Commit
•
22b43f2
1
Parent(s):
d0746b2
End of training
Browse files- README.md +2 -2
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Apr18_23-36-48_12e8e850ae9b/events.out.tfevents.1713484334.12e8e850ae9b.943.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +540 -0
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.6806
|
| 21 |
+
- Accuracy: 0.8333
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"eval_accuracy": 0.8333333333333334,
|
| 4 |
+
"eval_loss": 0.6805989146232605,
|
| 5 |
+
"eval_runtime": 1.5712,
|
| 6 |
+
"eval_samples_per_second": 38.186,
|
| 7 |
+
"eval_steps_per_second": 1.273,
|
| 8 |
+
"total_flos": 1.2027310550050406e+18,
|
| 9 |
+
"train_loss": 0.44187615152929915,
|
| 10 |
+
"train_runtime": 903.9492,
|
| 11 |
+
"train_samples_per_second": 45.255,
|
| 12 |
+
"train_steps_per_second": 0.325
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"eval_accuracy": 0.8333333333333334,
|
| 4 |
+
"eval_loss": 0.6805989146232605,
|
| 5 |
+
"eval_runtime": 1.5712,
|
| 6 |
+
"eval_samples_per_second": 38.186,
|
| 7 |
+
"eval_steps_per_second": 1.273
|
| 8 |
+
}
|
runs/Apr18_23-36-48_12e8e850ae9b/events.out.tfevents.1713484334.12e8e850ae9b.943.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1905ea8fc159e34369d8ddf30e6eda2239bf97de83ef89834c8bc2778eb21c77
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"total_flos": 1.2027310550050406e+18,
|
| 4 |
+
"train_loss": 0.44187615152929915,
|
| 5 |
+
"train_runtime": 903.9492,
|
| 6 |
+
"train_samples_per_second": 45.255,
|
| 7 |
+
"train_steps_per_second": 0.325
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,540 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8333333333333334,
|
| 3 |
+
"best_model_checkpoint": "swinv2-tiny-patch4-window8-256-dmae-va-U5-42/checkpoint-201",
|
| 4 |
+
"epoch": 37.935483870967744,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 294,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9,
|
| 13 |
+
"eval_accuracy": 0.4,
|
| 14 |
+
"eval_loss": 1.3298888206481934,
|
| 15 |
+
"eval_runtime": 2.4022,
|
| 16 |
+
"eval_samples_per_second": 24.977,
|
| 17 |
+
"eval_steps_per_second": 0.833,
|
| 18 |
+
"step": 7
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 1.55,
|
| 22 |
+
"grad_norm": 3.2318508625030518,
|
| 23 |
+
"learning_rate": 2e-05,
|
| 24 |
+
"loss": 1.3678,
|
| 25 |
+
"step": 12
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.94,
|
| 29 |
+
"eval_accuracy": 0.45,
|
| 30 |
+
"eval_loss": 1.2662369012832642,
|
| 31 |
+
"eval_runtime": 1.2745,
|
| 32 |
+
"eval_samples_per_second": 47.077,
|
| 33 |
+
"eval_steps_per_second": 1.569,
|
| 34 |
+
"step": 15
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.97,
|
| 38 |
+
"eval_accuracy": 0.5166666666666667,
|
| 39 |
+
"eval_loss": 1.09585440158844,
|
| 40 |
+
"eval_runtime": 1.282,
|
| 41 |
+
"eval_samples_per_second": 46.801,
|
| 42 |
+
"eval_steps_per_second": 1.56,
|
| 43 |
+
"step": 23
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.1,
|
| 47 |
+
"grad_norm": 5.9810967445373535,
|
| 48 |
+
"learning_rate": 4e-05,
|
| 49 |
+
"loss": 1.2546,
|
| 50 |
+
"step": 24
