End of training
Browse files
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/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 1.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.1410
|
| 21 |
+
- Accuracy: 0.6445
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
CHANGED
|
@@ -1,8 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
"train_steps_per_second": 0.012
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.94818652849741,
|
| 3 |
+
"eval_accuracy": 0.6445224246502909,
|
| 4 |
+
"eval_loss": 1.1410062313079834,
|
| 5 |
+
"eval_runtime": 3958.7465,
|
| 6 |
+
"eval_samples_per_second": 5.255,
|
| 7 |
+
"eval_steps_per_second": 0.082,
|
| 8 |
+
"total_flos": 6.10407050089078e+18,
|
| 9 |
+
"train_loss": 1.4552634547154109,
|
| 10 |
+
"train_runtime": 78972.9917,
|
| 11 |
+
"train_samples_per_second": 3.124,
|
| 12 |
"train_steps_per_second": 0.012
|
| 13 |
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.94818652849741,
|
| 3 |
+
"eval_accuracy": 0.6445224246502909,
|
| 4 |
+
"eval_loss": 1.1410062313079834,
|
| 5 |
+
"eval_runtime": 3958.7465,
|
| 6 |
+
"eval_samples_per_second": 5.255,
|
| 7 |
+
"eval_steps_per_second": 0.082
|
| 8 |
+
}
|
runs/Jun27_08-47-10_c3df91dfd599/events.out.tfevents.1719560986.c3df91dfd599
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab2e624d68f7a540c20264ece6d52fd6c1515358fa34d47e32d73e330993398a
|
| 3 |
+
size 140
|
runs/Jun27_10-04-20_79e23ecbe329/events.out.tfevents.1719482662.79e23ecbe329
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7283c41fe220096459cb54c21eef04c6fc32f896efe7179b4477384abcad9bf8
|
| 3 |
+
size 29470
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second": 3.
|
| 7 |
"train_steps_per_second": 0.012
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.94818652849741,
|
| 3 |
+
"total_flos": 6.10407050089078e+18,
|
| 4 |
+
"train_loss": 1.4552634547154109,
|
| 5 |
+
"train_runtime": 78972.9917,
|
| 6 |
+
"train_samples_per_second": 3.124,
|
| 7 |
"train_steps_per_second": 0.012
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,118 +1,787 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "swin-tiny-patch4-window7-224-category-classification/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 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"eval_steps_per_second": 0.079,
|
| 18 |
-
"step": 3
|
| 19 |
},
|
| 20 |
{
|
| 21 |
-
"epoch":
|
| 22 |
-
"
|
| 23 |
-
"
|
| 24 |
-
"
|
| 25 |
-
"
|
| 26 |
-
|
| 27 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch":
|
| 31 |
-
"grad_norm":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
"learning_rate": 3.7037037037037037e-05,
|
| 33 |
-
"loss":
|
| 34 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
},
|
| 36 |
{
|
| 37 |
-
"epoch":
|
| 38 |
-
"
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
"eval_steps_per_second": 0.096,
|
| 43 |
-
"step": 11
|
| 44 |
},
|
| 45 |
{
|
| 46 |
"epoch": 4.0,
|
| 47 |
-
"eval_accuracy": 0.
|
| 48 |
-
"eval_loss":
|
| 49 |
-
"eval_runtime":
|
| 50 |
-
"eval_samples_per_second": 5.
|
| 51 |
-
"eval_steps_per_second": 0.
|
| 52 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch":
|
| 56 |
-
"
|
| 57 |
-
"
|
| 58 |
-
"
|
| 59 |
-
"
|
| 60 |
-
"eval_steps_per_second": 0.101,
|
| 61 |
-
"step": 18
|
| 62 |
},
|
| 63 |
{
|
| 64 |
-
"epoch": 5.
|
| 65 |
-
"grad_norm":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 66 |
"learning_rate": 1.8518518518518518e-05,
|
| 67 |
-
"loss":
|
| 68 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch":
|
| 72 |
-
"
|
| 73 |
-
"
|
| 74 |
-
"
|
| 75 |
-
"
|
| 76 |
-
"eval_steps_per_second": 0.097,
|
| 77 |
-
"step": 22
|
| 78 |
},
|
| 79 |
{
|
| 80 |
-
"epoch":
|
| 81 |
-
"
|
| 82 |
-
"
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 8.0,
|
| 90 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
"learning_rate": 0.0,
|
| 92 |
-
"loss":
|
| 93 |
-
"step":
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"eval_accuracy": 0.
