SST-2-GLoRA-p40-seed62 / adapter_config.json
ekiprop's picture
Training in progress, step 200
d64d1dc verified
{
"alpha_pattern": {},
"auto_mapping": null,
"base_model_name_or_path": "roberta-base",
"bias": "none",
"corda_config": null,
"eva_config": null,
"exclude_modules": null,
"fan_in_fan_out": false,
"inference_mode": true,
"init_lora_weights": true,
"layer_replication": null,
"layers_pattern": null,
"layers_to_transform": null,
"loftq_config": {},
"lora_alpha": 16,
"lora_bias": false,
"lora_dropout": 0.1,
"megatron_config": null,
"megatron_core": "megatron.core",
"modules_to_save": [
"classifier",
"score"
],
"peft_type": "LORA",
"qalora_group_size": 16,
"r": 8,
"rank_pattern": {},
"revision": null,
"target_modules": "roberta\\.encoder\\.layer\\.11\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.10\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.11\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.10\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.9\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.11\\.output\\.dense|roberta\\.encoder\\.layer\\.8\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.10\\.output\\.dense|roberta\\.encoder\\.layer\\.6\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.7\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.9\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.5\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.9\\.output\\.dense|roberta\\.encoder\\.layer\\.3\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.2\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.8\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.4\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.8\\.output\\.dense|roberta\\.encoder\\.layer\\.7\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.6\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.5\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.2\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.1\\.attention\\.self\\.value|roberta\\.encoder\\.layer\\.11\\.attention\\.output\\.dense|roberta\\.encoder\\.layer\\.4\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.3\\.intermediate\\.dense|roberta\\.encoder\\.layer\\.7\\.output\\.dense|roberta\\.encoder\\.layer\\.10\\.attention\\.output\\.dense",
"task_type": "SEQ_CLS",
"trainable_token_indices": null,
"use_dora": false,
"use_qalora": false,
"use_rslora": false
}