End of training
Browse files- all_results.json +11 -11
- eval_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +660 -63
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.91044776119403,
|
| 3 |
+
"eval_accuracy": 0.9148148148148149,
|
| 4 |
+
"eval_loss": 0.30782097578048706,
|
| 5 |
+
"eval_runtime": 16.2443,
|
| 6 |
+
"eval_samples_per_second": 116.348,
|
| 7 |
+
"eval_steps_per_second": 3.694,
|
| 8 |
+
"total_flos": 1.0844463059730432e+19,
|
| 9 |
+
"train_loss": 0.6027212039533868,
|
| 10 |
+
"train_runtime": 2771.0404,
|
| 11 |
+
"train_samples_per_second": 38.65,
|
| 12 |
+
"train_steps_per_second": 0.3
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.91044776119403,
|
| 3 |
+
"eval_accuracy": 0.9148148148148149,
|
| 4 |
+
"eval_loss": 0.30782097578048706,
|
| 5 |
+
"eval_runtime": 16.2443,
|
| 6 |
+
"eval_samples_per_second": 116.348,
|
| 7 |
+
"eval_steps_per_second": 3.694
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.91044776119403,
|
| 3 |
+
"total_flos": 1.0844463059730432e+19,
|
| 4 |
+
"train_loss": 0.6027212039533868,
|
| 5 |
+
"train_runtime": 2771.0404,
|
| 6 |
+
"train_samples_per_second": 38.65,
|
| 7 |
+
"train_steps_per_second": 0.3
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,101 +1,698 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "dinov2-finetuned-har/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 2.
|
| 36 |
-
"loss":
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch":
|
| 41 |
-
"
|
| 42 |
-
"
|
| 43 |
-
"
|
| 44 |
-
"eval_samples_per_second": 49.239,
|
| 45 |
-
"eval_steps_per_second": 0.782,
|
| 46 |
-
"step": 42
|
| 47 |
-
},
|
| 48 |
-
{
|
| 49 |
-
"epoch": 1.1904761904761905,
|
| 50 |
-
"grad_norm": 23.17043685913086,
|
| 51 |
-
"learning_rate": 2.2666666666666668e-05,
|
| 52 |
-
"loss": 0.7596,
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch":
|
| 57 |
-
"grad_norm":
|
| 58 |
-
"learning_rate":
|
| 59 |
-
"loss": 0.
|
| 60 |
"step": 60
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch":
|
| 64 |
-
"grad_norm":
|
| 65 |
-
"learning_rate":
|
| 66 |
-
"loss": 0.
|
| 67 |
"step": 70
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"grad_norm":
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss": 0.
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch":
|
| 78 |
-
"eval_accuracy": 0.
|
| 79 |
-
"eval_loss": 0.
|
| 80 |
-
"eval_runtime":
|
| 81 |
-
"eval_samples_per_second":
|
| 82 |
-
"eval_steps_per_second":
|
| 83 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 84 |
},
|
| 85 |
{
|
| 86 |
-
"epoch":
|
| 87 |
-
"step":
|
| 88 |
-
"total_flos":
|
| 89 |
-
"train_loss": 0.
|
| 90 |
-
"train_runtime":
|
| 91 |
-
"train_samples_per_second":
|
| 92 |
-
"train_steps_per_second": 0.