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 4.0,
|
| 54 |
+
"eval_accuracy": 0.55,
|
| 55 |
+
"eval_loss": 0.9759470820426941,
|
| 56 |
+
"eval_runtime": 2.1999,
|
| 57 |
+
"eval_samples_per_second": 27.274,
|
| 58 |
+
"eval_steps_per_second": 0.909,
|
| 59 |
+
"step": 31
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 4.65,
|
| 63 |
+
"grad_norm": 8.929350852966309,
|
| 64 |
+
"learning_rate": 4.886363636363637e-05,
|
| 65 |
+
"loss": 1.0271,
|
| 66 |
+
"step": 36
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.9,
|
| 70 |
+
"eval_accuracy": 0.5666666666666667,
|
| 71 |
+
"eval_loss": 0.9374793767929077,
|
| 72 |
+
"eval_runtime": 1.3051,
|
| 73 |
+
"eval_samples_per_second": 45.972,
|
| 74 |
+
"eval_steps_per_second": 1.532,
|
| 75 |
+
"step": 38
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 5.94,
|
| 79 |
+
"eval_accuracy": 0.6,
|
| 80 |
+
"eval_loss": 0.8728086948394775,
|
| 81 |
+
"eval_runtime": 1.6572,
|
| 82 |
+
"eval_samples_per_second": 36.206,
|
| 83 |
+
"eval_steps_per_second": 1.207,
|
| 84 |
+
"step": 46
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 6.19,
|
| 88 |
+
"grad_norm": 12.113182067871094,
|
| 89 |
+
"learning_rate": 4.659090909090909e-05,
|
| 90 |
+
"loss": 0.8075,
|
| 91 |
+
"step": 48
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 6.97,
|
| 95 |
+
"eval_accuracy": 0.7166666666666667,
|
| 96 |
+
"eval_loss": 0.7359870076179504,
|
| 97 |
+
"eval_runtime": 1.8366,
|
| 98 |
+
"eval_samples_per_second": 32.668,
|
| 99 |
+
"eval_steps_per_second": 1.089,
|
| 100 |
+
"step": 54
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 7.74,
|
| 104 |
+
"grad_norm": 10.912686347961426,
|
| 105 |
+
"learning_rate": 4.431818181818182e-05,
|
| 106 |
+
"loss": 0.7026,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 8.0,
|
| 111 |
+
"eval_accuracy": 0.6666666666666666,
|
| 112 |
+
"eval_loss": 0.8097081780433655,
|
| 113 |
+
"eval_runtime": 1.5216,
|
| 114 |
+
"eval_samples_per_second": 39.432,
|
| 115 |
+
"eval_steps_per_second": 1.314,
|
| 116 |
+
"step": 62
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 8.9,
|
| 120 |
+
"eval_accuracy": 0.7,
|
| 121 |
+
"eval_loss": 0.7073581218719482,
|
| 122 |
+
"eval_runtime": 1.2986,
|
| 123 |
+
"eval_samples_per_second": 46.204,
|
| 124 |
+
"eval_steps_per_second": 1.54,
|
| 125 |
+
"step": 69
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 9.29,
|
| 129 |
+
"grad_norm": 7.962848663330078,
|
| 130 |
+
"learning_rate": 4.204545454545455e-05,
|
| 131 |
+
"loss": 0.5711,
|
| 132 |
+
"step": 72
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 9.94,
|
| 136 |
+
"eval_accuracy": 0.7833333333333333,
|
| 137 |
+
"eval_loss": 0.6912961602210999,
|
| 138 |
+
"eval_runtime": 1.3465,
|
| 139 |
+
"eval_samples_per_second": 44.561,
|
| 140 |
+
"eval_steps_per_second": 1.485,
|
| 141 |
+
"step": 77
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 10.84,
|
| 145 |
+
"grad_norm": 9.730597496032715,
|
| 146 |
+
"learning_rate": 3.9772727272727275e-05,
|
| 147 |
+
"loss": 0.5063,
|
| 148 |
+
"step": 84
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 10.97,
|
| 152 |
+
"eval_accuracy": 0.7166666666666667,
