bloom-560m_am_ia3_10000samples / adapter_config.json
Davlan's picture
Upload adapter_config.json
4b51a78
{
"config": {
"architecture": "union",
"configs": [
{
"alpha": 1,
"architecture": "lora",
"attn_matrices": [
"k",
"v"
],
"composition_mode": "scale",
"dropout": 0.0,
"init_weights": "ia3",
"intermediate_lora": false,
"output_lora": false,
"r": 1,
"selfattn_lora": true
},
{
"adapter_residual_before_ln": false,
"cross_adapter": false,
"factorized_phm_W": true,
"factorized_phm_rule": false,
"hypercomplex_nonlinearity": "glorot-uniform",
"init_weights": "bert",
"inv_adapter": "nice",
"inv_adapter_reduction_factor": 2,
"is_parallel": false,
"learn_phm": true,
"leave_out": [],
"ln_after": false,
"ln_before": false,
"mh_adapter": false,
"non_linearity": "relu",
"original_ln_after": false,
"original_ln_before": false,
"output_adapter": false,
"phm_bias": true,
"phm_c_init": "normal",
"phm_dim": 4,
"phm_init_range": 0.0001,
"phm_layer": false,
"phm_rank": 1,
"reduction_factor": 16,
"residual_before_ln": true,
"scaling": 1.0,
"shared_W_phm": false,
"shared_phm_rule": true
}
]
},
"config_id": "d2c8db3436274834",
"hidden_size": 1024,
"model_class": "BloomForCausalLM",
"model_name": "bigscience/bloom-560m",
"model_type": "bloom",
"name": "oscar_ia3+inv_am",
"version": "3.1.0a0"
}