|
| 93 |
}
|
| 94 |
],
|
| 95 |
"logging_steps": 10,
|
| 96 |
-
"max_steps":
|
| 97 |
"num_input_tokens_seen": 0,
|
| 98 |
-
"num_train_epochs":
|
| 99 |
"save_steps": 500,
|
| 100 |
"stateful_callbacks": {
|
| 101 |
"TrainerControl": {
|
|
@@ -109,8 +706,8 @@
|
|
| 109 |
"attributes": {}
|
| 110 |
}
|
| 111 |
},
|
| 112 |
-
"total_flos":
|
| 113 |
-
"train_batch_size":
|
| 114 |
"trial_name": null,
|
| 115 |
"trial_params": null
|
| 116 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9148148148148149,
|
| 3 |
+
"best_model_checkpoint": "dinov2-finetuned-har/checkpoint-830",
|
| 4 |
+
"epoch": 9.91044776119403,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 830,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.11940298507462686,
|
| 13 |
+
"grad_norm": 42.725677490234375,
|
| 14 |
+
"learning_rate": 6.024096385542169e-06,
|
| 15 |
+
"loss": 2.9994,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.23880597014925373,
|
| 20 |
+
"grad_norm": 40.75918197631836,
|
| 21 |
+
"learning_rate": 1.2048192771084338e-05,
|
| 22 |
+
"loss": 2.1467,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.3582089552238806,
|
| 27 |
+
"grad_norm": 36.34815979003906,
|
| 28 |
+
"learning_rate": 1.8072289156626505e-05,
|
| 29 |
+
"loss": 1.368,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.47761194029850745,
|
| 34 |
+
"grad_norm": 41.22068405151367,
|
| 35 |
+
"learning_rate": 2.4096385542168677e-05,
|
| 36 |
+
"loss": 1.0901,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.5970149253731343,
|
| 41 |
+
"grad_norm": 35.41105270385742,
|
| 42 |
+
"learning_rate": 3.012048192771085e-05,
|
| 43 |
+
"loss": 0.9148,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.7164179104477612,
|
| 48 |
+
"grad_norm": 45.00641632080078,
|
| 49 |
+
"learning_rate": 3.614457831325301e-05,
|
| 50 |
+
"loss": 0.8958,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.835820895522388,
|
| 55 |
+
"grad_norm": 32.627593994140625,
|
| 56 |
+
"learning_rate": 4.2168674698795186e-05,
|
| 57 |
+
"loss": 0.9745,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.9552238805970149,
|
| 62 |
+
"grad_norm": 44.3660888671875,
|
| 63 |
+
"learning_rate": 4.8192771084337354e-05,
|
| 64 |
+
"loss": 0.9429,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.991044776119403,
|
| 69 |
+
"eval_accuracy": 0.8328042328042328,
|
| 70 |
+
"eval_loss": 0.562383234500885,
|
| 71 |
+
"eval_runtime": 21.3212,
|
| 72 |
+
"eval_samples_per_second": 88.644,
|
| 73 |
+
"eval_steps_per_second": 2.814,
|
| 74 |
+
"step": 83
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0746268656716418,
|
| 78 |
+
"grad_norm": 66.64344024658203,
|
| 79 |
+
"learning_rate": 4.953145917001339e-05,
|
| 80 |
+
"loss": 0.8966,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.1940298507462686,
|
| 85 |
+
"grad_norm": 31.583904266357422,
|
| 86 |
+
"learning_rate": 4.886211512717537e-05,
|
| 87 |
+
"loss": 0.8292,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.3134328358208955,
|
| 92 |
+
"grad_norm": 70.04341888427734,
|
| 93 |
+
"learning_rate": 4.8192771084337354e-05,
|
| 94 |
+
"loss": 0.8166,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.4328358208955223,
|
| 99 |
+
"grad_norm": 42.080177307128906,
|
| 100 |
+
"learning_rate": 4.7523427041499336e-05,
|
| 101 |
+
"loss": 0.9254,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.5522388059701493,
|
| 106 |
+
"grad_norm": 123.05072784423828,
|
| 107 |
+
"learning_rate": 4.685408299866131e-05,
|
| 108 |
+
"loss": 0.9084,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.671641791044776,
|
| 113 |
+
"grad_norm": 25.725753784179688,
|
| 114 |
+
