T-pro-it-2.1-eagle3 / config.json
VirVen's picture
Add EAGLE-3 draft model for T-pro-it-2.1-FP8
1ccffe1 verified
Raw
History Blame Contribute Delete
623 Bytes
{
"depth": 5,
"draft_vocab_size": 151689,
"hidden_act": "silu",
"hidden_size": 5120,
"high_layer_index": 62,
"intermediate_size": 25600,
"low_layer_index": 8,
"max_position_embeddings": 40960,
"mid_layer_index": 32,
"model_type": "eagle3",
"num_attention_heads": 64,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 1000000.0,
"rope_type": "default"
},
"rope_theta": 1000000.0,
"target_hidden_size": 5120,
"threshold": 1.0,
"top_k": 8,
"total_tokens": 63,
"transformers_version": "5.11.0",
"vocab_size": 151689
}