|
| 153 |
+
"eval_loss": 0.7461908459663391,
|
| 154 |
+
"eval_runtime": 1.3164,
|
| 155 |
+
"eval_samples_per_second": 45.578,
|
| 156 |
+
"eval_steps_per_second": 1.519,
|
| 157 |
+
"step": 85
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 12.0,
|
| 161 |
+
"eval_accuracy": 0.5833333333333334,
|
| 162 |
+
"eval_loss": 0.8508856296539307,
|
| 163 |
+
"eval_runtime": 1.2995,
|
| 164 |
+
"eval_samples_per_second": 46.172,
|
| 165 |
+
"eval_steps_per_second": 1.539,
|
| 166 |
+
"step": 93
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 12.39,
|
| 170 |
+
"grad_norm": 6.257162570953369,
|
| 171 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 172 |
+
"loss": 0.4701,
|
| 173 |
+
"step": 96
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 12.9,
|
| 177 |
+
"eval_accuracy": 0.7333333333333333,
|
| 178 |
+
"eval_loss": 0.6895251870155334,
|
| 179 |
+
"eval_runtime": 1.2873,
|
| 180 |
+
"eval_samples_per_second": 46.61,
|
| 181 |
+
"eval_steps_per_second": 1.554,
|
| 182 |
+
"step": 100
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 13.94,
|
| 186 |
+
"grad_norm": 8.040353775024414,
|
| 187 |
+
"learning_rate": 3.522727272727273e-05,
|
| 188 |
+
"loss": 0.3708,
|
| 189 |
+
"step": 108
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 13.94,
|
| 193 |
+
"eval_accuracy": 0.6833333333333333,
|
| 194 |
+
"eval_loss": 0.7592976689338684,
|
| 195 |
+
"eval_runtime": 1.3064,
|
| 196 |
+
"eval_samples_per_second": 45.927,
|
| 197 |
+
"eval_steps_per_second": 1.531,
|
| 198 |
+
"step": 108
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 14.97,
|
| 202 |
+
"eval_accuracy": 0.7166666666666667,
|
| 203 |
+
"eval_loss": 0.8621997833251953,
|
| 204 |
+
"eval_runtime": 1.8003,
|
| 205 |
+
"eval_samples_per_second": 33.327,
|
| 206 |
+
"eval_steps_per_second": 1.111,
|
| 207 |
+
"step": 116
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 15.48,
|
| 211 |
+
"grad_norm": 9.414116859436035,
|
| 212 |
+
"learning_rate": 3.295454545454545e-05,
|
| 213 |
+
"loss": 0.3581,
|
| 214 |
+
"step": 120
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 16.0,
|
| 218 |
+
"eval_accuracy": 0.7666666666666667,
|
| 219 |
+
"eval_loss": 0.750374972820282,
|
| 220 |
+
"eval_runtime": 1.348,
|
| 221 |
+
"eval_samples_per_second": 44.509,
|
| 222 |
+
"eval_steps_per_second": 1.484,
|
| 223 |
+
"step": 124
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 16.9,
|
| 227 |
+
"eval_accuracy": 0.75,
|
| 228 |
+
"eval_loss": 0.669380247592926,
|
| 229 |
+
"eval_runtime": 1.2906,
|
| 230 |
+
"eval_samples_per_second": 46.49,
|
| 231 |
+
"eval_steps_per_second": 1.55,
|
| 232 |
+
"step": 131
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 17.03,
|
| 236 |
+
"grad_norm": 10.203311920166016,
|
| 237 |
+
"learning_rate": 3.068181818181818e-05,
|
| 238 |
+
"loss": 0.3342,
|
| 239 |
+
"step": 132
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 17.94,
|
| 243 |
+
"eval_accuracy": 0.7333333333333333,
|
| 244 |
+
"eval_loss": 0.726170003414154,
|
| 245 |
+
"eval_runtime": 1.2961,
|
| 246 |
+
"eval_samples_per_second": 46.292,
|
| 247 |
+
"eval_steps_per_second": 1.543,
|
