slm-125m-instruct / config.json
tohio's picture
Fix tokenizer layout and remote code auto_map
623df2d verified
Raw
History Blame Contribute Delete
741 Bytes
{
"architectures": [
"SLMForCausalLM"
],
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "config.SLMConfig",
"AutoModelForCausalLM": "model.SLMForCausalLM"
},
"bos_token_id": 2,
"dtype": "bfloat16",
"eos_token_id": 3,
"hidden_size": 768,
"initializer_range": 0.02,
"intermediate_size": 2048,
"max_position_embeddings": 2048,
"model_type": "slm",
"num_attention_heads": 12,
"num_hidden_layers": 16,
"num_key_value_heads": 4,
"pad_token_id": 0,
"rms_norm_eps": 1e-05,
"rope_parameters": {
"rope_theta": 500000.0,
"rope_type": "default"
},
"rope_theta": 500000.0,
"tie_word_embeddings": true,
"transformers_version": "5.5.4",
"use_cache": false,
"vocab_size": 32000
}