 2 7B",
  "identifiers": {
    "huggingface": [
      "microsoft/Orca-2-7b"
    ]
  },
  "developer": "Microsoft Research",
  "release_date": {
    "value": "2023-11-18",
    "source": "https://arxiv.org/abs/2311.11045",
    "status": "recorded",
    "note": "arXiv v1 submission date; the abstract announces the weights ('We make Orca 2 weights publicly available at aka.ms/orca-lm'). The HF repo's first commit is 2023-11-21 (HF API); repo creation 2023-11-14."
  },
  "weights_status": "open",
  "availability": {
    "value": "available",
    "checked": "2026-09-24",
    "source": "https://huggingface.co/microsoft/Orca-2-7b"
  },
  "license": {
    "value": "Microsoft Research License",
    "source": "https://huggingface.co/microsoft/Orca-2-7b/blob/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/README.md",
    "status": "recorded",
    "note": "Card: 'Orca 2 is licensed under the Microsoft Research License.' It adds: 'Llama 2 is licensed under the LLAMA 2 Community License', which applies to the base."
  },
  "architecture": {
    "family": "decoder_only",
    "note": "family read from config 'architectures': ['LlamaForCausalLM'].",
    "n_layers": {
      "value": 32,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "hidden_size": {
      "value": 4096,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "n_heads": {
      "value": 32,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "vocab_size": {
      "value": 32003,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "context_length": {
      "value": 4096,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "n_kv_heads": {
      "value": 32,
      "source": "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
      "status": "recorded"
    },
    "positional_encoding": {
      "value": "rotary (RoPE)",
      "source": "https://arxiv.org/abs/2307.09288",
      "status": "recorded",
      "note": "Propagated from llama-2-7b: the card says 'Please refer to LLaMA-2 technical report for details on the model architecture' and the config matches."
    }
  },
  "training_data": {
    "value": "Synthetic data created to improve small-model reasoning. Paper (progressive learning): start from the LLaMA-2-7B checkpoint; fine-tune on the FLAN-v2 train split (1 epoch); then on 5 million ChatGPT examples from Orca 1 (3 epochs); then on 1 million GPT-4 examples from Orca 1 plus Orca 2's 817K examples (4 epochs). Card: all synthetic data moderated with Azure content filters.",
    "source": "https://arxiv.org/abs/2311.11045",
    "status": "recorded",
    "note": "Paper section 4.2, 'Progressive Learning'. Loss is computed only on teacher-generated tokens."
  },
  "techniques": [],
  "primary_sources": [
    "https://huggingface.co/microsoft/Orca-2-7b",
    "https://huggingface.co/microsoft/Orca-2-7b/raw/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/config.json",
    "https://huggingface.co/microsoft/Orca-2-7b/blob/60e31e6bdcf582ad103b807cb74b73ee1d2c4b17/README.md",
    "https://arxiv.org/abs/2311.11045"
  ],
  "record_history": [
    {
      "date": "2026-09-24",
      "change": "ingested as candidate from HF (microsoft/Orca-2-7b@60e31e6bdcf582ad103b807cb74b73ee1d2c4b17)",
      "by": "ingest_hf.py"
    },
    {
      "date": "2026-09-24",
      "change": "preparer: developer, release date (arXiv), license, positional encoding (propagated from llama-2-7b), training data, 1 edge accepted + 2 PANEL-NEEDED; sources: card, arXiv 2311.11045",
      "by": "claude (preparer, Sonnet 5)"
    },
    {
      "date": "2026-09-24",
      "change": "ruling: 3-tier panel (2/3 dataset routing), see session log: 2 direct edges rejected, 3 trained_on edges added",
      "by": "wilson-pruitt + claude (Sonnet 5)"
    },
    {
      "date": "2026-09-24",
      "change": "reviewed and promoted from staging (4 edge(s) accepted)",
      "by": "Wilson Pruitt"
    }
  ]
}