🍻 cheers
Browse files- README.md +7 -6
- all_results.json +16 -0
- eval_results.json +11 -0
- runs/Mar21_16-19-41_45242e0ea5c0/events.out.tfevents.1711041100.45242e0ea5c0.20273.24 +3 -0
- train_results.json +8 -0
- trainer_state.json +505 -0
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -18,13 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# vit-weight-decay-1e-4
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
-
- Accuracy: 0.
|
| 25 |
-
- Precision: 0.
|
| 26 |
-
- Recall: 0.
|
| 27 |
-
- F1: 0.
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
|
|
| 19 |
|
| 20 |
# vit-weight-decay-1e-4
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the skin-cancer dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.5277
|
| 25 |
+
- Accuracy: 0.8263
|
| 26 |
+
- Precision: 0.8467
|
| 27 |
+
- Recall: 0.8263
|
| 28 |
+
- F1: 0.8324
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 25.0,
|
| 3 |
+
"eval_accuracy": 0.8262829403606102,
|
| 4 |
+
"eval_f1": 0.832391972746498,
|
| 5 |
+
"eval_loss": 0.5276548266410828,
|
| 6 |
+
"eval_precision": 0.8466524464455557,
|
| 7 |
+
"eval_recall": 0.8262829403606102,
|
| 8 |
+
"eval_runtime": 22.57,
|
| 9 |
+
"eval_samples_per_second": 127.78,
|
| 10 |
+
"eval_steps_per_second": 15.995,
|
| 11 |
+
"total_flos": 9.934918266289766e+18,
|
| 12 |
+
"train_loss": 0.5703618476843908,
|
| 13 |
+
"train_runtime": 3054.4722,
|
| 14 |
+
"train_samples_per_second": 167.885,
|
| 15 |
+
"train_steps_per_second": 10.509
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 25.0,
|
| 3 |
+
"eval_accuracy": 0.8262829403606102,
|
| 4 |
+
"eval_f1": 0.832391972746498,
|
| 5 |
+
"eval_loss": 0.5276548266410828,
|
| 6 |
+
"eval_precision": 0.8466524464455557,
|
| 7 |
+
"eval_recall": 0.8262829403606102,
|
| 8 |
+
"eval_runtime": 22.57,
|
| 9 |
+
"eval_samples_per_second": 127.78,
|
| 10 |
+
"eval_steps_per_second": 15.995
|
| 11 |
+
}
|
runs/Mar21_16-19-41_45242e0ea5c0/events.out.tfevents.1711041100.45242e0ea5c0.20273.24
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7eaf187cc62d414cf5f53bde73b2062acebd76605dcf338e4f70fc90d0492637
|
| 3 |
+
size 560
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 25.0,
|
| 3 |
+
"total_flos": 9.934918266289766e+18,
|
| 4 |
+
"train_loss": 0.5703618476843908,
|
| 5 |
+
"train_runtime": 3054.4722,
|
| 6 |
+
"train_samples_per_second": 167.885,
|
| 7 |
+
"train_steps_per_second": 10.509
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,505 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5276548266410828,
|
| 3 |
+
"best_model_checkpoint": "./vit-weight-decay-1e-4/checkpoint-4815",
|
| 4 |
+
"epoch": 25.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8025,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"grad_norm": 25.431089401245117,
|
| 14 |
+
"learning_rate": 1.8349682631275247e-05,
|
| 15 |
+
"loss": 1.7856,
|
| 16 |
+
"step": 321
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.6868932038834952,
|
| 21 |
+
"eval_f1": 0.6229244062646352,
|
| 22 |
+
"eval_loss": 0.9535064697265625,
|
| 23 |
+
"eval_precision": 0.6411826114675676,
|
| 24 |
+
"eval_recall": 0.6868932038834952,
|
| 25 |
+
"eval_runtime": 25.4898,
|
| 26 |
+
"eval_samples_per_second": 113.143,
|
| 27 |
+
"eval_steps_per_second": 14.163,
|
| 28 |
+