| 248 |
+
"step": 139
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 18.58,
|
| 252 |
+
"grad_norm": 6.9751410484313965,
|
| 253 |
+
"learning_rate": 2.8409090909090912e-05,
|
| 254 |
+
"loss": 0.2979,
|
| 255 |
+
"step": 144
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 18.97,
|
| 259 |
+
"eval_accuracy": 0.7166666666666667,
|
| 260 |
+
"eval_loss": 0.7233790159225464,
|
| 261 |
+
"eval_runtime": 1.5486,
|
| 262 |
+
"eval_samples_per_second": 38.744,
|
| 263 |
+
"eval_steps_per_second": 1.291,
|
| 264 |
+
"step": 147
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 20.0,
|
| 268 |
+
"eval_accuracy": 0.7833333333333333,
|
| 269 |
+
"eval_loss": 0.6403045058250427,
|
| 270 |
+
"eval_runtime": 1.8291,
|
| 271 |
+
"eval_samples_per_second": 32.804,
|
| 272 |
+
"eval_steps_per_second": 1.093,
|
| 273 |
+
"step": 155
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 20.13,
|
| 277 |
+
"grad_norm": 11.188940048217773,
|
| 278 |
+
"learning_rate": 2.6136363636363637e-05,
|
| 279 |
+
"loss": 0.2919,
|
| 280 |
+
"step": 156
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 20.9,
|
| 284 |
+
"eval_accuracy": 0.7666666666666667,
|
| 285 |
+
"eval_loss": 0.6847358345985413,
|
| 286 |
+
"eval_runtime": 1.8691,
|
| 287 |
+
"eval_samples_per_second": 32.102,
|
| 288 |
+
"eval_steps_per_second": 1.07,
|
| 289 |
+
"step": 162
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 21.68,
|
| 293 |
+
"grad_norm": 8.915830612182617,
|
| 294 |
+
"learning_rate": 2.3863636363636365e-05,
|
| 295 |
+
"loss": 0.274,
|
| 296 |
+
"step": 168
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 21.94,
|
| 300 |
+
"eval_accuracy": 0.75,
|
| 301 |
+
"eval_loss": 0.6942654848098755,
|
| 302 |
+
"eval_runtime": 1.6216,
|
| 303 |
+
"eval_samples_per_second": 37.0,
|
| 304 |
+
"eval_steps_per_second": 1.233,
|
| 305 |
+
"step": 170
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 22.97,
|
| 309 |
+
"eval_accuracy": 0.7833333333333333,
|
| 310 |
+
"eval_loss": 0.7235167026519775,
|
| 311 |
+
"eval_runtime": 1.2987,
|
| 312 |
+
"eval_samples_per_second": 46.199,
|
| 313 |
+
"eval_steps_per_second": 1.54,
|
| 314 |
+
"step": 178
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 23.23,
|
| 318 |
+
"grad_norm": 5.901014804840088,
|
| 319 |
+
"learning_rate": 2.1590909090909093e-05,
|
| 320 |
+
"loss": 0.2434,
|
| 321 |
+
"step": 180
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 24.0,
|
| 325 |
+
"eval_accuracy": 0.75,
|
| 326 |
+
"eval_loss": 0.7835569977760315,
|
| 327 |
+
"eval_runtime": 1.3206,
|
| 328 |
+
"eval_samples_per_second": 45.434,
|
| 329 |
+
"eval_steps_per_second": 1.514,
|
| 330 |
+
"step": 186
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 24.77,
|
| 334 |
+
"grad_norm": 9.987520217895508,
|
| 335 |
+
"learning_rate": 1.9318181818181818e-05,
|
| 336 |
+
"loss": 0.239,
|
| 337 |
+
"step": 192
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 24.9,
|
| 341 |
+
"eval_accuracy": 0.8166666666666667,
|
| 342 |
+
"eval_loss": 0.7199376225471497,
|
| 343 |
+
"eval_runtime": 1.3056,
|
| 344 |
+
"eval_samples_per_second": 45.956,
|
| 345 |
+
"eval_steps_per_second": 1.532,