|
| 98 |
-
"eval_loss":
|
| 99 |
-
"eval_runtime":
|
| 100 |
-
"eval_samples_per_second": 5.
|
| 101 |
-
"eval_steps_per_second": 0.
|
| 102 |
-
"step":
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"step":
|
| 107 |
-
"total_flos":
|
| 108 |
-
"train_loss":
|
| 109 |
-
"train_runtime":
|
| 110 |
-
"train_samples_per_second": 3.
|
| 111 |
"train_steps_per_second": 0.012
|
| 112 |
}
|
| 113 |
],
|
| 114 |
"logging_steps": 10,
|
| 115 |
-
"max_steps":
|
| 116 |
"num_input_tokens_seen": 0,
|
| 117 |
"num_train_epochs": 10,
|
| 118 |
"save_steps": 500,
|
|
@@ -128,7 +797,7 @@
|
|
| 128 |
"attributes": {}
|
| 129 |
}
|
| 130 |
},
|
| 131 |
-
"total_flos":
|
| 132 |
"train_batch_size": 64,
|
| 133 |
"trial_name": null,
|
| 134 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.6431283949430371,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-category-classification/checkpoint-868",
|
| 4 |
+
"epoch": 9.94818652849741,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 960,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.10362694300518134,
|
| 13 |
+
"grad_norm": 6.313995838165283,
|
| 14 |
+
"learning_rate": 5.208333333333334e-06,
|
| 15 |
+
"loss": 3.047,
|
| 16 |
+
"step": 10
|
|
|
|
|
|
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.20725388601036268,
|
| 20 |
+
"grad_norm": 3.603754758834839,
|
| 21 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 22 |
+
"loss": 2.9583,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.31088082901554404,
|
| 27 |
+
"grad_norm": 3.8277997970581055,
|
| 28 |
+
"learning_rate": 1.5625e-05,
|
| 29 |
+
"loss": 2.8186,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.41450777202072536,
|
| 34 |
+
"grad_norm": 3.712352752685547,
|
| 35 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 36 |
+
"loss": 2.5968,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5181347150259067,
|
| 41 |
+
"grad_norm": 6.0071611404418945,
|
| 42 |
+
"learning_rate": 2.604166666666667e-05,
|
| 43 |
+
"loss": 2.3988,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6217616580310881,
|
| 48 |
+
"grad_norm": 5.359269142150879,
|
| 49 |
+
"learning_rate": 3.125e-05,
|
| 50 |
+
"loss": 2.1299,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7253886010362695,
|
| 55 |
+
"grad_norm": 6.4594244956970215,
|
| 56 |
+
"learning_rate": 3.6458333333333336e-05,
|
| 57 |
+
"loss": 2.0938,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.8290155440414507,
|
| 62 |
+
"grad_norm": 6.613985538482666,
|
| 63 |
+
"learning_rate": 4.166666666666667e-05,
|
| 64 |
+
"loss": 1.9437,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.9326424870466321,
|
| 69 |
+
"grad_norm": 4.681708812713623,
|
| 70 |
+
"learning_rate": 4.6875e-05,
|
| 71 |
+
"loss": 1.9251,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.9948186528497409,
|
| 76 |
+
"eval_accuracy": 0.5348747776762967,
|
| 77 |
+
"eval_loss": 1.5363707542419434,
|
| 78 |
+
"eval_runtime": 4339.9701,
|
| 79 |
+
"eval_samples_per_second": 4.793,
|
| 80 |
+
"eval_steps_per_second": 0.075,
|
| 81 |
+
"step": 96
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0362694300518134,
|
| 85 |
+
"grad_norm": 6.605560779571533,
|
| 86 |
+
"learning_rate": 4.976851851851852e-05,
|
| 87 |
+
"loss": 1.8135,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.1398963730569949,
|
| 92 |
+
"grad_norm": 5.990049839019775,
|
| 93 |
+
"learning_rate": 4.9189814814814815e-05,
|
| 94 |
+