"step": 321
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 2.0,
|
| 32 |
+
"grad_norm": 61.20114517211914,
|
| 33 |
+
"learning_rate": 3.681477207155222e-05,
|
| 34 |
+
"loss": 1.1412,
|
| 35 |
+
"step": 642
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 2.0,
|
| 39 |
+
"eval_accuracy": 0.742371705963939,
|
| 40 |
+
"eval_f1": 0.7310806521188435,
|
| 41 |
+
"eval_loss": 0.6927886605262756,
|
| 42 |
+
"eval_precision": 0.743980694623501,
|
| 43 |
+
"eval_recall": 0.742371705963939,
|
| 44 |
+
"eval_runtime": 25.3401,
|
| 45 |
+
"eval_samples_per_second": 113.812,
|
| 46 |
+
"eval_steps_per_second": 14.246,
|
| 47 |
+
"step": 642
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.0,
|
| 51 |
+
"grad_norm": 5.373477458953857,
|
| 52 |
+
"learning_rate": 5.5337564916330066e-05,
|
| 53 |
+
"loss": 1.0297,
|
| 54 |
+
"step": 963
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 3.0,
|
| 58 |
+
"eval_accuracy": 0.7489597780859917,
|
| 59 |
+
"eval_f1": 0.7057462869226876,
|
| 60 |
+
"eval_loss": 0.6862976551055908,
|
| 61 |
+
"eval_precision": 0.7362120996210428,
|
| 62 |
+
"eval_recall": 0.7489597780859917,
|
| 63 |
+
"eval_runtime": 25.4318,
|
| 64 |
+
"eval_samples_per_second": 113.401,
|
| 65 |
+
"eval_steps_per_second": 14.195,
|
| 66 |
+
"step": 963
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.0,
|
| 70 |
+
"grad_norm": 37.48583221435547,
|
| 71 |
+
"learning_rate": 7.380265435660705e-05,
|
| 72 |
+
"loss": 0.9571,
|
| 73 |
+
"step": 1284
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.0,
|
| 77 |
+
"eval_accuracy": 0.7694174757281553,
|
| 78 |
+
"eval_f1": 0.773635549987236,
|
| 79 |
+
"eval_loss": 0.5587358474731445,
|
| 80 |
+
"eval_precision": 0.7900997688719571,
|
| 81 |
+
"eval_recall": 0.7694174757281553,
|
| 82 |
+
"eval_runtime": 25.2721,
|
| 83 |
+
"eval_samples_per_second": 114.118,
|
| 84 |
+
"eval_steps_per_second": 14.285,
|
| 85 |
+
"step": 1284
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 5.0,
|
| 89 |
+
"grad_norm": 4.737943649291992,
|
| 90 |
+
"learning_rate": 9.232544720138489e-05,
|
| 91 |
+
"loss": 0.9346,
|
| 92 |
+
"step": 1605
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 5.0,
|
| 96 |
+
"eval_accuracy": 0.7940360610263523,
|
| 97 |
+
"eval_f1": 0.7919455286020083,
|
| 98 |
+
"eval_loss": 0.5654485821723938,
|
| 99 |
+
"eval_precision": 0.8057986798329398,
|
| 100 |
+
"eval_recall": 0.7940360610263523,
|
| 101 |
+
"eval_runtime": 24.9343,
|
| 102 |
+
"eval_samples_per_second": 115.664,
|
| 103 |
+
"eval_steps_per_second": 14.478,
|
| 104 |
+
"step": 1605
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 6.0,
|
| 108 |
+
"grad_norm": 8.350205421447754,
|
| 109 |
+
"learning_rate": 9.984876489938473e-05,
|
| 110 |
+
"loss": 0.9802,
|
| 111 |
+
"step": 1926
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 6.0,
|
| 115 |
+
"eval_accuracy": 0.7746185852981969,
|
| 116 |
+
"eval_f1": 0.779379969410103,
|
| 117 |
+
"eval_loss": 0.6317604184150696,
|
| 118 |
+
"eval_precision": 0.7928135059188427,
|
| 119 |
+
"eval_recall": 0.7746185852981969,
|
| 120 |
+
"eval_runtime": 24.8108,
|
| 121 |
+
"eval_samples_per_second": 116.239,
|
| 122 |
+
"eval_steps_per_second": 14.55,
|
| 123 |
+
"step": 1926
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 7.0,
