End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +85 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9208715596330275,
|
| 4 |
+
"eval_loss": 0.3048288822174072,
|
| 5 |
+
"eval_runtime": 5.4607,
|
| 6 |
+
"eval_samples": 872,
|
| 7 |
+
"eval_samples_per_second": 159.688,
|
| 8 |
+
"eval_steps_per_second": 5.128,
|
| 9 |
+
"train_loss": 0.11159764983808133,
|
| 10 |
+
"train_runtime": 5170.0912,
|
| 11 |
+
"train_samples": 67349,
|
| 12 |
+
"train_samples_per_second": 52.107,
|
| 13 |
+
"train_steps_per_second": 0.407
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9208715596330275,
|
| 4 |
+
"eval_loss": 0.3048288822174072,
|
| 5 |
+
"eval_runtime": 5.4607,
|
| 6 |
+
"eval_samples": 872,
|
| 7 |
+
"eval_samples_per_second": 159.688,
|
| 8 |
+
"eval_steps_per_second": 5.128
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.11159764983808133,
|
| 4 |
+
"train_runtime": 5170.0912,
|
| 5 |
+
"train_samples": 67349,
|
| 6 |
+
"train_samples_per_second": 52.107,
|
| 7 |
+
"train_steps_per_second": 0.407
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.998099762470309,
|
| 5 |
+
"global_step": 2104,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.95,
|
| 12 |
+
"learning_rate": 3.811787072243346e-05,
|
| 13 |
+
"loss": 0.2524,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_accuracy": 0.908256880733945,
|
| 19 |
+
"eval_loss": 0.21832765638828278,
|
| 20 |
+
"eval_runtime": 5.5594,
|
| 21 |
+
"eval_samples_per_second": 156.85,
|
| 22 |
+
"eval_steps_per_second": 5.036,
|
| 23 |
+
"step": 526
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.9,
|
| 27 |
+
"learning_rate": 2.6235741444866924e-05,
|
| 28 |
+
"loss": 0.1152,
|
| 29 |
+
"step": 1000
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 2.0,
|
| 33 |
+
"eval_accuracy": 0.8944954128440367,
|
| 34 |
+
"eval_loss": 0.27505654096603394,
|
| 35 |
+
"eval_runtime": 5.5334,
|
| 36 |
+
"eval_samples_per_second": 157.59,
|
| 37 |
+
"eval_steps_per_second": 5.06,
|
| 38 |
+
"step": 1052
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.85,
|
| 42 |
+
"learning_rate": 1.4353612167300381e-05,
|
| 43 |
+
"loss": 0.0647,
|
| 44 |
+
"step": 1500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.0,
|
| 48 |
+
"eval_accuracy": 0.9059633027522935,
|
| 49 |
+
"eval_loss": 0.3214015066623688,
|
| 50 |
+
"eval_runtime": 5.5423,
|
| 51 |
+
"eval_samples_per_second": 157.334,
|
| 52 |
+
"eval_steps_per_second": 5.052,
|
| 53 |
+
"step": 1578
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.8,
|
| 57 |
+
"learning_rate": 2.4714828897338406e-06,
|
| 58 |
+
"loss": 0.0322,
|
| 59 |
+
"step": 2000
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 4.0,
|
| 63 |
+
"eval_accuracy": 0.9208715596330275,
|
| 64 |
+
"eval_loss": 0.3048288822174072,
|
| 65 |
+
"eval_runtime": 5.5583,
|
| 66 |
+
"eval_samples_per_second": 156.883,
|
| 67 |
+
"eval_steps_per_second": 5.038,
|
| 68 |
+
"step": 2104
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.0,
|
| 72 |
+
"step": 2104,
|
| 73 |
+
"total_flos": 2.509461556171899e+17,
|
| 74 |
+
"train_loss": 0.11159764983808133,
|
| 75 |
+
"train_runtime": 5170.0912,
|
| 76 |
+
"train_samples_per_second": 52.107,
|
| 77 |
+
"train_steps_per_second": 0.407
|
| 78 |
+
}
|
| 79 |
+
],
|
| 80 |
+
"max_steps": 2104,
|
| 81 |
+
"num_train_epochs": 4,
|
| 82 |
+
"total_flos": 2.509461556171899e+17,
|
| 83 |
+
"trial_name": null,
|
| 84 |
+
"trial_params": null
|
| 85 |
+
}
|