"loss": 1.7992,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.2435233160621761,
|
| 99 |
+
"grad_norm": 5.5032124519348145,
|
| 100 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 101 |
+
"loss": 1.7406,
|
| 102 |
+
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 1.3471502590673574,
|
| 106 |
+
"grad_norm": 6.0698771476745605,
|
| 107 |
+
"learning_rate": 4.803240740740741e-05,
|
| 108 |
+
"loss": 1.7347,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.450777202072539,
|
| 113 |
+
"grad_norm": 5.383842468261719,
|
| 114 |
+
"learning_rate": 4.745370370370371e-05,
|
| 115 |
+
"loss": 1.7108,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.5544041450777202,
|
| 120 |
+
"grad_norm": 5.526336193084717,
|
| 121 |
+
"learning_rate": 4.6875e-05,
|
| 122 |
+
"loss": 1.6553,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.6580310880829017,
|
| 127 |
+
"grad_norm": 6.193297863006592,
|
| 128 |
+
"learning_rate": 4.62962962962963e-05,
|
| 129 |
+
"loss": 1.584,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.761658031088083,
|
| 134 |
+
"grad_norm": 5.479032039642334,
|
| 135 |
+
"learning_rate": 4.5717592592592594e-05,
|
| 136 |
+
"loss": 1.6467,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.8652849740932642,
|
| 141 |
+
"grad_norm": 5.958849906921387,
|
| 142 |
+
"learning_rate": 4.5138888888888894e-05,
|
| 143 |
+
"loss": 1.6197,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.9689119170984455,
|
| 148 |
+
"grad_norm": 6.240902423858643,
|
| 149 |
+
"learning_rate": 4.456018518518519e-05,
|
| 150 |
+
"loss": 1.6301,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"eval_accuracy": 0.5978464644522424,
|
| 156 |
+
"eval_loss": 1.309254765510559,
|
| 157 |
+
"eval_runtime": 3808.3973,
|
| 158 |
+
"eval_samples_per_second": 5.462,
|
| 159 |
+
"eval_steps_per_second": 0.086,
|
| 160 |
+
"step": 193
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.0725388601036268,
|
| 164 |
+
"grad_norm": 6.163332939147949,
|
| 165 |
+
"learning_rate": 4.3981481481481486e-05,
|
| 166 |
+
"loss": 1.5208,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.1761658031088085,
|
| 171 |
+
"grad_norm": 4.633750915527344,
|
| 172 |
+
"learning_rate": 4.340277777777778e-05,
|
| 173 |
+
"loss": 1.5478,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.2797927461139897,
|
| 178 |
+
"grad_norm": 6.3137288093566895,
|
| 179 |
+
"learning_rate": 4.282407407407408e-05,
|
| 180 |
+
"loss": 1.5114,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.383419689119171,
|
| 185 |
+
"grad_norm": 6.306565761566162,
|
| 186 |
+
"learning_rate": 4.224537037037037e-05,
|
| 187 |
+
"loss": 1.5908,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.4870466321243523,
|
| 192 |
+
"grad_norm": 6.069610595703125,
|
| 193 |
+
"learning_rate": 4.166666666666667e-05,
|
| 194 |
+
"loss": 1.5485,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.5906735751295336,
|
| 199 |
+
"grad_norm": 5.604846477508545,
|
| 200 |
+
"learning_rate": 4.1087962962962965e-05,
|
| 201 |
+
"loss": 1.4435,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.694300518134715,
|
| 206 |
+
"grad_norm": 5.180357933044434,
|
| 207 |
+
"learning_rate": 4.0509259259259265e-05,
|
| 208 |
+
"loss": 1.4953,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.7979274611398965,
|
| 213 |
+
"grad_norm": 5.613372325897217,
|
| 214 |
+
"learning_rate": 3.993055555555556e-05,
|
| 215 |
+
"loss": 1.4898,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.901554404145078,
|