|
| 127 |
+
"grad_norm": 9.141142845153809,
|
| 128 |
+
"learning_rate": 9.889494151200358e-05,
|
| 129 |
+
"loss": 0.8352,
|
| 130 |
+
"step": 2247
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 7.0,
|
| 134 |
+
"eval_accuracy": 0.7295423023578363,
|
| 135 |
+
"eval_f1": 0.7497746149490648,
|
| 136 |
+
"eval_loss": 0.6611436009407043,
|
| 137 |
+
"eval_precision": 0.8144719729782962,
|
| 138 |
+
"eval_recall": 0.7295423023578363,
|
| 139 |
+
"eval_runtime": 24.9184,
|
| 140 |
+
"eval_samples_per_second": 115.738,
|
| 141 |
+
"eval_steps_per_second": 14.487,
|
| 142 |
+
"step": 2247
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 8.0,
|
| 146 |
+
"grad_norm": 2.7915797233581543,
|
| 147 |
+
"learning_rate": 9.707962612088379e-05,
|
| 148 |
+
"loss": 0.7621,
|
| 149 |
+
"step": 2568
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 8.0,
|
| 153 |
+
"eval_accuracy": 0.7666435506241331,
|
| 154 |
+
"eval_f1": 0.7780963506303339,
|
| 155 |
+
"eval_loss": 0.5766130685806274,
|
| 156 |
+
"eval_precision": 0.8161643971869461,
|
| 157 |
+
"eval_recall": 0.7666435506241331,
|
| 158 |
+
"eval_runtime": 25.8655,
|
| 159 |
+
"eval_samples_per_second": 111.5,
|
| 160 |
+
"eval_steps_per_second": 13.957,
|
| 161 |
+
"step": 2568
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 9.0,
|
| 165 |
+
"grad_norm": 8.544121742248535,
|
| 166 |
+
"learning_rate": 9.443480321450928e-05,
|
| 167 |
+
"loss": 0.7352,
|
| 168 |
+
"step": 2889
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 9.0,
|
| 172 |
+
"eval_accuracy": 0.7995839112343966,
|
| 173 |
+
"eval_f1": 0.8078907785166459,
|
| 174 |
+
"eval_loss": 0.5368876457214355,
|
| 175 |
+
"eval_precision": 0.8268758324015973,
|
| 176 |
+
"eval_recall": 0.7995839112343966,
|
| 177 |
+
"eval_runtime": 25.7681,
|
| 178 |
+
"eval_samples_per_second": 111.921,
|
| 179 |
+
"eval_steps_per_second": 14.01,
|
| 180 |
+
"step": 2889
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 10.0,
|
| 184 |
+
"grad_norm": 14.100479125976562,
|
| 185 |
+
"learning_rate": 9.100707257835249e-05,
|
| 186 |
+
"loss": 0.6919,
|
| 187 |
+
"step": 3210
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 10.0,
|
| 191 |
+
"eval_accuracy": 0.7753120665742025,
|
| 192 |
+
"eval_f1": 0.790016007945942,
|
| 193 |
+
"eval_loss": 0.5499810576438904,
|
| 194 |
+
"eval_precision": 0.8270227193923745,
|
| 195 |
+
"eval_recall": 0.7753120665742025,
|
| 196 |
+
"eval_runtime": 25.3767,
|
| 197 |
+
"eval_samples_per_second": 113.648,
|
| 198 |
+
"eval_steps_per_second": 14.226,
|
| 199 |
+
"step": 3210
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 11.0,
|
| 203 |
+
"grad_norm": 22.55378532409668,
|
| 204 |
+
"learning_rate": 8.685682824178951e-05,
|
| 205 |
+
"loss": 0.6105,
|
| 206 |
+
"step": 3531
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 11.0,
|
| 210 |
+
"eval_accuracy": 0.8061719833564494,
|
| 211 |
+
"eval_f1": 0.8129475017689415,
|
| 212 |
+
"eval_loss": 0.5561972856521606,
|
| 213 |
+
"eval_precision": 0.8309911311946955,
|
| 214 |
+
"eval_recall": 0.8061719833564494,
|
| 215 |
+
"eval_runtime": 24.4163,
|
| 216 |
+
"eval_samples_per_second": 118.118,
|
| 217 |
+
"eval_steps_per_second": 14.785,
|
| 218 |
+
"step": 3531
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 12.0,