"learning_rate": 4.61847389558233e-05,
|
| 115 |
+
"loss": 0.8126,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.7910447761194028,
|
| 120 |
+
"grad_norm": 20.468259811401367,
|
| 121 |
+
"learning_rate": 4.5515394912985275e-05,
|
| 122 |
+
"loss": 0.8057,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.9104477611940298,
|
| 127 |
+
"grad_norm": 24.806772232055664,
|
| 128 |
+
"learning_rate": 4.484605087014726e-05,
|
| 129 |
+
"loss": 0.7912,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.9940298507462688,
|
| 134 |
+
"eval_accuracy": 0.8587301587301587,
|
| 135 |
+
"eval_loss": 0.47548335790634155,
|
| 136 |
+
"eval_runtime": 21.1591,
|
| 137 |
+
"eval_samples_per_second": 89.323,
|
| 138 |
+
"eval_steps_per_second": 2.836,
|
| 139 |
+
"step": 167
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.029850746268657,
|
| 143 |
+
"grad_norm": 26.82819175720215,
|
| 144 |
+
"learning_rate": 4.417670682730924e-05,
|
| 145 |
+
"loss": 0.7938,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.1492537313432836,
|
| 150 |
+
"grad_norm": 30.11927604675293,
|
| 151 |
+
"learning_rate": 4.350736278447122e-05,
|
| 152 |
+
"loss": 0.7185,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.2686567164179103,
|
| 157 |
+
"grad_norm": 32.696815490722656,
|
| 158 |
+
"learning_rate": 4.2838018741633203e-05,
|
| 159 |
+
"loss": 0.7642,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.388059701492537,
|
| 164 |
+
"grad_norm": 34.70363998413086,
|
| 165 |
+
"learning_rate": 4.2168674698795186e-05,
|
| 166 |
+
"loss": 0.7867,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.5074626865671643,
|
| 171 |
+
"grad_norm": 21.691709518432617,
|
| 172 |
+
"learning_rate": 4.149933065595716e-05,
|
| 173 |
+
"loss": 0.7376,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.626865671641791,
|
| 178 |
+
"grad_norm": 37.57550811767578,
|
| 179 |
+
"learning_rate": 4.082998661311915e-05,
|
| 180 |
+
"loss": 0.7446,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.746268656716418,
|
| 185 |
+
"grad_norm": 21.075855255126953,
|
| 186 |
+
"learning_rate": 4.0160642570281125e-05,
|
| 187 |
+
"loss": 0.6815,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.8656716417910446,
|
| 192 |
+
"grad_norm": 29.882211685180664,
|
| 193 |
+
"learning_rate": 3.949129852744311e-05,
|
| 194 |
+
"loss": 0.686,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.9850746268656714,
|
| 199 |
+
"grad_norm": 31.888534545898438,
|
| 200 |
+
"learning_rate": 3.882195448460509e-05,
|
| 201 |
+
"loss": 0.7371,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.997014925373134,
|
| 206 |
+
"eval_accuracy": 0.855026455026455,
|
| 207 |
+
"eval_loss": 0.4584466516971588,
|
| 208 |
+
"eval_runtime": 21.3468,
|
| 209 |
+
"eval_samples_per_second": 88.538,
|
| 210 |
+
"eval_steps_per_second": 2.811,
|
| 211 |
+
"step": 251
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.1044776119402986,
|
| 215 |
+
"grad_norm": 31.14760971069336,
|
| 216 |
+
"learning_rate": 3.815261044176707e-05,
|
| 217 |
+
"loss": 0.6986,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.2238805970149254,
|
| 222 |
+
"grad_norm": 20.720378875732422,
|
| 223 |
+
"learning_rate": 3.748326639892905e-05,
|
| 224 |
+
"loss": 0.6277,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.343283582089552,
|
| 229 |
+
"grad_norm": 22.004247665405273,
|
| 230 |
+
"learning_rate": 3.6813922356091035e-05,
|
| 231 |
+
"loss": 0.5936,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.4626865671641793,
|
| 236 |
+
"grad_norm": 24.986722946166992,
|
| 237 |
+
"learning_rate": 3.614457831325301e-05,
|
| 238 |
+
