moe-hindi-fm-base-1 / config.json
convaiinnovations's picture
Upload Hindi MoE model - step 7000
b18a717 verified
Raw History Blame Contribute Delete
537 Bytes
{
"vocab_size": 16000,
"hidden_size": 768,
"num_hidden_layers": 12,
"num_attention_heads": 16,
"num_key_value_heads": 4,
"intermediate_size": 3072,
"hidden_act": "silu",
"max_position_embeddings": 512,
"initializer_range": 0.02,
"layer_norm_eps": 1e-05,
"use_cache": true,
"pad_token_id": 0,
"bos_token_id": 1,
"eos_token_id": 2,
"tie_word_embeddings": false,
"attention_dropout": 0.0,
"use_moe": true,
"num_local_experts": 8,
"num_experts_per_tok": 2,
"rope_theta": 10000.0,
"mlp_bias": false
}