testypoos2 / last-checkpoint /lora_lower /adapter_config.json
alicegoesdown's picture
Training in progress, step 150, checkpoint
8cacbab verified
{
"alpha_pattern": {
"transformer.h.0.mlp.dense_4h_to_h": 8,
"transformer.h.0.mlp.dense_h_to_4h": 8,
"transformer.h.0.self_attention.dense": 8,
"transformer.h.0.self_attention.query_key_value": 8,
"transformer.h.1.mlp.dense_4h_to_h": 8,
"transformer.h.1.mlp.dense_h_to_4h": 8,
"transformer.h.1.self_attention.dense": 8,
"transformer.h.1.self_attention.query_key_value": 8,
"transformer.h.10.mlp.dense_4h_to_h": 8,
"transformer.h.10.mlp.dense_h_to_4h": 8,
"transformer.h.10.self_attention.dense": 8,
"transformer.h.10.self_attention.query_key_value": 8,
"transformer.h.11.mlp.dense_4h_to_h": 8,
"transformer.h.11.mlp.dense_h_to_4h": 8,
"transformer.h.11.self_attention.dense": 8,
"transformer.h.11.self_attention.query_key_value": 8,
"transformer.h.12.mlp.dense_4h_to_h": 16,
"transformer.h.12.mlp.dense_h_to_4h": 16,
"transformer.h.12.self_attention.dense": 16,
"transformer.h.12.self_attention.query_key_value": 16,
"transformer.h.13.mlp.dense_4h_to_h": 16,
"transformer.h.13.mlp.dense_h_to_4h": 16,
"transformer.h.13.self_attention.dense": 16,
"transformer.h.13.self_attention.query_key_value": 16,
"transformer.h.14.mlp.dense_4h_to_h": 16,
"transformer.h.14.mlp.dense_h_to_4h": 16,
"transformer.h.14.self_attention.dense": 16,
"transformer.h.14.self_attention.query_key_value": 16,
"transformer.h.15.mlp.dense_4h_to_h": 16,
"transformer.h.15.mlp.dense_h_to_4h": 16,
"transformer.h.15.self_attention.dense": 16,
"transformer.h.15.self_attention.query_key_value": 16,
"transformer.h.16.mlp.dense_4h_to_h": 16,
"transformer.h.16.mlp.dense_h_to_4h": 16,
"transformer.h.16.self_attention.dense": 16,
"transformer.h.16.self_attention.query_key_value": 16,
"transformer.h.17.mlp.dense_4h_to_h": 16,
"transformer.h.17.mlp.dense_h_to_4h": 16,
"transformer.h.17.self_attention.dense": 16,
"transformer.h.17.self_attention.query_key_value": 16,
"transformer.h.18.mlp.dense_4h_to_h": 16,
"transformer.h.18.mlp.dense_h_to_4h": 16,
"transformer.h.18.self_attention.dense": 16,
"transformer.h.18.self_attention.query_key_value": 16,
"transformer.h.19.mlp.dense_4h_to_h": 128,
"transformer.h.19.mlp.dense_h_to_4h": 128,
"transformer.h.19.self_attention.dense": 128,
"transformer.h.19.self_attention.query_key_value": 128,
"transformer.h.2.mlp.dense_4h_to_h": 8,
"transformer.h.2.mlp.dense_h_to_4h": 8,
"transformer.h.2.self_attention.dense": 8,
"transformer.h.2.self_attention.query_key_value": 8,
"transformer.h.20.mlp.dense_4h_to_h": 128,
"transformer.h.20.mlp.dense_h_to_4h": 128,
"transformer.h.20.self_attention.dense": 128,
"transformer.h.20.self_attention.query_key_value": 128,
"transformer.h.21.mlp.dense_4h_to_h": 128,
"transformer.h.21.mlp.dense_h_to_4h": 128,
"transformer.h.21.self_attention.dense": 128,
"transformer.h.21.self_attention.query_key_value": 128,
"transformer.h.22.mlp.dense_4h_to_h": 128,
"transformer.h.22.mlp.dense_h_to_4h": 128,
"transformer.h.22.self_attention.dense": 128,
"transformer.h.22.self_attention.query_key_value": 128,
"transformer.h.23.mlp.dense_4h_to_h": 128,
"transformer.h.23.mlp.dense_h_to_4h": 128,
"transformer.h.23.self_attention.dense": 128,
"transformer.h.23.self_attention.query_key_value": 128,
"transformer.h.3.mlp.dense_4h_to_h": 8,
"transformer.h.3.mlp.dense_h_to_4h": 8,
"transformer.h.3.self_attention.dense": 8,
"transformer.h.3.self_attention.query_key_value": 8,
"transformer.h.4.mlp.dense_4h_to_h": 8,
"transformer.h.4.mlp.dense_h_to_4h": 8,
"transformer.h.4.self_attention.dense": 8,
"transformer.h.4.self_attention.query_key_value": 8,
"transformer.h.5.mlp.dense_4h_to_h": 8,
"transformer.h.5.mlp.dense_h_to_4h": 8,