"loss": 0.5917,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.582089552238806,
|
| 243 |
+
"grad_norm": 27.637083053588867,
|
| 244 |
+
"learning_rate": 3.5475234270415e-05,
|
| 245 |
+
"loss": 0.5619,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.701492537313433,
|
| 250 |
+
"grad_norm": 37.6093864440918,
|
| 251 |
+
"learning_rate": 3.4805890227576974e-05,
|
| 252 |
+
"loss": 0.6538,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.8208955223880596,
|
| 257 |
+
"grad_norm": 23.496280670166016,
|
| 258 |
+
"learning_rate": 3.413654618473896e-05,
|
| 259 |
+
"loss": 0.6455,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.9402985074626864,
|
| 264 |
+
"grad_norm": 23.340776443481445,
|
| 265 |
+
"learning_rate": 3.346720214190094e-05,
|
| 266 |
+
"loss": 0.5915,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.0,
|
| 271 |
+
"eval_accuracy": 0.8761904761904762,
|
| 272 |
+
"eval_loss": 0.38699546456336975,
|
| 273 |
+
"eval_runtime": 21.0126,
|
| 274 |
+
"eval_samples_per_second": 89.946,
|
| 275 |
+
"eval_steps_per_second": 2.855,
|
| 276 |
+
"step": 335
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.059701492537314,
|
| 280 |
+
"grad_norm": 24.867502212524414,
|
| 281 |
+
"learning_rate": 3.279785809906292e-05,
|
| 282 |
+
"loss": 0.5275,
|
| 283 |
+
"step": 340
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.17910447761194,
|
| 287 |
+
"grad_norm": 22.04759407043457,
|
| 288 |
+
"learning_rate": 3.21285140562249e-05,
|
| 289 |
+
"loss": 0.5317,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.298507462686567,
|
| 294 |
+
"grad_norm": 23.60555076599121,
|
| 295 |
+
"learning_rate": 3.1459170013386885e-05,
|
| 296 |
+
"loss": 0.6104,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.417910447761194,
|
| 301 |
+
"grad_norm": 30.320343017578125,
|
| 302 |
+
"learning_rate": 3.078982597054887e-05,
|
| 303 |
+
"loss": 0.5362,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.537313432835821,
|
| 308 |
+
"grad_norm": 19.35739517211914,
|
| 309 |
+
"learning_rate": 3.012048192771085e-05,
|
| 310 |
+
"loss": 0.5487,
|
| 311 |
+
"step": 380
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.656716417910448,
|
| 315 |
+
"grad_norm": 27.066503524780273,
|
| 316 |
+
"learning_rate": 2.9451137884872827e-05,
|
| 317 |
+
"loss": 0.5296,
|
| 318 |
+
"step": 390
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.776119402985074,
|
| 322 |
+
"grad_norm": 35.0575065612793,
|
| 323 |
+
"learning_rate": 2.878179384203481e-05,
|
| 324 |
+
"loss": 0.5051,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.895522388059701,
|
| 329 |
+
"grad_norm": 19.96647071838379,
|
| 330 |
+
"learning_rate": 2.8112449799196788e-05,
|
| 331 |
+
"loss": 0.5635,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.991044776119403,
|
| 336 |
+
"eval_accuracy": 0.8703703703703703,
|
| 337 |
+
"eval_loss": 0.4037255644798279,
|
| 338 |
+
"eval_runtime": 21.0465,
|
| 339 |
+
"eval_samples_per_second": 89.801,
|
| 340 |
+
"eval_steps_per_second": 2.851,
|
| 341 |
+
"step": 418
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 5.014925373134329,
|
| 345 |
+
"grad_norm": 18.035507202148438,
|
| 346 |
+
"learning_rate": 2.7443105756358774e-05,
|
| 347 |
+
"loss": 0.5311,
|
| 348 |
+
"step": 420
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.134328358208955,
|
| 352 |
+
"grad_norm": 27.073444366455078,
|
| 353 |
+
"learning_rate": 2.6773761713520752e-05,
|
| 354 |
+
"loss": 0.4792,
|
| 355 |
+
"step": 430
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 5.253731343283582,
|
| 359 |
+
"grad_norm": 17.406234741210938,
|
| 360 |
+
"learning_rate": 2.6104417670682734e-05,
|
| 361 |