| 220 |
+
"grad_norm": 6.961531162261963,
|
| 221 |
+
"learning_rate": 3.935185185185186e-05,
|
| 222 |
+
"loss": 1.4793,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.994818652849741,
|
| 227 |
+
"eval_accuracy": 0.6074123924433976,
|
| 228 |
+
"eval_loss": 1.259105920791626,
|
| 229 |
+
"eval_runtime": 3877.0711,
|
| 230 |
+
"eval_samples_per_second": 5.366,
|
| 231 |
+
"eval_steps_per_second": 0.084,
|
| 232 |
+
"step": 289
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.005181347150259,
|
| 236 |
+
"grad_norm": 6.35546350479126,
|
| 237 |
+
"learning_rate": 3.877314814814815e-05,
|
| 238 |
+
"loss": 1.4606,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.1088082901554404,
|
| 243 |
+
"grad_norm": 4.932433128356934,
|
| 244 |
+
"learning_rate": 3.8194444444444444e-05,
|
| 245 |
+
"loss": 1.423,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.2124352331606216,
|
| 250 |
+
"grad_norm": 6.546709060668945,
|
| 251 |
+
"learning_rate": 3.7615740740740744e-05,
|
| 252 |
+
"loss": 1.406,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.3160621761658033,
|
| 257 |
+
"grad_norm": 9.50334644317627,
|
| 258 |
"learning_rate": 3.7037037037037037e-05,
|
| 259 |
+
"loss": 1.4365,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.4196891191709846,
|
| 264 |
+
"grad_norm": 5.8740363121032715,
|
| 265 |
+
"learning_rate": 3.6458333333333336e-05,
|
| 266 |
+
"loss": 1.4108,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.523316062176166,
|
| 271 |
+
"grad_norm": 6.044996738433838,
|
| 272 |
+
"learning_rate": 3.587962962962963e-05,
|
| 273 |
+
"loss": 1.4176,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.626943005181347,
|
| 278 |
+
"grad_norm": 5.8521552085876465,
|
| 279 |
+
"learning_rate": 3.530092592592593e-05,
|
| 280 |
+
"loss": 1.4497,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.7305699481865284,
|
| 285 |
+
"grad_norm": 5.023159027099609,
|
| 286 |
+
"learning_rate": 3.472222222222222e-05,
|
| 287 |
+
"loss": 1.3827,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.8341968911917097,
|
| 292 |
+
"grad_norm": 7.735539436340332,
|
| 293 |
+
"learning_rate": 3.414351851851852e-05,
|
| 294 |
+
"loss": 1.4248,
|
| 295 |
+
"step": 370
|
| 296 |
},
|
| 297 |
{
|
| 298 |
+
"epoch": 3.937823834196891,
|
| 299 |
+
"grad_norm": 6.757364273071289,
|
| 300 |
+
"learning_rate": 3.3564814814814815e-05,
|
| 301 |
+
"loss": 1.4316,
|
| 302 |
+
"step": 380
|
|
|
|
|
|
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 4.0,
|
| 306 |
+
"eval_accuracy": 0.6127962313127914,
|
| 307 |
+
"eval_loss": 1.2433934211730957,
|
| 308 |
+
"eval_runtime": 3727.6262,
|
| 309 |
+
"eval_samples_per_second": 5.581,
|
| 310 |
+
"eval_steps_per_second": 0.087,
|
| 311 |
+
"step": 386
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.041450777202073,
|
| 315 |
+
"grad_norm": 5.170826435089111,
|
| 316 |
+
"learning_rate": 3.2986111111111115e-05,
|
| 317 |
+
"loss": 1.3815,
|
| 318 |
+
"step": 390
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.1450777202072535,
|
| 322 |
+
"grad_norm": 5.3575053215026855,
|
| 323 |
+
"learning_rate": 3.240740740740741e-05,
|
| 324 |
+
"loss": 1.3492,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.248704663212435,
|
| 329 |
+
"grad_norm": 6.391707897186279,
|
| 330 |
+
"learning_rate": 3.182870370370371e-05,
|
| 331 |
+
"loss": 1.3728,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.352331606217617,
|
| 336 |
+
"grad_norm": 6.244965553283691,
