Commit
•
2f6f192
1
Parent(s):
fcd73b6
End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +61 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.23,
|
| 3 |
+
"eval_accuracy": 0.986989308918276,
|
| 4 |
+
"eval_loss": 0.0699462890625,
|
| 5 |
+
"eval_runtime": 192.7346,
|
| 6 |
+
"eval_samples": 11583,
|
| 7 |
+
"eval_samples_per_second": 60.098,
|
| 8 |
+
"eval_steps_per_second": 1.878,
|
| 9 |
+
"perplexity": 1.0724505773813178,
|
| 10 |
+
"train_loss": 0.12210027694702148,
|
| 11 |
+
"train_runtime": 1976.8574,
|
| 12 |
+
"train_samples": 220074,
|
| 13 |
+
"train_samples_per_second": 25.9,
|
| 14 |
+
"train_steps_per_second": 0.202
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.23,
|
| 3 |
+
"eval_accuracy": 0.986989308918276,
|
| 4 |
+
"eval_loss": 0.0699462890625,
|
| 5 |
+
"eval_runtime": 192.7346,
|
| 6 |
+
"eval_samples": 11583,
|
| 7 |
+
"eval_samples_per_second": 60.098,
|
| 8 |
+
"eval_steps_per_second": 1.878,
|
| 9 |
+
"perplexity": 1.0724505773813178
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.23,
|
| 3 |
+
"train_loss": 0.12210027694702148,
|
| 4 |
+
"train_runtime": 1976.8574,
|
| 5 |
+
"train_samples": 220074,
|
| 6 |
+
"train_samples_per_second": 25.9,
|
| 7 |
+
"train_steps_per_second": 0.202
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.23262576330328583,
|
| 5 |
+
"global_step": 400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"eval_accuracy": 0.9866142691167129,
|
| 13 |
+
"eval_loss": 0.07293701171875,
|
| 14 |
+
"eval_runtime": 192.7185,
|
| 15 |
+
"eval_samples_per_second": 60.103,
|
| 16 |
+
"eval_steps_per_second": 1.878,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.12,
|
| 21 |
+
"eval_accuracy": 0.9868109877885048,
|
| 22 |
+
"eval_loss": 0.07159423828125,
|
| 23 |
+
"eval_runtime": 192.6928,
|
| 24 |
+
"eval_samples_per_second": 60.111,
|
| 25 |
+
"eval_steps_per_second": 1.879,
|
| 26 |
+
"step": 200
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.17,
|
| 30 |
+
"eval_accuracy": 0.9869265209767002,
|
| 31 |
+
"eval_loss": 0.07049560546875,
|
| 32 |
+
"eval_runtime": 193.5043,
|
| 33 |
+
"eval_samples_per_second": 59.859,
|
| 34 |
+
"eval_steps_per_second": 1.871,
|
| 35 |
+
"step": 300
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.23,
|
| 39 |
+
"eval_accuracy": 0.986989308918276,
|
| 40 |
+
"eval_loss": 0.0699462890625,
|
| 41 |
+
"eval_runtime": 192.6707,
|
| 42 |
+
"eval_samples_per_second": 60.118,
|
| 43 |
+
"eval_steps_per_second": 1.879,
|
| 44 |
+
"step": 400
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.23,
|
| 48 |
+
"step": 400,
|
| 49 |
+
"total_flos": 4.12601633927594e+17,
|
| 50 |
+
"train_loss": 0.12210027694702148,
|
| 51 |
+
"train_runtime": 1976.8574,
|
| 52 |
+
"train_samples_per_second": 25.9,
|
| 53 |
+
"train_steps_per_second": 0.202
|
| 54 |
+
}
|
| 55 |
+
],
|
| 56 |
+
"max_steps": 400,
|
| 57 |
+
"num_train_epochs": 1,
|
| 58 |
+
"total_flos": 4.12601633927594e+17,
|
| 59 |
+
"trial_name": null,
|
| 60 |
+
"trial_params": null
|
| 61 |
+
}
|