|
| 222 |
+
"grad_norm": 49.789276123046875,
|
| 223 |
+
"learning_rate": 8.205719438083829e-05,
|
| 224 |
+
"loss": 0.5808,
|
| 225 |
+
"step": 3852
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 12.0,
|
| 229 |
+
"eval_accuracy": 0.7708044382801664,
|
| 230 |
+
"eval_f1": 0.7870590251597063,
|
| 231 |
+
"eval_loss": 0.6607648134231567,
|
| 232 |
+
"eval_precision": 0.8277847611250527,
|
| 233 |
+
"eval_recall": 0.7708044382801664,
|
| 234 |
+
"eval_runtime": 24.4945,
|
| 235 |
+
"eval_samples_per_second": 117.741,
|
| 236 |
+
"eval_steps_per_second": 14.738,
|
| 237 |
+
"step": 3852
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 13.0,
|
| 241 |
+
"grad_norm": 11.32583999633789,
|
| 242 |
+
"learning_rate": 7.669273692531118e-05,
|
| 243 |
+
"loss": 0.4534,
|
| 244 |
+
"step": 4173
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 13.0,
|
| 248 |
+
"eval_accuracy": 0.8300970873786407,
|
| 249 |
+
"eval_f1": 0.8290964834873092,
|
| 250 |
+
"eval_loss": 0.5684113502502441,
|
| 251 |
+
"eval_precision": 0.8483268608964948,
|
| 252 |
+
"eval_recall": 0.8300970873786407,
|
| 253 |
+
"eval_runtime": 24.424,
|
| 254 |
+
"eval_samples_per_second": 118.08,
|
| 255 |
+
"eval_steps_per_second": 14.781,
|
| 256 |
+
"step": 4173
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 14.0,
|
| 260 |
+
"grad_norm": 4.780898094177246,
|
| 261 |
+
"learning_rate": 7.085797357089247e-05,
|
| 262 |
+
"loss": 0.4519,
|
| 263 |
+
"step": 4494
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 14.0,
|
| 267 |
+
"eval_accuracy": 0.8120665742024965,
|
| 268 |
+
"eval_f1": 0.8201472901135594,
|
| 269 |
+
"eval_loss": 0.5927954316139221,
|
| 270 |
+
"eval_precision": 0.8387522929504424,
|
| 271 |
+
"eval_recall": 0.8120665742024965,
|
| 272 |
+
"eval_runtime": 23.8992,
|
| 273 |
+
"eval_samples_per_second": 120.673,
|
| 274 |
+
"eval_steps_per_second": 15.105,
|
| 275 |
+
"step": 4494
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 15.0,
|
| 279 |
+
"grad_norm": 3.468371629714966,
|
| 280 |
+
"learning_rate": 6.46557084486047e-05,
|
| 281 |
+
"loss": 0.3998,
|
| 282 |
+
"step": 4815
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 15.0,
|
| 286 |
+
"eval_accuracy": 0.8262829403606102,
|
| 287 |
+
"eval_f1": 0.832391972746498,
|
| 288 |
+
"eval_loss": 0.5276548266410828,
|
| 289 |
+
"eval_precision": 0.8466524464455557,
|
| 290 |
+
"eval_recall": 0.8262829403606102,
|
| 291 |
+
"eval_runtime": 24.4747,
|
| 292 |
+
"eval_samples_per_second": 117.836,
|
| 293 |
+
"eval_steps_per_second": 14.75,
|
| 294 |
+
"step": 4815
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 16.0,
|
| 298 |
+
"grad_norm": 22.259729385375977,
|
| 299 |
+
"learning_rate": 5.8195220793532045e-05,
|
| 300 |
+
"loss": 0.3307,
|
| 301 |
+
"step": 5136
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 16.0,
|
| 305 |
+
"eval_accuracy": 0.826629680998613,
|
| 306 |
+
"eval_f1": 0.8330263843090071,
|
| 307 |
+
"eval_loss": 0.5944316387176514,
|
| 308 |
+
"eval_precision": 0.845755454394291,
|
| 309 |
+
"eval_recall": 0.826629680998613,
|
| 310 |
+
"eval_runtime": 24.425,
|
| 311 |
+
"eval_samples_per_second": 118.076,
|
| 312 |
+
"eval_steps_per_second": 14.78,
|
| 313 |
+
"step": 5136
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 17.0,