|
| 337 |
+
"learning_rate": 3.125e-05,
|
| 338 |
+
"loss": 1.367,
|
| 339 |
+
"step": 420
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.455958549222798,
|
| 343 |
+
"grad_norm": 5.091292381286621,
|
| 344 |
+
"learning_rate": 3.06712962962963e-05,
|
| 345 |
+
"loss": 1.3732,
|
| 346 |
+
"step": 430
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.5595854922279795,
|
| 350 |
+
"grad_norm": 4.755329608917236,
|
| 351 |
+
"learning_rate": 3.0092592592592593e-05,
|
| 352 |
+
"loss": 1.3418,
|
| 353 |
+
"step": 440
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.66321243523316,
|
| 357 |
+
"grad_norm": 6.017394542694092,
|
| 358 |
+
"learning_rate": 2.951388888888889e-05,
|
| 359 |
+
"loss": 1.3519,
|
| 360 |
+
"step": 450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.766839378238342,
|
| 364 |
+
"grad_norm": 6.148505687713623,
|
| 365 |
+
"learning_rate": 2.8935185185185186e-05,
|
| 366 |
+
"loss": 1.3473,
|
| 367 |
+
"step": 460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.870466321243523,
|
| 371 |
+
"grad_norm": 5.228733062744141,
|
| 372 |
+
"learning_rate": 2.8356481481481483e-05,
|
| 373 |
+
"loss": 1.2983,
|
| 374 |
+
"step": 470
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.974093264248705,
|
| 378 |
+
"grad_norm": 5.9043803215026855,
|
| 379 |
+
"learning_rate": 2.777777777777778e-05,
|
| 380 |
+
"loss": 1.3424,
|
| 381 |
+
"step": 480
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.994818652849741,
|
| 385 |
+
"eval_accuracy": 0.6309666874969956,
|
| 386 |
+
"eval_loss": 1.1883153915405273,
|
| 387 |
+
"eval_runtime": 3739.3568,
|
| 388 |
+
"eval_samples_per_second": 5.563,
|
| 389 |
+
"eval_steps_per_second": 0.087,
|
| 390 |
+
"step": 482
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 5.077720207253886,
|
| 394 |
+
"grad_norm": 6.2051310539245605,
|
| 395 |
+
"learning_rate": 2.7199074074074076e-05,
|
| 396 |
+
"loss": 1.3081,
|
| 397 |
+
"step": 490
|
|
|
|
|
|
|
| 398 |
},
|
| 399 |
{
|
| 400 |
+
"epoch": 5.181347150259067,
|
| 401 |
+
"grad_norm": 5.718914985656738,
|
| 402 |
+
"learning_rate": 2.6620370370370372e-05,
|
| 403 |
+
"loss": 1.2773,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.284974093264249,
|
| 408 |
+
"grad_norm": 7.097866535186768,
|
| 409 |
+
"learning_rate": 2.604166666666667e-05,
|
| 410 |
+
"loss": 1.2914,
|
| 411 |
+
"step": 510
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.38860103626943,
|
| 415 |
+
"grad_norm": 5.6485724449157715,
|
| 416 |
+
"learning_rate": 2.5462962962962965e-05,
|
| 417 |
+
"loss": 1.3093,
|
| 418 |
+
"step": 520
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 5.492227979274611,
|
| 422 |
+
"grad_norm": 6.204397201538086,
|
| 423 |
+
"learning_rate": 2.488425925925926e-05,
|
| 424 |
+
"loss": 1.3067,
|
| 425 |
+
"step": 530
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.595854922279793,
|
| 429 |
+
"grad_norm": 6.455079555511475,
|
| 430 |
+
"learning_rate": 2.4305555555555558e-05,
|
| 431 |
+
"loss": 1.2826,
|
| 432 |
+
"step": 540
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 5.699481865284974,
|
| 436 |
+
"grad_norm": 5.391714572906494,
|
| 437 |
+
"learning_rate": 2.3726851851851854e-05,
|
| 438 |
+
"loss": 1.2741,
|
| 439 |
+
"step": 550
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 5.803108808290156,
|
| 443 |
+
"grad_norm": 5.677307605743408,
|
| 444 |
+
"learning_rate": 2.314814814814815e-05,
|
| 445 |
+
"loss": 1.3511,
|
| 446 |
+
