{
 "schema_version": "1.7.0",
 "id": "x_NXMC-2024-0229",
 "modified": "2026-11-04T07:40:29.540Z",
 "published": "2026-11-01T12:02:00.579Z",
 "summary": "llama-3-8b-Instruct-bnb-4bit: tokenizer change at 7a83eed (2024-08-06)",
 "details": "llama-3-8b-Instruct-bnb-4bit: generation_config.json pad_token_id unset to 128255; generate() fills finished sequences in a batch with this id, while training pads with the tokenizer's pad token, which did not change; generation defaults changed (do_sample unset to true, max_length unset to 4096, temperature unset to 0.6, top_p unset to 0.9); a call that does not pass these explicitly now generates differently (2024-08-06).\n\nRepository: https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit. Revisions from 7a83eed3a6850fefe81b0250fbe5bc563fecf620 onward carry this change; a pin at 7400af6f3bfa4bc4f7e1afb04dc681329a0dc01f or earlier does not.\n\nChanged keys:\n- generation do_sample: unset -> true (generation-defaults)\n- generation max_length: unset -> 4096 (generation-defaults)\n- generation pad_token_id: unset -> 128255 (tokenizer)\n- generation temperature: unset -> 0.6 (generation-defaults)\n- generation top_p: unset -> 0.9 (generation-defaults)\n\nThis record says what the change does mechanically. It is not a report that anything broke for anyone.",
 "affected": [
  {
   "ranges": [
    {
     "type": "GIT",
     "repo": "https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit",
     "events": [
      {
       "introduced": "7a83eed3a6850fefe81b0250fbe5bc563fecf620"
      }
     ]
    }
   ],
   "database_specific": {
    "model": "unsloth/llama-3-8b-Instruct-bnb-4bit",
    "previous_revision": "7400af6f3bfa4bc4f7e1afb04dc681329a0dc01f"
   }
  }
 ],
 "references": [
  {
   "type": "INTRODUCED",
   "url": "https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit/commit/7a83eed3a6850fefe81b0250fbe5bc563fecf620"
  },
  {
   "type": "PACKAGE",
   "url": "https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit"
  },
  {
   "type": "WEB",
   "url": "https://www.neurarch.com/changes-feeds/models/unsloth/llama-3-8b-Instruct-bnb-4bit.html"
  }
 ],
 "database_specific": {
  "source": "neurarch model change tracker",
  "model": "unsloth/llama-3-8b-Instruct-bnb-4bit",
  "kind": "tokenizer",
  "classes": [
   "tokenizer",
   "generation-defaults"
  ],
  "changed_keys": [
   {
    "file": "generation",
    "key": "do_sample",
    "before": null,
    "after": true,
    "class": "generation-defaults"
   },
   {
    "file": "generation",
    "key": "max_length",
    "before": null,
    "after": 4096,
    "class": "generation-defaults"
   },
   {
    "file": "generation",
    "key": "pad_token_id",
    "before": null,
    "after": 128255,
    "class": "tokenizer"
   },
   {
    "file": "generation",
    "key": "temperature",
    "before": null,
    "after": 0.6,
    "class": "generation-defaults"
   },
   {
    "file": "generation",
    "key": "top_p",
    "before": null,
    "after": 0.9,
    "class": "generation-defaults"
   }
  ],
  "impact": {
   "memory": false,
   "serving": true,
   "tokenizer": true,
   "licence": null
  },
  "breaking": true,
  "breaking_rule": "https://www.neurarch.com/changes-feeds/advisories/index.json#breaking_rule",
  "consequence": {
   "params": null,
   "weightsGbBf16": null,
   "trainPeakGb": null,
   "fingerprint": null,
   "context": null,
   "eos": null,
   "bos": null,
   "derivedSame": true
  },
  "commit_date": "2024-08-06T23:43:01.000Z",
  "detected_at": "2026-11-01T12:02:00.579Z",
  "detected_by": "backfill",
  "days_since_first_commit": 110,
  "license": "CC-BY-4.0"
 }
}