|
| 317 |
+
"grad_norm": 28.815109252929688,
|
| 318 |
+
"learning_rate": 5.15903395270923e-05,
|
| 319 |
+
"loss": 0.2899,
|
| 320 |
+
"step": 5457
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 17.0,
|
| 324 |
+
"eval_accuracy": 0.852635228848821,
|
| 325 |
+
"eval_f1": 0.8523735375042804,
|
| 326 |
+
"eval_loss": 0.538686990737915,
|
| 327 |
+
"eval_precision": 0.8545742100242152,
|
| 328 |
+
"eval_recall": 0.852635228848821,
|
| 329 |
+
"eval_runtime": 24.6184,
|
| 330 |
+
"eval_samples_per_second": 117.148,
|
| 331 |
+
"eval_steps_per_second": 14.664,
|
| 332 |
+
"step": 5457
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 18.0,
|
| 336 |
+
"grad_norm": 23.336013793945312,
|
| 337 |
+
"learning_rate": 4.495743767726598e-05,
|
| 338 |
+
"loss": 0.2466,
|
| 339 |
+
"step": 5778
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 18.0,
|
| 343 |
+
"eval_accuracy": 0.8495145631067961,
|
| 344 |
+
"eval_f1": 0.8505845839961479,
|
| 345 |
+
"eval_loss": 0.5559062361717224,
|
| 346 |
+
"eval_precision": 0.8540896208450929,
|
| 347 |
+
"eval_recall": 0.8495145631067961,
|
| 348 |
+
"eval_runtime": 24.3265,
|
| 349 |
+
"eval_samples_per_second": 118.554,
|
| 350 |
+
"eval_steps_per_second": 14.84,
|
| 351 |
+
"step": 5778
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 19.0,
|
| 355 |
+
"grad_norm": 0.06215408816933632,
|
| 356 |
+
"learning_rate": 3.841338197358591e-05,
|
| 357 |
+
"loss": 0.201,
|
| 358 |
+
"step": 6099
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 19.0,
|
| 362 |
+
"eval_accuracy": 0.8335644937586685,
|
| 363 |
+
"eval_f1": 0.8426585405075842,
|
| 364 |
+
"eval_loss": 0.6359649300575256,
|
| 365 |
+
"eval_precision": 0.8670540297358912,
|
| 366 |
+
"eval_recall": 0.8335644937586685,
|
| 367 |
+
"eval_runtime": 23.9522,
|
| 368 |
+
"eval_samples_per_second": 120.406,
|
| 369 |
+
"eval_steps_per_second": 15.072,
|
| 370 |
+
"step": 6099
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 20.0,
|
| 374 |
+
"grad_norm": 1.8757870197296143,
|
| 375 |
+
"learning_rate": 3.2073473743477955e-05,
|
| 376 |
+
"loss": 0.2163,
|
| 377 |
+
"step": 6420
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 20.0,
|
| 381 |
+
"eval_accuracy": 0.8599167822468793,
|
| 382 |
+
"eval_f1": 0.8581261251415077,
|
| 383 |
+
"eval_loss": 0.6009441614151001,
|
| 384 |
+
"eval_precision": 0.8575486655359866,
|
| 385 |
+
"eval_recall": 0.8599167822468793,
|
| 386 |
+
"eval_runtime": 23.6371,
|
| 387 |
+
"eval_samples_per_second": 122.011,
|
| 388 |
+
"eval_steps_per_second": 15.273,
|
| 389 |
+
"step": 6420
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 21.0,
|
| 393 |
+
"grad_norm": 13.443089485168457,
|
| 394 |
+
"learning_rate": 2.604941738980618e-05,
|
| 395 |
+
"loss": 0.1614,
|
| 396 |
+
"step": 6741
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 21.0,
|
| 400 |
+
"eval_accuracy": 0.8689320388349514,
|
| 401 |
+
"eval_f1": 0.8630350691040827,
|
| 402 |
+
"eval_loss": 0.5803854465484619,
|
| 403 |
+
"eval_precision": 0.8647592823276374,
|
| 404 |
+
"eval_recall": 0.8689320388349514,
|
| 405 |
+
"eval_runtime": 24.1018,
|
| 406 |
+
"eval_samples_per_second": 119.659,
|
| 407 |
+
"eval_steps_per_second": 14.978,
|
| 408 |
+
"step": 6741
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 22.0,