"transformer.h.5.self_attention.dense": 8,
"transformer.h.5.self_attention.query_key_value": 8,
"transformer.h.6.mlp.dense_4h_to_h": 8,
"transformer.h.6.mlp.dense_h_to_4h": 8,
"transformer.h.6.self_attention.dense": 8,
"transformer.h.6.self_attention.query_key_value": 8,
"transformer.h.7.mlp.dense_4h_to_h": 8,
"transformer.h.7.mlp.dense_h_to_4h": 8,
"transformer.h.7.self_attention.dense": 8,
"transformer.h.7.self_attention.query_key_value": 8,
"transformer.h.8.mlp.dense_4h_to_h": 8,
"transformer.h.8.mlp.dense_h_to_4h": 8,
"transformer.h.8.self_attention.dense": 8,
"transformer.h.8.self_attention.query_key_value": 8,
"transformer.h.9.mlp.dense_4h_to_h": 8,
"transformer.h.9.mlp.dense_h_to_4h": 8,
"transformer.h.9.self_attention.dense": 8,
"transformer.h.9.self_attention.query_key_value": 8
},
"auto_mapping": null,
"base_model_name_or_path": "bigscience/bloomz-560m",
"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": 8,
"lora_dropout": 0.0,
"megatron_config": null,
"megatron_core": "megatron.core",
"modules_to_save": null,
"peft_type": "LORA",
"r": 4,
"rank_pattern": {
"transformer.h.0.mlp.dense_4h_to_h": 4,
"transformer.h.0.mlp.dense_h_to_4h": 4,
"transformer.h.0.self_attention.dense": 4,
"transformer.h.0.self_attention.query_key_value": 4,
"transformer.h.1.mlp.dense_4h_to_h": 4,
"transformer.h.1.mlp.dense_h_to_4h": 4,
"transformer.h.1.self_attention.dense": 4,
"transformer.h.1.self_attention.query_key_value": 4,
"transformer.h.10.mlp.dense_4h_to_h": 4,
"transformer.h.10.mlp.dense_h_to_4h": 4,
"transformer.h.10.self_attention.dense": 4,
"transformer.h.10.self_attention.query_key_value": 4,
"transformer.h.11.mlp.dense_4h_to_h": 4,
"transformer.h.11.mlp.dense_h_to_4h": 4,
"transformer.h.11.self_attention.dense": 4,
"transformer.h.11.self_attention.query_key_value": 4,
"transformer.h.12.mlp.dense_4h_to_h": 8,
"transformer.h.12.mlp.dense_h_to_4h": 8,
"transformer.h.12.self_attention.dense": 8,
"transformer.h.12.self_attention.query_key_value": 8,
"transformer.h.13.mlp.dense_4h_to_h": 8,
"transformer.h.13.mlp.dense_h_to_4h": 8,
"transformer.h.13.self_attention.dense": 8,
"transformer.h.13.self_attention.query_key_value": 8,
"transformer.h.14.mlp.dense_4h_to_h": 8,
"transformer.h.14.mlp.dense_h_to_4h": 8,
"transformer.h.14.self_attention.dense": 8,
"transformer.h.14.self_attention.query_key_value": 8,
"transformer.h.15.mlp.dense_4h_to_h": 8,
"transformer.h.15.mlp.dense_h_to_4h": 8,
"transformer.h.15.self_attention.dense": 8,
"transformer.h.15.self_attention.query_key_value": 8,
"transformer.h.16.mlp.dense_4h_to_h": 8,
"transformer.h.16.mlp.dense_h_to_4h": 8,
"transformer.h.16.self_attention.dense": 8,
"transformer.h.16.self_attention.query_key_value": 8,
"transformer.h.17.mlp.dense_4h_to_h": 8,
"transformer.h.17.mlp.dense_h_to_4h": 8,
"transformer.h.17.self_attention.dense": 8,
"transformer.h.17.self_attention.query_key_value": 8,
"transformer.h.18.mlp.dense_4h_to_h": 8,
"transformer.h.18.mlp.dense_h_to_4h": 8,
"transformer.h.18.self_attention.dense": 8,
"transformer.h.18.self_attention.query_key_value": 8,
"transformer.h.19.mlp.dense_4h_to_h": 64,
"transformer.h.19.mlp.dense_h_to_4h": 64,
"transformer.h.19.self_attention.dense": 64,
"transformer.h.19.self_attention.query_key_value": 64,
"transformer.h.2.mlp.dense_4h_to_h": 4,
"transformer.h.2.mlp.dense_h_to_4h": 4,
"transformer.h.2.self_attention.dense": 4,
"transformer.h.2.self_attention.query_key_value": 4,
"transformer.h.20.mlp.dense_4h_to_h": 64,
"transformer.h.20.mlp.dense_h_to_4h": 64,
"transformer.h.20.self_attention.dense": 64,
"transformer.h.20.self_attention.query_key_value": 64,
"transformer.h.21.mlp.dense_4h_to_h": 64,
"transformer.h.21.mlp.dense_h_to_4h": 64,
"transformer.h.21.self_attention.dense": 64,