+
"loss": 0.5018,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.373134328358209,
|
| 366 |
+
"grad_norm": 19.9872989654541,
|
| 367 |
+
"learning_rate": 2.5435073627844713e-05,
|
| 368 |
+
"loss": 0.4636,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.492537313432836,
|
| 373 |
+
"grad_norm": 20.06911849975586,
|
| 374 |
+
"learning_rate": 2.4765729585006695e-05,
|
| 375 |
+
"loss": 0.4911,
|
| 376 |
+
"step": 460
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.611940298507463,
|
| 380 |
+
"grad_norm": 19.70343780517578,
|
| 381 |
+
"learning_rate": 2.4096385542168677e-05,
|
| 382 |
+
"loss": 0.4889,
|
| 383 |
+
"step": 470
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.731343283582089,
|
| 387 |
+
"grad_norm": 21.2330379486084,
|
| 388 |
+
"learning_rate": 2.3427041499330656e-05,
|
| 389 |
+
"loss": 0.4411,
|
| 390 |
+
"step": 480
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.850746268656716,
|
| 394 |
+
"grad_norm": 20.05991554260254,
|
| 395 |
+
"learning_rate": 2.2757697456492638e-05,
|
| 396 |
+
"loss": 0.4584,
|
| 397 |
+
"step": 490
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.970149253731344,
|
| 401 |
+
"grad_norm": 19.27924919128418,
|
| 402 |
+
"learning_rate": 2.208835341365462e-05,
|
| 403 |
+
"loss": 0.498,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.994029850746268,
|
| 408 |
+
"eval_accuracy": 0.8804232804232804,
|
| 409 |
+
"eval_loss": 0.3876107931137085,
|
| 410 |
+
"eval_runtime": 21.1814,
|
| 411 |
+
"eval_samples_per_second": 89.229,
|
| 412 |
+
"eval_steps_per_second": 2.833,
|
| 413 |
+
"step": 502
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 6.08955223880597,
|
| 417 |
+
"grad_norm": 23.8977108001709,
|
| 418 |
+
"learning_rate": 2.1419009370816602e-05,
|
| 419 |
+
"loss": 0.4212,
|
| 420 |
+
"step": 510
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 6.208955223880597,
|
| 424 |
+
"grad_norm": 25.704334259033203,
|
| 425 |
+
"learning_rate": 2.074966532797858e-05,
|
| 426 |
+
"loss": 0.4529,
|
| 427 |
+
"step": 520
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 6.3283582089552235,
|
| 431 |
+
"grad_norm": 24.296876907348633,
|
| 432 |
+
"learning_rate": 2.0080321285140562e-05,
|
| 433 |
+
"loss": 0.4707,
|
| 434 |
+
"step": 530
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.447761194029851,
|
| 438 |
+
"grad_norm": 17.248048782348633,
|
| 439 |
+
"learning_rate": 1.9410977242302544e-05,
|
| 440 |
+
"loss": 0.4178,
|
| 441 |
+
"step": 540
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.567164179104478,
|
| 445 |
+
"grad_norm": 18.4821834564209,
|
| 446 |
+
"learning_rate": 1.8741633199464527e-05,
|
| 447 |
+
"loss": 0.449,
|
| 448 |
+
"step": 550
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.686567164179104,
|
| 452 |
+
"grad_norm": 19.246505737304688,
|
| 453 |
+
"learning_rate": 1.8072289156626505e-05,
|
| 454 |
+
"loss": 0.4209,
|
| 455 |
+
"step": 560
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.8059701492537314,
|
| 459 |
+
"grad_norm": 20.905685424804688,
|
| 460 |
+
"learning_rate": 1.7402945113788487e-05,
|
| 461 |
+
"loss": 0.4344,
|
| 462 |
+
"step": 570
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.925373134328359,
|
| 466 |
+
"grad_norm": 16.702898025512695,
|
| 467 |
+
"learning_rate": 1.673360107095047e-05,
|
| 468 |
+
"loss": 0.4541,
|
| 469 |
+
"step": 580
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.997014925373135,
|
| 473 |
+
"eval_accuracy": 0.8883597883597883,
|
| 474 |
+
"eval_loss": 0.3611726760864258,
|
| 475 |
+
"eval_runtime": 21.6673,
|
| 476 |
+
"eval_samples_per_second": 87.228,
|
| 477 |
+