|
| 412 |
+
"grad_norm": 44.889469146728516,
|
| 413 |
+
"learning_rate": 2.0447352243517255e-05,
|
| 414 |
+
"loss": 0.1106,
|
| 415 |
+
"step": 7062
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 22.0,
|
| 419 |
+
"eval_accuracy": 0.8689320388349514,
|
| 420 |
+
"eval_f1": 0.8669998549584287,
|
| 421 |
+
"eval_loss": 0.5798132419586182,
|
| 422 |
+
"eval_precision": 0.8660861822399937,
|
| 423 |
+
"eval_recall": 0.8689320388349514,
|
| 424 |
+
"eval_runtime": 23.3011,
|
| 425 |
+
"eval_samples_per_second": 123.771,
|
| 426 |
+
"eval_steps_per_second": 15.493,
|
| 427 |
+
"step": 7062
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 23.0,
|
| 431 |
+
"grad_norm": 0.35838574171066284,
|
| 432 |
+
"learning_rate": 1.536598246865575e-05,
|
| 433 |
+
"loss": 0.1243,
|
| 434 |
+
"step": 7383
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 23.0,
|
| 438 |
+
"eval_accuracy": 0.8703190013869625,
|
| 439 |
+
"eval_f1": 0.8671952665464066,
|
| 440 |
+
"eval_loss": 0.6228230595588684,
|
| 441 |
+
"eval_precision": 0.8686274693929571,
|
| 442 |
+
"eval_recall": 0.8703190013869625,
|
| 443 |
+
"eval_runtime": 23.1472,
|
| 444 |
+
"eval_samples_per_second": 124.594,
|
| 445 |
+
"eval_steps_per_second": 15.596,
|
| 446 |
+
"step": 7383
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 24.0,
|
| 450 |
+
"grad_norm": 0.4248667359352112,
|
| 451 |
+
"learning_rate": 1.0894837969414489e-05,
|
| 452 |
+
"loss": 0.1251,
|
| 453 |
+
"step": 7704
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 24.0,
|
| 457 |
+
"eval_accuracy": 0.8727461858529819,
|
| 458 |
+
"eval_f1": 0.8697599306698692,
|
| 459 |
+
"eval_loss": 0.5986893177032471,
|
| 460 |
+
"eval_precision": 0.8694921227485961,
|
| 461 |
+
"eval_recall": 0.8727461858529819,
|
| 462 |
+
"eval_runtime": 22.875,
|
| 463 |
+
"eval_samples_per_second": 126.076,
|
| 464 |
+
"eval_steps_per_second": 15.781,
|
| 465 |
+
"step": 7704
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 25.0,
|
| 469 |
+
"grad_norm": 0.290132075548172,
|
| 470 |
+
"learning_rate": 7.112696940726155e-06,
|
| 471 |
+
"loss": 0.1038,
|
| 472 |
+
"step": 8025
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 25.0,
|
| 476 |
+
"eval_accuracy": 0.8769070735090153,
|
| 477 |
+
"eval_f1": 0.8753178399399126,
|
| 478 |
+
"eval_loss": 0.5806095004081726,
|
| 479 |
+
"eval_precision": 0.8755589237651312,
|
| 480 |
+
"eval_recall": 0.8769070735090153,
|
| 481 |
+
"eval_runtime": 22.7161,
|
| 482 |
+
"eval_samples_per_second": 126.958,
|
| 483 |
+
"eval_steps_per_second": 15.892,
|
| 484 |
+
"step": 8025
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 25.0,
|
| 488 |
+
"step": 8025,
|
| 489 |
+
"total_flos": 9.934918266289766e+18,
|
| 490 |
+
"train_loss": 0.5703618476843908,
|
| 491 |
+
"train_runtime": 3054.4722,
|
| 492 |
+
"train_samples_per_second": 167.885,
|
| 493 |
+
"train_steps_per_second": 10.509
|
| 494 |
+
}
|
| 495 |
+
],
|
| 496 |
+
"logging_steps": 500,
|
| 497 |
+
"max_steps": 32100,
|
| 498 |
+
"num_input_tokens_seen": 0,
|
| 499 |
+
"num_train_epochs": 100,
|
| 500 |
+
"save_steps": 500,
|
| 501 |
+
"total_flos": 9.934918266289766e+18,
|
| 502 |
+
"train_batch_size": 16,
|
| 503 |
+
"trial_name": null,
|
| 504 |
+
"trial_params": null
|
| 505 |
+
}
|