long_llama_3b_instruct / config.json
Szymon Tworkowski
init release
4552b82
{
"architectures": [
"LongLlamaForCausalLM"
],
"auto_map": {
"AutoConfig": "configuration_longllama.LongLlamaConfig",
"AutoModel": "modeling_longllama.LongLlamaModel",
"AutoModelForCausalLM": "modeling_longllama.LongLlamaForCausalLM",
"AutoModelForSequenceClassification": "modeling_longllama.LongLlamaForSequenceClassification"
},
"bos_token_id": 1,
"eos_token_id": 2,
"gradient_checkpoint_every_ith": 1,
"hidden_act": "silu",
"hidden_size": 3200,
"initializer_range": 0.02,
"intermediate_size": 8640,
"last_context_length": 1024,
"max_position_embeddings": 2048,
"mem_attention_grouping": null,
"mem_dtype": "bfloat16",
"mem_layers": [
6,
12,
18
],
"mem_positionals": true,
"model_type": "longllama",
"num_attention_heads": 32,
"num_hidden_layers": 26,
"pad_token_id": 0,
"rms_norm_eps": 1e-06,
"tie_word_embeddings": false,
"torch_attention": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.30.0",
"use_cache": true,
"vocab_size": 32000
}