pravsels's picture
Training in progress, step 500
bd2746d verified
raw
history blame contribute delete
791 Bytes
{
"_name_or_path": "smallcloudai/Refact-1_6B-fim",
"architectures": [
"GPTRefactForCausalLM"
],
"attention_bias_in_fp32": true,
"attention_softmax_in_fp32": true,
"auto_map": {
"AutoConfig": "smallcloudai/Refact-1_6B-fim--configuration_gpt_refact.GPTRefactConfig",
"AutoModelForCausalLM": "smallcloudai/Refact-1_6B-fim--modeling_gpt_refact.GPTRefactForCausalLM"
},
"do_sample": true,
"eos_token_id": 0,
"initializer_range": 0.02,
"layer_norm_epsilon": 1e-05,
"model_type": "gpt_refact",
"multi_query": true,
"n_embd": 2048,
"n_head": 32,
"n_inner": null,
"n_layer": 32,
"n_positions": 4096,
"scale_attention_softmax_in_fp32": true,
"torch_dtype": "float32",
"transformers_version": "4.35.2",
"use_cache": true,
"vocab_size": 49216
}