26 lines
591 B
JSON
26 lines
591 B
JSON
{
|
|
"architectures": [
|
|
"Fairseq2LlamaForCausalLM"
|
|
],
|
|
"bos_token_id": 128000,
|
|
"eos_token_id": 128001,
|
|
"hidden_size": 2048,
|
|
"intermediate_size": 8192,
|
|
"max_position_embeddings": 131072,
|
|
"model_type": "llama",
|
|
"num_attention_heads": 32,
|
|
"num_hidden_layers": 16,
|
|
"num_key_value_heads": 8,
|
|
"rms_norm_eps": 1e-05,
|
|
"rope_scaling": {
|
|
"factor": 32.0,
|
|
"high_freq_factor": 5.0,
|
|
"low_freq_factor": 1.0,
|
|
"original_max_position_embeddings": 8192,
|
|
"rope_type": "llama3"
|
|
},
|
|
"rope_theta": 500000.0,
|
|
"tie_word_embeddings": false,
|
|
"vocab_size": 128256
|
|
}
|