{
 "schema_version": "1.7.0",
 "id": "x_NXMC-2024-0172",
 "modified": "2026-10-19T02:33:38.839Z",
 "published": "2026-10-18T09:00:52.565Z",
 "summary": "saiga_llama3_8b: tokenizer change at bc54c81 (2024-04-26)",
 "details": "saiga_llama3_8b: EOS changed from \"<|im_end|>\" to \"<|eot_id|>\"; anything fine-tuned or served against the old value stops at a different token; BOS changed from \"<|im_start|>\" to \"<|begin_of_text|>\"; prompts built by the old tokenizer start with a different token; pad token \"<|im_start|>\" to \"<|begin_of_text|>\"; batched training masks a different id (2024-04-26).\n\nRepository: https://huggingface.co/IlyaGusev/saiga_llama3_8b. Revisions from bc54c81cc63aa4836f64171f6fbb07d1f904b674 onward carry this change; a pin at 91d7b8a51d0dde9658d6f715a01f1ec2cc2a9f50 or earlier does not.\n\nChanged keys:\n- tokenizer bos_token: <|im_start|> -> <|begin_of_text|> (tokenizer)\n- tokenizer chat_template_sha256: 64cc1405a15b6b72 -> ba03a121d097859c (template)\n- tokenizer eos_token: <|im_end|> -> <|eot_id|> (tokenizer)\n- tokenizer pad_token: <|im_start|> -> <|begin_of_text|> (tokenizer)\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/IlyaGusev/saiga_llama3_8b",
     "events": [
      {
       "introduced": "bc54c81cc63aa4836f64171f6fbb07d1f904b674"
      }
     ]
    }
   ],
   "database_specific": {
    "model": "IlyaGusev/saiga_llama3_8b",
    "previous_revision": "91d7b8a51d0dde9658d6f715a01f1ec2cc2a9f50"
   }
  }
 ],
 "references": [
  {
   "type": "INTRODUCED",
   "url": "https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/bc54c81cc63aa4836f64171f6fbb07d1f904b674"
  },
  {
   "type": "PACKAGE",
   "url": "https://huggingface.co/IlyaGusev/saiga_llama3_8b"
  },
  {
   "type": "WEB",
   "url": "https://www.neurarch.com/changes-feeds/models/IlyaGusev/saiga_llama3_8b.html"
  }
 ],
 "database_specific": {
  "source": "neurarch model change tracker",
  "model": "IlyaGusev/saiga_llama3_8b",
  "kind": "tokenizer",
  "classes": [
   "tokenizer",
   "template"
  ],
  "changed_keys": [
   {
    "file": "tokenizer",
    "key": "bos_token",
    "before": "<|im_start|>",
    "after": "<|begin_of_text|>",
    "class": "tokenizer"
   },
   {
    "file": "tokenizer",
    "key": "chat_template_sha256",
    "before": "64cc1405a15b6b72",
    "after": "ba03a121d097859c",
    "class": "template"
   },
   {
    "file": "tokenizer",
    "key": "eos_token",
    "before": "<|im_end|>",
    "after": "<|eot_id|>",
    "class": "tokenizer"
   },
   {
    "file": "tokenizer",
    "key": "pad_token",
    "before": "<|im_start|>",
    "after": "<|begin_of_text|>",
    "class": "tokenizer"
   }
  ],
  "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": {
    "before": "<|im_end|>",
    "after": "<|eot_id|>"
   },
   "bos": {
    "before": "<|im_start|>",
    "after": "<|begin_of_text|>"
   }
  },
  "commit_date": "2024-04-26T06:20:18.000Z",
  "detected_at": "2026-10-18T09:00:52.565Z",
  "detected_by": "backfill",
  "days_since_first_commit": 7,
  "license": "CC-BY-4.0"
 }
}