49 lines
1.5 KiB
JSON
49 lines
1.5 KiB
JSON
|
{
|
||
|
|
"adapter_dir": "/NHNHOME/0226010080_A/BASE/jongbokwon/FLS/fsl/outputs/fedavg/8/2",
|
||
|
|
"base_model": "meta-llama/Meta-Llama-3-8B-Instruct",
|
||
|
|
"peft_type": "LORA",
|
||
|
|
"task_type": "CAUSAL_LM",
|
||
|
|
"lora_r": 16,
|
||
|
|
"lora_alpha": 16,
|
||
|
|
"target_modules": [
|
||
|
|
"up_proj",
|
||
|
|
"v_proj",
|
||
|
|
"gate_proj",
|
||
|
|
"q_proj",
|
||
|
|
"k_proj",
|
||
|
|
"o_proj",
|
||
|
|
"down_proj"
|
||
|
|
],
|
||
|
|
"merged_at_utc": "2026-05-22T05:24:31.213169+00:00",
|
||
|
|
"output_dir": "/NHNHOME/0226010080_A/BASE/jongbokwon/FLS/fsl/outputs/fedavg/8/merged",
|
||
|
|
"repo_id": "wvnvwn/Meta-Llama-3-8B-Instruct-fedavg-v0",
|
||
|
|
"private": false,
|
||
|
|
"revision": null,
|
||
|
|
"dtype": "bfloat16",
|
||
|
|
"max_shard_size": "5GB",
|
||
|
|
"training": {
|
||
|
|
"procedure": "Federated LoRA fine-tuning followed by adapter aggregation. The resulting PEFT LoRA adapter was merged into the base model for reproducible evaluation.",
|
||
|
|
"algorithm": "fedavg",
|
||
|
|
"training_data": "data_hetero_with_4_tasks",
|
||
|
|
"training_command": null,
|
||
|
|
"num_clients": 8,
|
||
|
|
"communication_round": "3",
|
||
|
|
"local_epochs": "3",
|
||
|
|
"local_batch_size": "256",
|
||
|
|
"local_micro_batch_size": "16",
|
||
|
|
"local_learning_rate": "0.0003"
|
||
|
|
},
|
||
|
|
"framework_versions": {
|
||
|
|
"python": "3.10.20",
|
||
|
|
"platform": "Linux-6.8.0-100-generic-x86_64-with-glibc2.39",
|
||
|
|
"torch": "2.11.0+cu130",
|
||
|
|
"cuda": "13.0",
|
||
|
|
"transformers": "4.57.1",
|
||
|
|
"peft": "0.19.1",
|
||
|
|
"accelerate": "1.13.0",
|
||
|
|
"huggingface_hub": "0.36.2",
|
||
|
|
"safetensors": "0.7.0",
|
||
|
|
"vllm": "0.20.2",
|
||
|
|
"wandb": "0.27.0"
|
||
|
|
}
|
||
|
|
}
|