|
| 346 |
+
"step": 193
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 25.94,
|
| 350 |
+
"eval_accuracy": 0.8333333333333334,
|
| 351 |
+
"eval_loss": 0.6805989146232605,
|
| 352 |
+
"eval_runtime": 1.2869,
|
| 353 |
+
"eval_samples_per_second": 46.625,
|
| 354 |
+
"eval_steps_per_second": 1.554,
|
| 355 |
+
"step": 201
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 26.32,
|
| 359 |
+
"grad_norm": 9.460673332214355,
|
| 360 |
+
"learning_rate": 1.7045454545454546e-05,
|
| 361 |
+
"loss": 0.2184,
|
| 362 |
+
"step": 204
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 26.97,
|
| 366 |
+
"eval_accuracy": 0.8,
|
| 367 |
+
"eval_loss": 0.6923350095748901,
|
| 368 |
+
"eval_runtime": 1.3321,
|
| 369 |
+
"eval_samples_per_second": 45.043,
|
| 370 |
+
"eval_steps_per_second": 1.501,
|
| 371 |
+
"step": 209
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 27.87,
|
| 375 |
+
"grad_norm": 7.031528472900391,
|
| 376 |
+
"learning_rate": 1.4772727272727274e-05,
|
| 377 |
+
"loss": 0.2176,
|
| 378 |
+
"step": 216
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 28.0,
|
| 382 |
+
"eval_accuracy": 0.7833333333333333,
|
| 383 |
+
"eval_loss": 0.7069784998893738,
|
| 384 |
+
"eval_runtime": 1.3346,
|
| 385 |
+
"eval_samples_per_second": 44.957,
|
| 386 |
+
"eval_steps_per_second": 1.499,
|
| 387 |
+
"step": 217
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 28.9,
|
| 391 |
+
"eval_accuracy": 0.7666666666666667,
|
| 392 |
+
"eval_loss": 0.6990731954574585,
|
| 393 |
+
"eval_runtime": 1.3399,
|
| 394 |
+
"eval_samples_per_second": 44.779,
|
| 395 |
+
"eval_steps_per_second": 1.493,
|
| 396 |
+
"step": 224
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 29.42,
|
| 400 |
+
"grad_norm": 6.380643367767334,
|
| 401 |
+
"learning_rate": 1.25e-05,
|
| 402 |
+
"loss": 0.231,
|
| 403 |
+
"step": 228
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 29.94,
|
| 407 |
+
"eval_accuracy": 0.7833333333333333,
|
| 408 |
+
"eval_loss": 0.7043098211288452,
|
| 409 |
+
"eval_runtime": 1.3053,
|
| 410 |
+
"eval_samples_per_second": 45.968,
|
| 411 |
+
"eval_steps_per_second": 1.532,
|
| 412 |
+
"step": 232
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 30.97,
|
| 416 |
+
"grad_norm": 7.81527042388916,
|
| 417 |
+
"learning_rate": 1.0227272727272729e-05,
|
| 418 |
+
"loss": 0.1889,
|
| 419 |
+
"step": 240
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 30.97,
|
| 423 |
+
"eval_accuracy": 0.7666666666666667,
|
| 424 |
+
"eval_loss": 0.6575068235397339,
|
| 425 |
+
"eval_runtime": 1.3306,
|
| 426 |
+
"eval_samples_per_second": 45.092,
|
| 427 |
+
"eval_steps_per_second": 1.503,
|
| 428 |
+
"step": 240
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 32.0,
|
| 432 |
+
"eval_accuracy": 0.75,
|
| 433 |
+
"eval_loss": 0.752100944519043,
|
| 434 |
+
"eval_runtime": 1.3091,
|
| 435 |
+
"eval_samples_per_second": 45.833,
|
| 436 |
+
"eval_steps_per_second": 1.528,
|
| 437 |
+
"step": 248
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 32.52,
|
| 441 |
+
"grad_norm": 12.994545936584473,
|
| 442 |
+
"learning_rate": 7.954545454545455e-06,
|
| 443 |