"transformer.h.21.self_attention.query_key_value": 64,
"transformer.h.22.mlp.dense_4h_to_h": 64,
"transformer.h.22.mlp.dense_h_to_4h": 64,
"transformer.h.22.self_attention.dense": 64,
"transformer.h.22.self_attention.query_key_value": 64,
"transformer.h.23.mlp.dense_4h_to_h": 64,
"transformer.h.23.mlp.dense_h_to_4h": 64,
"transformer.h.23.self_attention.dense": 64,
"transformer.h.23.self_attention.query_key_value": 64,
"transformer.h.3.mlp.dense_4h_to_h": 4,
"transformer.h.3.mlp.dense_h_to_4h": 4,
"transformer.h.3.self_attention.dense": 4,
"transformer.h.3.self_attention.query_key_value": 4,
"transformer.h.4.mlp.dense_4h_to_h": 4,
"transformer.h.4.mlp.dense_h_to_4h": 4,
"transformer.h.4.self_attention.dense": 4,
"transformer.h.4.self_attention.query_key_value": 4,
"transformer.h.5.mlp.dense_4h_to_h": 4,
"transformer.h.5.mlp.dense_h_to_4h": 4,
"transformer.h.5.self_attention.dense": 4,
"transformer.h.5.self_attention.query_key_value": 4,
"transformer.h.6.mlp.dense_4h_to_h": 4,
"transformer.h.6.mlp.dense_h_to_4h": 4,
"transformer.h.6.self_attention.dense": 4,
"transformer.h.6.self_attention.query_key_value": 4,
"transformer.h.7.mlp.dense_4h_to_h": 4,
"transformer.h.7.mlp.dense_h_to_4h": 4,
"transformer.h.7.self_attention.dense": 4,
"transformer.h.7.self_attention.query_key_value": 4,
"transformer.h.8.mlp.dense_4h_to_h": 4,
"transformer.h.8.mlp.dense_h_to_4h": 4,
"transformer.h.8.self_attention.dense": 4,
"transformer.h.8.self_attention.query_key_value": 4,
"transformer.h.9.mlp.dense_4h_to_h": 4,
"transformer.h.9.mlp.dense_h_to_4h": 4,
"transformer.h.9.self_attention.dense": 4,
"transformer.h.9.self_attention.query_key_value": 4
},
"revision": null,
"target_modules": [
"transformer.h.0.mlp.dense_h_to_4h",
"transformer.h.8.mlp.dense_h_to_4h",
"transformer.h.4.mlp.dense_4h_to_h",
"transformer.h.2.self_attention.dense",
"transformer.h.9.mlp.dense_4h_to_h",
"transformer.h.11.mlp.dense_4h_to_h",
"transformer.h.9.mlp.dense_h_to_4h",
"transformer.h.8.self_attention.query_key_value",
"transformer.h.6.self_attention.dense",
"transformer.h.3.self_attention.query_key_value",
"transformer.h.3.self_attention.dense",
"transformer.h.11.self_attention.dense",
"transformer.h.1.self_attention.query_key_value",
"transformer.h.5.mlp.dense_h_to_4h",
"transformer.h.2.self_attention.query_key_value",
"transformer.h.10.mlp.dense_h_to_4h",
"transformer.h.3.mlp.dense_4h_to_h",
"transformer.h.4.self_attention.query_key_value",
"transformer.h.10.mlp.dense_4h_to_h",
"transformer.h.4.mlp.dense_h_to_4h",
"transformer.h.7.self_attention.dense",
"transformer.h.0.self_attention.dense",
"transformer.h.7.mlp.dense_h_to_4h",
"transformer.h.2.mlp.dense_h_to_4h",
"transformer.h.8.self_attention.dense",
"transformer.h.11.mlp.dense_h_to_4h",
"transformer.h.7.mlp.dense_4h_to_h",
"transformer.h.7.self_attention.query_key_value",
"transformer.h.0.mlp.dense_4h_to_h",
"transformer.h.10.self_attention.query_key_value",
"transformer.h.11.self_attention.query_key_value",
"transformer.h.1.mlp.dense_4h_to_h",
"transformer.h.1.mlp.dense_h_to_4h",
"transformer.h.5.self_attention.dense",
"transformer.h.10.self_attention.dense",
"transformer.h.6.self_attention.query_key_value",
"transformer.h.6.mlp.dense_h_to_4h",
"transformer.h.5.self_attention.query_key_value",
"transformer.h.4.self_attention.dense",
"transformer.h.8.mlp.dense_4h_to_h",
"transformer.h.3.mlp.dense_h_to_4h",
"transformer.h.1.self_attention.dense",
"transformer.h.9.self_attention.query_key_value",
"transformer.h.9.self_attention.dense",
"transformer.h.0.self_attention.query_key_value",
"transformer.h.5.mlp.dense_4h_to_h",
"transformer.h.6.mlp.dense_4h_to_h",
"transformer.h.2.mlp.dense_4h_to_h"
],
"task_type": "CAUSAL_LM",
"use_dora": false,
"use_rslora": false
}