urdshals's picture
Upload final model (step 0) and all checkpoints at 2024-12-06T19:22:20.047757
de5550b verified
raw
history blame
218 Bytes
{
"architectures": [
"HFHookedTransformer"
],
"hidden_size": 128,
"num_attention_heads": 4,
"num_hidden_layers": 6,
"torch_dtype": "float32",
"transformers_version": "4.45.2",
"vocab_size": 50304
}