AndreaUnibo's picture
Training in progress, step 117000, checkpoint
d1c4c25 verified
{
"alpha_pattern": {},
"auto_mapping": null,
"base_model_name_or_path": "rank_lstm_full_trained_depth3_n2",
"bias": "none",
"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_dropout": 0.05,
"megatron_config": null,
"megatron_core": "megatron.core",
"modules_to_save": null,
"peft_type": "LORA",
"r": 8,
"rank_pattern": {},
"revision": null,
"target_modules": [
"model.layers.1.mlp.router.layer",
"model.layers.21.self_attention.experts.router.layer",
"model.layers.20.self_attention.experts.router.layer",
"model.layers.12.self_attention.kv_proj",
"model.layers.20.self_attention.kv_proj",
"model.layers.15.self_attention.kv_proj",
"model.layers.18.self_attention.kv_proj",
"model.lstm_gates.0.transform",
"model.lstm_gates.1.transform",
"model.layers.6.self_attention.kv_proj",
"model.layers.10.self_attention.kv_proj",
"model.layers.7.self_attention.kv_proj",
"model.lstm_gates.2.transform",
"model.lstm_gates.0.input_gate",
"model.layers.2.self_attention.kv_proj",
"model.layers.20.mlp.router.layer",
"model.layers.23.self_attention.kv_proj",
"model.layers.19.mlp.router.layer",
"model.layers.16.self_attention.kv_proj",
"model.layers.4.self_attention.kv_proj",
"model.layers.1.mlp.router.perturbation",
"model.layers.19.self_attention.kv_proj",
"model.lstm_gates.0.forget_gate",
"model.layers.1.self_attention.experts.router.layer",
"model.layers.21.mlp.router.layer",
"model.layers.9.self_attention.kv_proj",
"model.lstm_gates.1.input_gate",
"model.layers.21.self_attention.kv_proj",
"model.layers.22.self_attention.kv_proj",
"model.lstm_gates.1.output_gate",
"model.layers.5.self_attention.kv_proj",
"model.layers.8.self_attention.kv_proj",
"model.layers.17.self_attention.kv_proj",
"model.layers.3.self_attention.kv_proj",
"model.layers.11.self_attention.kv_proj",
"model.lstm_gates.0.output_gate",
"model.lstm_gates.2.forget_gate",
"model.layers.0.self_attention.kv_proj",
"model.lstm_gates.1.forget_gate",
"model.layers.13.self_attention.kv_proj",
"model.lstm_gates.2.output_gate",
"model.layers.1.self_attention.kv_proj",
"model.layers.19.self_attention.experts.router.layer",
"model.layers.14.self_attention.kv_proj",
"model.lstm_gates.2.input_gate"
],
"task_type": "CAUSAL_LM",
"use_dora": false,
"use_rslora": false
}