|
{ |
|
"adapter_labels": [ |
|
"DA", |
|
"S_FR", |
|
"B_FR", |
|
"IT", |
|
"S_DE", |
|
"G_DE" |
|
], |
|
"alpha_pattern": {}, |
|
"auto_mapping": null, |
|
"base_model_name_or_path": "ai-forever/mGPT", |
|
"bias": "none", |
|
"exclude_modules": null, |
|
"fan_in_fan_out": true, |
|
"inference_mode": true, |
|
"init_lora_weights": true, |
|
"layer_group": 3, |
|
"layer_replication": null, |
|
"layers_pattern": "h", |
|
"layers_to_transform": [ |
|
18, |
|
19, |
|
20, |
|
21, |
|
22, |
|
23 |
|
], |
|
"loftq_config": {}, |
|
"lora_alpha": 26, |
|
"lora_dropout": 0.1, |
|
"megatron_config": null, |
|
"megatron_core": "megatron.core", |
|
"modules_to_save": null, |
|
"num_adapters_per_layer": 6, |
|
"peft_type": "LORA", |
|
"r": 13, |
|
"r_a": [ |
|
3, |
|
12, |
|
2, |
|
21, |
|
16, |
|
10 |
|
], |
|
"rank_pattern": {}, |
|
"revision": null, |
|
"target_modules": [ |
|
"c_attn", |
|
"c_proj" |
|
], |
|
"task_type": "CAUSAL_LM", |
|
"use_dora": false, |
|
"use_rslora": false |
|
} |