"eval_steps_per_second": 2.769,
|
| 478 |
+
"step": 586
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.044776119402985,
|
| 482 |
+
"grad_norm": 13.757901191711426,
|
| 483 |
+
"learning_rate": 1.606425702811245e-05,
|
| 484 |
+
"loss": 0.3419,
|
| 485 |
+
"step": 590
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.164179104477612,
|
| 489 |
+
"grad_norm": 20.324989318847656,
|
| 490 |
+
"learning_rate": 1.5394912985274433e-05,
|
| 491 |
+
"loss": 0.3985,
|
| 492 |
+
"step": 600
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.2835820895522385,
|
| 496 |
+
"grad_norm": 15.066502571105957,
|
| 497 |
+
"learning_rate": 1.4725568942436414e-05,
|
| 498 |
+
"loss": 0.3917,
|
| 499 |
+
"step": 610
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.402985074626866,
|
| 503 |
+
"grad_norm": 14.179686546325684,
|
| 504 |
+
"learning_rate": 1.4056224899598394e-05,
|
| 505 |
+
"loss": 0.372,
|
| 506 |
+
"step": 620
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.522388059701493,
|
| 510 |
+
"grad_norm": 18.657678604125977,
|
| 511 |
+
"learning_rate": 1.3386880856760376e-05,
|
| 512 |
+
"loss": 0.3744,
|
| 513 |
+
"step": 630
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.641791044776119,
|
| 517 |
+
"grad_norm": 20.829837799072266,
|
| 518 |
+
"learning_rate": 1.2717536813922356e-05,
|
| 519 |
+
"loss": 0.3809,
|
| 520 |
+
"step": 640
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.7611940298507465,
|
| 524 |
+
"grad_norm": 16.62371063232422,
|
| 525 |
+
"learning_rate": 1.2048192771084338e-05,
|
| 526 |
+
"loss": 0.3727,
|
| 527 |
+
"step": 650
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.880597014925373,
|
| 531 |
+
"grad_norm": 14.337791442871094,
|
| 532 |
+
"learning_rate": 1.1378848728246319e-05,
|
| 533 |
+
"loss": 0.3274,
|
| 534 |
+
"step": 660
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.0,
|
| 538 |
+
"grad_norm": 19.433561325073242,
|
| 539 |
+
"learning_rate": 1.0709504685408301e-05,
|
| 540 |
+
"loss": 0.3513,
|
| 541 |
+
"step": 670
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.0,
|
| 545 |
+
"eval_accuracy": 0.9052910052910053,
|
| 546 |
+
"eval_loss": 0.32395628094673157,
|
| 547 |
+
"eval_runtime": 21.5863,
|
| 548 |
+
"eval_samples_per_second": 87.555,
|
| 549 |
+
"eval_steps_per_second": 2.78,
|
| 550 |
+
"step": 670
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 8.119402985074627,
|
| 554 |
+
"grad_norm": 21.07122802734375,
|
| 555 |
+
"learning_rate": 1.0040160642570281e-05,
|
| 556 |
+
"loss": 0.3392,
|
| 557 |
+
"step": 680
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 8.238805970149254,
|
| 561 |
+
"grad_norm": 21.168922424316406,
|
| 562 |
+
"learning_rate": 9.370816599732263e-06,
|
| 563 |
+
"loss": 0.3038,
|
| 564 |
+
"step": 690
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 8.35820895522388,
|
| 568 |
+
"grad_norm": 17.61219024658203,
|
| 569 |
+
"learning_rate": 8.701472556894244e-06,
|
| 570 |
+
"loss": 0.3049,
|
| 571 |
+
"step": 700
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 8.477611940298507,
|
| 575 |
+
"grad_norm": 18.213272094726562,
|
| 576 |
+
"learning_rate": 8.032128514056226e-06,
|
| 577 |
+
"loss": 0.3423,
|
| 578 |
+
"step": 710
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 8.597014925373134,
|
| 582 |
+
"grad_norm": 16.038349151611328,
|
| 583 |
+
"learning_rate": 7.362784471218207e-06,
|
| 584 |
+
"loss": 0.2895,
|
| 585 |
+
"step": 720
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.716417910447761,
|
| 589 |
+
"grad_norm": 14.367926597595215,
|
| 590 |
+
"learning_rate": 6.693440428380188e-06,
|
| 591 |
+
"loss": 0.3142,
|
| 592 |
