dart-v1-sft / config.json
p1atdev's picture
Upload OPTForCausalLM
cc93325 verified
raw history blame
No virus
928 Bytes
{
"_name_or_path": "p1atdev/dart-test-3-sft-2",
"_remove_final_layer_norm": false,
"activation_function": "relu",
"architectures": [
"OPTForCausalLM"
],
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "p1atdev/dart-test-3-sft-2--configuration_dart.DartConfig",
"AutoModelForCausalLM": "p1atdev/dart-test-3-sft-2--modeling_dart.DartForCausalLM"
},
"bos_token_id": 0,
"do_layer_norm_before": true,
"dropout": 0.1,
"enable_bias": true,
"eos_token_id": 1,
"ffn_dim": 3072,
"hidden_size": 768,
"init_std": 0.02,
"layer_norm_elementwise_affine": true,
"layerdrop": 0.0,
"max_position_embeddings": 2048,
"model_type": "opt",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"pad_token_id": 2,
"torch_dtype": "float32",
"transformers_version": "4.38.0",
"use_cache": true,
"use_position_embeddings": false,
"vocab_size": 67996,
"word_embed_proj_dim": 768
}