Yishi Xu
added CerebrasGPT-13B_SP checkpoint 20000
31b8767
raw
history blame
360 Bytes
{
"attn_pdrop": 0.0,
"scale_attn_weights": true,
"resid_pdrop": 0.0,
"n_inner": 20480,
"n_embd": 5120,
"layer_norm_epsilon": 1e-05,
"n_positions": 2048,
"activation_function": "gelu",
"n_head": 40,
"n_layer": 40,
"tie_word_embeddings": true,
"vocab_size": 50257,
"embd_pdrop": 0.0,
"model_type": "gpt2"
}