+
"step": 730
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 8.835820895522389,
|
| 596 |
+
"grad_norm": 16.960412979125977,
|
| 597 |
+
"learning_rate": 6.024096385542169e-06,
|
| 598 |
+
"loss": 0.3338,
|
| 599 |
+
"step": 740
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 8.955223880597014,
|
| 603 |
+
"grad_norm": 13.258671760559082,
|
| 604 |
+
"learning_rate": 5.3547523427041504e-06,
|
| 605 |
+
"loss": 0.2963,
|
| 606 |
+
"step": 750
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 8.991044776119402,
|
| 610 |
+
"eval_accuracy": 0.9116402116402117,
|
| 611 |
+
"eval_loss": 0.3176145553588867,
|
| 612 |
+
"eval_runtime": 21.2968,
|
| 613 |
+
"eval_samples_per_second": 88.746,
|
| 614 |
+
"eval_steps_per_second": 2.817,
|
| 615 |
+
"step": 753
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 9.074626865671641,
|
| 619 |
+
"grad_norm": 17.589509963989258,
|
| 620 |
+
"learning_rate": 4.685408299866132e-06,
|
| 621 |
+
"loss": 0.265,
|
| 622 |
+
"step": 760
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 9.194029850746269,
|
| 626 |
+
"grad_norm": 13.759172439575195,
|
| 627 |
+
"learning_rate": 4.016064257028113e-06,
|
| 628 |
+
"loss": 0.2718,
|
| 629 |
+
"step": 770
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 9.313432835820896,
|
| 633 |
+
"grad_norm": 13.576536178588867,
|
| 634 |
+
"learning_rate": 3.346720214190094e-06,
|
| 635 |
+
"loss": 0.2744,
|
| 636 |
+
"step": 780
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 9.432835820895523,
|
| 640 |
+
"grad_norm": 13.831005096435547,
|
| 641 |
+
"learning_rate": 2.6773761713520752e-06,
|
| 642 |
+
"loss": 0.2789,
|
| 643 |
+
"step": 790
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 9.552238805970148,
|
| 647 |
+
"grad_norm": 14.289161682128906,
|
| 648 |
+
"learning_rate": 2.0080321285140564e-06,
|
| 649 |
+
"loss": 0.2793,
|
| 650 |
+
"step": 800
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 9.671641791044776,
|
| 654 |
+
"grad_norm": 13.557323455810547,
|
| 655 |
+
"learning_rate": 1.3386880856760376e-06,
|
| 656 |
+
"loss": 0.2893,
|
| 657 |
+
"step": 810
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 9.791044776119403,
|
| 661 |
+
"grad_norm": 13.76667308807373,
|
| 662 |
+
"learning_rate": 6.693440428380188e-07,
|
| 663 |
+
"loss": 0.2918,
|
| 664 |
+
"step": 820
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 9.91044776119403,
|
| 668 |
+
"grad_norm": 15.072478294372559,
|
| 669 |
+
"learning_rate": 0.0,
|
| 670 |
+
"loss": 0.2815,
|
| 671 |
+
"step": 830
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 9.91044776119403,
|
| 675 |
+
"eval_accuracy": 0.9148148148148149,
|
| 676 |
+
"eval_loss": 0.30782097578048706,
|
| 677 |
+
"eval_runtime": 22.2064,
|
| 678 |
+
"eval_samples_per_second": 85.111,
|
| 679 |
+
"eval_steps_per_second": 2.702,
|
| 680 |
+
"step": 830
|
| 681 |
},
|
| 682 |
{
|
| 683 |
+
"epoch": 9.91044776119403,
|
| 684 |
+
"step": 830,
|
| 685 |
+
"total_flos": 1.0844463059730432e+19,
|
| 686 |
+
"train_loss": 0.6027212039533868,
|
| 687 |
+
"train_runtime": 2771.0404,
|
| 688 |
+
"train_samples_per_second": 38.65,
|
| 689 |
+
"train_steps_per_second": 0.3
|
| 690 |
}
|
| 691 |
],
|
| 692 |
"logging_steps": 10,
|
| 693 |
+
"max_steps": 830,
|
| 694 |
"num_input_tokens_seen": 0,
|
| 695 |
+
"num_train_epochs": 10,
|
| 696 |
"save_steps": 500,
|
| 697 |
"stateful_callbacks": {
|
| 698 |
"TrainerControl": {
|
|
|
|
| 706 |
"attributes": {}
|
| 707 |
}
|
| 708 |
},
|
| 709 |
+
"total_flos": 1.0844463059730432e+19,
|
| 710 |
+
"train_batch_size": 32,
|
| 711 |
"trial_name": null,
|
| 712 |
"trial_params": null
|
| 713 |
}
|