4M-ACT / config.json
w-ahmad's picture
Training in progress, step 2000
965c195 verified
Raw
History Blame Contribute Delete
829 Bytes
{
"activation": "s10",
"architectures": [
"TinyLlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 1,
"dtype": "bfloat16",
"eos_token_id": 2,
"head_dim": 32,
"hidden_act": "silu",
"hidden_size": 128,
"initializer_range": 0.02,
"intermediate_size": 256,
"max_position_embeddings": 512,
"mlp_bias": false,
"mlp_type": "glu",
"model_type": "tiny_llama",
"num_attention_heads": 4,
"num_hidden_layers": 9,
"num_key_value_heads": 4,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 10000.0,
"rope_type": "default"
},
"tie_word_embeddings": true,
"tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
"transformers_version": "5.15.0.dev0",
"use_cache": false,
"vocab_size": 4096
}