+
"loss": 0.2033,
|
| 444 |
+
"step": 252
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 32.9,
|
| 448 |
+
"eval_accuracy": 0.7833333333333333,
|
| 449 |
+
"eval_loss": 0.7062455415725708,
|
| 450 |
+
"eval_runtime": 1.3337,
|
| 451 |
+
"eval_samples_per_second": 44.989,
|
| 452 |
+
"eval_steps_per_second": 1.5,
|
| 453 |
+
"step": 255
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 33.94,
|
| 457 |
+
"eval_accuracy": 0.8,
|
| 458 |
+
"eval_loss": 0.6957877278327942,
|
| 459 |
+
"eval_runtime": 1.3005,
|
| 460 |
+
"eval_samples_per_second": 46.136,
|
| 461 |
+
"eval_steps_per_second": 1.538,
|
| 462 |
+
"step": 263
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 34.06,
|
| 466 |
+
"grad_norm": 6.9790143966674805,
|
| 467 |
+
"learning_rate": 5.681818181818182e-06,
|
| 468 |
+
"loss": 0.1891,
|
| 469 |
+
"step": 264
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 34.97,
|
| 473 |
+
"eval_accuracy": 0.8,
|
| 474 |
+
"eval_loss": 0.7188975811004639,
|
| 475 |
+
"eval_runtime": 1.2948,
|
| 476 |
+
"eval_samples_per_second": 46.34,
|
| 477 |
+
"eval_steps_per_second": 1.545,
|
| 478 |
+
"step": 271
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 35.61,
|
| 482 |
+
"grad_norm": 10.62868881225586,
|
| 483 |
+
"learning_rate": 3.409090909090909e-06,
|
| 484 |
+
"loss": 0.1739,
|
| 485 |
+
"step": 276
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 36.0,
|
| 489 |
+
"eval_accuracy": 0.8,
|
| 490 |
+
"eval_loss": 0.7456963658332825,
|
| 491 |
+
"eval_runtime": 1.3109,
|
| 492 |
+
"eval_samples_per_second": 45.772,
|
| 493 |
+
"eval_steps_per_second": 1.526,
|
| 494 |
+
"step": 279
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 36.9,
|
| 498 |
+
"eval_accuracy": 0.7833333333333333,
|
| 499 |
+
"eval_loss": 0.7766147255897522,
|
| 500 |
+
"eval_runtime": 1.3196,
|
| 501 |
+
"eval_samples_per_second": 45.469,
|
| 502 |
+
"eval_steps_per_second": 1.516,
|
| 503 |
+
"step": 286
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 37.16,
|
| 507 |
+
"grad_norm": 11.091560363769531,
|
| 508 |
+
"learning_rate": 1.1363636363636364e-06,
|
| 509 |
+
"loss": 0.1949,
|
| 510 |
+
"step": 288
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 37.94,
|
| 514 |
+
"eval_accuracy": 0.7666666666666667,
|
| 515 |
+
"eval_loss": 0.7808470129966736,
|
| 516 |
+
"eval_runtime": 1.3104,
|
| 517 |
+
"eval_samples_per_second": 45.788,
|
| 518 |
+
"eval_steps_per_second": 1.526,
|
| 519 |
+
"step": 294
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 37.94,
|
| 523 |
+
"step": 294,
|
| 524 |
+
"total_flos": 1.2027310550050406e+18,
|
| 525 |
+
"train_loss": 0.44187615152929915,
|
| 526 |
+
"train_runtime": 903.9492,
|
| 527 |
+
"train_samples_per_second": 45.255,
|
| 528 |
+
"train_steps_per_second": 0.325
|
| 529 |
+
}
|
| 530 |
+
],
|
| 531 |
+
"logging_steps": 12,
|
| 532 |
+
"max_steps": 294,
|
| 533 |
+
"num_input_tokens_seen": 0,
|
| 534 |
+
"num_train_epochs": 42,
|
| 535 |
+
"save_steps": 500,
|
| 536 |
+
"total_flos": 1.2027310550050406e+18,
|
| 537 |
+
"train_batch_size": 32,
|
| 538 |
+
"trial_name": null,
|
| 539 |
+
"trial_params": null
|
| 540 |
+
}
|