{
 "recorded_with": {
  "contract": "huggingface-mic/conv-bert-base-mic",
  "contract_name": "ConvBERT",
  "contract_revision": "c82aa40351add05270079c854e9282ea8956166e",
  "framework": "transformers",
  "variant": "bfloat16",
  "target": "conv-bert-base",
  "kind": "checkpoint",
  "revision": "5cb451936b5c4a96562d8b146de85f64f9cf2c22",
  "environment": {
   "python": "3.11.18",
   "transformers": "5.22.0.dev0",
   "torch": "2.14.0+cu130",
   "torchvision": "0.29.2",
   "torchaudio": "2.11.0+cu130",
   "tokenizers": "0.23.3",
   "numpy": "2.4.6",
   "pillow": "13.0.0",
   "accelerate": "1.16.0.dev0",
   "safetensors": "0.8.0",
   "huggingface-hub": "2.3.0",
   "sentencepiece": "0.2.2",
   "protobuf": "7.36.2",
   "librosa": "0.11.0",
   "soundfile": "0.14.0",
   "scipy": "1.17.1"
  },
  "seconds": 8.55,
  "environment_differs": {
   "torchvision": "0.29.2 (pinned 0.29.1)",
   "numpy": "2.4.6 (pinned 2.3.5)",
   "pillow": "13.0.0 (pinned 12.3.0)",
   "accelerate": "1.16.0.dev0 (pinned 1.15.0)"
  },
  "platform": "cuda"
 },
 "outputs": {
  "runtime": {
   "device": "cuda",
   "dtype": "bfloat16",
   "attn_implementation": "eager"
  },
  "hidden_probe": [
   [
    -0.022216796875,
    -0.4140625
   ],
   [
    0.2353515625,
    0.55078125
   ],
   [
    0.2041015625,
    -0.0283203125
   ]
  ],
  "pooled_probe": [
   -0.022216796875,
   -0.4140625,
   -0.490234375,
   -0.2890625
  ],
  "training": {
   "loss": 0.73046875,
   "gradients": {
    "encoder.layer.0.attention.self.query.weight": true,
    "encoder.layer.0.attention.self.key_conv_attn_layer.depthwise.weight": true,
    "encoder.layer.0.intermediate.dense.weight": true
   },
   "updated": true
  }
 }
}