Wiola13M / config.json
aryuemaan
Release Wiola13M v1.0.0
be81bae
Raw
History Blame Contribute Delete
572 Bytes
{
"architectures": [
"WiolaForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 1,
"eos_token_id": 2,
"gate_pre_rope": true,
"hidden_size": 256,
"initializer_range": 0.02,
"intermediate_size": 512,
"max_position_embeddings": 512,
"model_type": "wiola",
"num_attention_heads": 8,
"num_hidden_layers": 6,
"pad_token_id": 3,
"rms_norm_eps": 1e-06,
"rope_theta": 10000.0,
"spiral_alpha": 0.05,
"torch_dtype": "float32",
"transformers_version": "4.45.2",
"use_attention_gate": true,
"use_cache": true,
"vocab_size": 32000
}