mgpt2-dpo / config.json
ace-1's picture
Publish mgpt2 dpo checkpoint (step 420, val_loss 0.001878)
b640ec6 verified
{
"architectures": [
"GPT"
],
"model_type": "mgpt2",
"block_size": 1024,
"vocab_size": 50304,
"n_layer": 12,
"n_head": 12,
"n_embd": 768,
"tokenizer_kind": "mgpt2"
}