Doge-120M-MoE / config.json
JingzeShi's picture
Training in progress, step 100
3aa6cfb verified
raw
history blame contribute delete
919 Bytes
{
"_name_or_path": "./Doge-60M-checkpoint",
"architectures": [
"DogeForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 0,
"dynamic_mask_ratio": 0.0,
"eos_token_id": 1,
"hidden_act": "silu",
"hidden_bias": false,
"hidden_dropout": 0.0,
"hidden_size": 512,
"initializer_range": 0.02,
"intermediate_size": 1024,
"is_moe": true,
"keep_window_size": 2048,
"max_position_embeddings": 2048,
"model_type": "doge",
"num_attention_heads": 4,
"num_experts": 4096,
"num_experts_per_tok": 32,
"num_hidden_layers": 16,
"num_key_value_heads": 2,
"pad_token_id": 2,
"rms_norm_eps": 1e-06,
"rope_scaling": {
"factor": 4.0,
"original_max_position_embeddings": 2048,
"rope_type": "dynamic"
},
"rope_theta": 10000.0,
"tie_word_embeddings": true,
"torch_dtype": "float32",
"transformers_version": "4.50.0.dev0",
"use_cache": true,
"vocab_size": 32768
}