308 lines
6.7 KiB
JSON
308 lines
6.7 KiB
JSON
|
|
{
|
||
|
|
"run_name": "pilot-2k-v2",
|
||
|
|
"finetune_name": "slm-125m-sft-pilot-2k-e3",
|
||
|
|
"base_model_repo": "thesreedath/slm-125m-base",
|
||
|
|
"output_dir": "/data/sft/thesreedath-slm-125m/finetunes/slm-125m-sft-pilot-2k-e3",
|
||
|
|
"epochs": 3,
|
||
|
|
"optimizer_steps": 339,
|
||
|
|
"train_examples": 1800,
|
||
|
|
"val_examples": 100,
|
||
|
|
"train_tokens_per_epoch": 1262981,
|
||
|
|
"train_assistant_tokens_per_epoch": 62190,
|
||
|
|
"tokens_seen": 3788943,
|
||
|
|
"assistant_tokens_seen": 186570,
|
||
|
|
"initial_val_loss": 3.6076530987264204,
|
||
|
|
"initial_val_ppl": 36.87939886230172,
|
||
|
|
"final_val_loss": 1.5027814324426207,
|
||
|
|
"final_val_ppl": 4.494171937794634,
|
||
|
|
"elapsed_seconds": 160.08841013908386,
|
||
|
|
"estimated_gpu_usd": 0.03553962705087662,
|
||
|
|
"history": [
|
||
|
|
{
|
||
|
|
"step": 0,
|
||
|
|
"event": "initial_eval",
|
||
|
|
"val_loss": 3.6076530987264204,
|
||
|
|
"val_ppl": 36.87939886230172
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 1,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 3.3445167541503906,
|
||
|
|
"lr": 3.125e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 10,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 3.2654700875282288,
|
||
|
|
"lr": 3.125e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 20,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 2.046467971801758,
|
||
|
|
"lr": 4.9981082231081334e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 30,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.6243530839681626,
|
||
|
|
"lr": 4.976858599798299e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 40,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.5121346324682237,
|
||
|
|
"lr": 4.932196167813559e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 50,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.4252592638134955,
|
||
|
|
"lr": 4.864543104251587e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 50,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 1,
|
||
|
|
"val_loss": 1.6063852179899305,
|
||
|
|
"val_ppl": 4.9847598014490835
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 60,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.4405100345611572,
|
||
|
|
"lr": 4.774538907965318e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 70,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.3378228411078452,
|
||
|
|
"lr": 4.663034354621863e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 80,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.404534536600113,
|
||
|
|
"lr": 4.531083454641531e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 90,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.3485690742731093,
|
||
|
|
"lr": 4.3799334900354815e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 100,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.5101487711071968,
|
||
|
|
"lr": 4.211013224320062e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 100,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 1,
|
||
|
|
"val_loss": 1.5369550711230229,
|
||
|
|
"val_ppl": 4.650408527530095
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 110,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 1,
|
||
|
|
"train_loss": 1.3192693248391152,
|
||
|
|
"lr": 4.025919396955019e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 120,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.2987151977262998,
|
||
|
|
"lr": 3.826401629968575e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 130,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.1192938029766082,
|
||
|
|
"lr": 3.6143458894413465e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 140,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.1743364408612251,
|
||
|
|
"lr": 3.3917566581814276e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 150,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.1988286420702934,
|
||
|
|
"lr": 3.160737988105633e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 150,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 2,
|
||
|
|
"val_loss": 1.5214578169787263,
|
||
|
|
"val_ppl": 4.578895523539861
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 160,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.192124105989933,
|
||
|
|
"lr": 2.9234736114315616e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 170,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.1844297125935555,
|
||
|
|
"lr": 2.6822062986818796e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 180,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.2872302904725075,
|
||
|
|
"lr": 2.4392166586218103e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 190,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.18301559984684,
|
||
|
|
"lr": 2.1968015805260062e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 200,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.217595275491476,
|
||
|
|
"lr": 1.9572525225519227e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 200,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 2,
|
||
|
|
"val_loss": 1.5065097094689361,
|
||
|
|
"val_ppl": 4.51095872924231
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 210,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.2071400985121727,
|
||
|
|
"lr": 1.7228338514514846e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 220,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 2,
|
||
|
|
"train_loss": 1.2201956912875176,
|
||
|
|
"lr": 1.495761438367577e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 230,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.162185350531026,
|
||
|
|
"lr": 1.2781817130412088e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 240,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.2239526972174644,
|
||
|
|
"lr": 1.0721513744220169e-05
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 250,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.0888006776571273,
|
||
|
|
"lr": 8.796179494700441e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 250,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 3,
|
||
|
|
"val_loss": 1.5030882143383795,
|
||
|
|
"val_ppl": 4.495550879887996
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 260,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.166143298894167,
|
||
|
|
"lr": 7.02401383919088e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 270,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.1675694316625596,
|
||
|
|
"lr": 5.421768390171933e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 280,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.1555261366069316,
|
||
|
|
"lr": 4.004588568601897e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 290,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.0684210263192653,
|
||
|
|
"lr": 2.7858704399739895e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 300,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.1304051853716373,
|
||
|
|
"lr": 1.7771340863698306e-06
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 300,
|
||
|
|
"event": "eval",
|
||
|
|
"epoch": 3,
|
||
|
|
"val_loss": 1.5036853406266901,
|
||
|
|
"val_ppl": 4.498236093124735
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 310,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.17097322717309,
|
||
|
|
"lr": 9.879147114753545e-07
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 320,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.0470945492386818,
|
||
|
|
"lr": 4.256725079024554e-07
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"step": 330,
|
||
|
|
"event": "train_log",
|
||
|
|
"epoch": 3,
|
||
|
|
"train_loss": 1.1463549926877021,
|
||
|
|
"lr": 9.572213880643111e-08
|
||
|
|
}
|
||
|
|
]
|
||
|
|
}
|