coeiico's picture
Add every run in its own folder and the model card
4a570eb verified
Raw History Blame Contribute Delete
6.05 kB
{
"config": {
"run_name": "default-seed0",
"seed": 0,
"quick": false,
"device": "auto",
"eval_every": 250,
"task": {
"n_values": 64,
"length": 16
},
"splits": {
"val_seed": 1000003,
"val_size": 1024,
"test_seed": 2000003,
"test_size": 2048
},
"model": {
"n_layers": 4,
"d_model": 256,
"n_heads": 8,
"d_mlp": 1024,
"mlp": "gelu",
"norm": "layernorm",
"dropout": 0.0
},
"optim": {
"steps": 4000,
"batch_size": 256,
"lr": 0.001,
"warmup_steps": 100,
"min_lr_ratio": 0.1,
"weight_decay": 0.1,
"grad_clip": 1.0
}
},
"model_config": {
"vocab_size": 65,
"n_layers": 4,
"d_model": 256,
"n_heads": 8,
"d_mlp": 1024,
"max_seq_len": 33,
"mlp": "gelu",
"norm": "layernorm",
"dropout": 0.0
},
"parameters": 3201280,
"data": {
"task": "sorting",
"version": 1,
"n_values": 64,
"length": 16,
"val_seed": 1000003,
"val_size": 1024,
"test_seed": 2000003,
"test_size": 2048,
"train_seed": 0
},
"code": {
"commit": "19276f75392bee9d5f6f61820395002a0746e349",
"dirty": true,
"snapshot": "source-snapshot.tar.gz"
},
"environment": {
"python": "3.12.3",
"torch": "2.15.0+cu132",
"device": "cuda",
"precision": "torch.float32",
"uv_lock_sha256": "60eba889e8dcd90d00fdcc282a51fc42df3869f3ef3807b1946aaa3cfd5f62e2",
"gpu": "NVIDIA GeForce RTX 3060",
"cuda": "13.2"
},
"resumable": false,
"status": "finished",
"seconds": 646.7,
"history": [
{
"step": 250,
"train_loss": 0.07291533797979355,
"val_loss": 0.0985519140958786,
"val_token_accuracy": 0.9683837890625,
"val_exact_match": 0.595703125,
"grad_norm": 2.3497650623321533,
"lr": 0.0009967189933441244
},
{
"step": 500,
"train_loss": 0.014207392930984497,
"val_loss": 0.031683966517448425,
"val_token_accuracy": 0.99053955078125,
"val_exact_match": 0.8720703125,
"grad_norm": 0.7331729531288147,
"lr": 0.0009768413988762155
},
{
"step": 750,
"train_loss": 0.02052781730890274,
"val_loss": 0.023844445124268532,
"val_token_accuracy": 0.99273681640625,
"val_exact_match": 0.890625,
"grad_norm": 1.3566714525222778,
"lr": 0.0009397114317029974
},
{
"step": 1000,
"train_loss": 0.009082821197807789,
"val_loss": 0.01231020875275135,
"val_token_accuracy": 0.996826171875,
"val_exact_match": 0.951171875,
"grad_norm": 0.7233428359031677,
"lr": 0.0008868298366769955
},
{
"step": 1250,
"train_loss": 0.00922011211514473,
"val_loss": 0.019146941602230072,
"val_token_accuracy": 0.994873046875,
"val_exact_match": 0.9208984375,
"grad_norm": 0.5756050944328308,
"lr": 0.0008203340189070906
},
{
"step": 1500,
"train_loss": 0.01364622451364994,
"val_loss": 0.005576150491833687,
"val_token_accuracy": 0.998779296875,
"val_exact_match": 0.9814453125,
"grad_norm": 1.3124696016311646,
"lr": 0.0007429116526313744
},
{
"step": 1750,
"train_loss": 0.003266311716288328,
"val_loss": 0.001853839261457324,
"val_token_accuracy": 0.9996337890625,
"val_exact_match": 0.994140625,
"grad_norm": 0.28147438168525696,
"lr": 0.0006576920489294011
},
{
"step": 2000,
"train_loss": 0.005052485503256321,
"val_loss": 0.0006970376125536859,
"val_token_accuracy": 0.9998779296875,
"val_exact_match": 0.998046875,
"grad_norm": 1.1249206066131592,
"lr": 0.0005681196730492369
},
{
"step": 2250,
"train_loss": 0.00020874346955679357,
"val_loss": 0.0009217980550602078,
"val_token_accuracy": 0.99981689453125,
"val_exact_match": 0.9970703125,
"grad_norm": 0.03951434791088104,
"lr": 0.00047781492361400794
},
{
"step": 2500,
"train_loss": 0.001243826001882553,
"val_loss": 0.0005396516062319279,
"val_token_accuracy": 0.99981689453125,
"val_exact_match": 0.9970703125,
"grad_norm": 0.23733462393283844,
"lr": 0.0003904278008308589
},
{
"step": 2750,
"train_loss": 0.00022199669911060482,
"val_loss": 0.00017582655709702522,
"val_token_accuracy": 0.99993896484375,
"val_exact_match": 0.9990234375,
"grad_norm": 0.11565989255905151,
"lr": 0.0003094903782424895
},
{
"step": 3000,
"train_loss": 0.00021956569980829954,
"val_loss": 8.435545896645635e-05,
"val_token_accuracy": 1.0,
"val_exact_match": 1.0,
"grad_norm": 0.061819761991500854,
"lr": 0.0002382740409206803
},
{
"step": 3250,
"train_loss": 3.6058951081940904e-05,
"val_loss": 8.97766076377593e-05,
"val_token_accuracy": 0.99993896484375,
"val_exact_match": 0.9990234375,
"grad_norm": 0.004151737317442894,
"lr": 0.00017965726034785466
},
{
"step": 3500,
"train_loss": 4.75610613648314e-06,
"val_loss": 9.887960004562046e-06,
"val_token_accuracy": 1.0,
"val_exact_match": 1.0,
"grad_norm": 0.0010888022370636463,
"lr": 0.00013600925035352912
},
{
"step": 3750,
"train_loss": 0.00010177387594012544,
"val_loss": 0.00023778356262482703,
"val_token_accuracy": 0.99993896484375,
"val_exact_match": 0.9990234375,
"grad_norm": 0.09358467906713486,
"lr": 0.00010909420658097956
},
{
"step": 4000,
"train_loss": 1.727779454085976e-05,
"val_loss": 1.847820749389939e-05,
"val_token_accuracy": 1.0,
"val_exact_match": 1.0,
"grad_norm": 0.014979858882725239,
"lr": 0.0001
}
],
"final": {
"val_loss": 1.847820749389939e-05,
"val_token_accuracy": 1.0,
"val_exact_match": 1.0
}
}