"step": 560
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 5.9067357512953365,
|
| 450 |
+
"grad_norm": 6.441585063934326,
|
| 451 |
+
"learning_rate": 2.2569444444444447e-05,
|
| 452 |
+
"loss": 1.2867,
|
| 453 |
+
"step": 570
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 6.0,
|
| 457 |
+
"eval_accuracy": 0.6371196462048743,
|
| 458 |
+
"eval_loss": 1.1587461233139038,
|
| 459 |
+
"eval_runtime": 3638.0616,
|
| 460 |
+
"eval_samples_per_second": 5.718,
|
| 461 |
+
"eval_steps_per_second": 0.09,
|
| 462 |
+
"step": 579
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.010362694300518,
|
| 466 |
+
"grad_norm": 5.769332408905029,
|
| 467 |
+
"learning_rate": 2.1990740740740743e-05,
|
| 468 |
+
"loss": 1.3059,
|
| 469 |
+
"step": 580
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.1139896373057,
|
| 473 |
+
"grad_norm": 4.571792125701904,
|
| 474 |
+
"learning_rate": 2.141203703703704e-05,
|
| 475 |
+
"loss": 1.2681,
|
| 476 |
+
"step": 590
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.217616580310881,
|
| 480 |
+
"grad_norm": 5.652552127838135,
|
| 481 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 482 |
+
"loss": 1.2911,
|
| 483 |
+
"step": 600
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.321243523316062,
|
| 487 |
+
"grad_norm": 5.243156909942627,
|
| 488 |
+
"learning_rate": 2.0254629629629632e-05,
|
| 489 |
+
"loss": 1.2597,
|
| 490 |
+
"step": 610
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.424870466321243,
|
| 494 |
+
"grad_norm": 6.888844966888428,
|
| 495 |
+
"learning_rate": 1.967592592592593e-05,
|
| 496 |
+
"loss": 1.2195,
|
| 497 |
+
"step": 620
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 6.528497409326425,
|
| 501 |
+
"grad_norm": 6.571014404296875,
|
| 502 |
+
"learning_rate": 1.9097222222222222e-05,
|
| 503 |
+
"loss": 1.2624,
|
| 504 |
+
"step": 630
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 6.632124352331607,
|
| 508 |
+
"grad_norm": 7.260251998901367,
|
| 509 |
"learning_rate": 1.8518518518518518e-05,
|
| 510 |
+
"loss": 1.2817,
|
| 511 |
+
"step": 640
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 6.7357512953367875,
|
| 515 |
+
"grad_norm": 5.61851167678833,
|
| 516 |
+
"learning_rate": 1.7939814814814815e-05,
|
| 517 |
+
"loss": 1.2261,
|
| 518 |
+
"step": 650
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 6.839378238341969,
|
| 522 |
+
"grad_norm": 5.466727256774902,
|
| 523 |
+
"learning_rate": 1.736111111111111e-05,
|
| 524 |
+
"loss": 1.2524,
|
| 525 |
+
"step": 660
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 6.94300518134715,
|
| 529 |
+
"grad_norm": 6.450463771820068,
|
| 530 |
+
"learning_rate": 1.6782407407407408e-05,
|
| 531 |
+
"loss": 1.2177,
|
| 532 |
+
"step": 670
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 6.994818652849741,
|
| 536 |
+
"eval_accuracy": 0.6366389463058213,
|
| 537 |
+
"eval_loss": 1.161221981048584,
|
| 538 |
+
"eval_runtime": 3629.1662,
|
| 539 |
+
"eval_samples_per_second": 5.732,
|
| 540 |
+
"eval_steps_per_second": 0.09,
|
| 541 |
+
"step": 675
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.046632124352332,
|
| 545 |
+
"grad_norm": 6.2261786460876465,
|
| 546 |
+
"learning_rate": 1.6203703703703704e-05,
|
| 547 |
+
"loss": 1.2867,
|
| 548 |
+
"step": 680
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.150259067357513,
|
| 552 |
+
"grad_norm": 5.230395317077637,
|
| 553 |
+
"learning_rate": 1.5625e-05,
|
| 554 |
+
"loss": 1.2294,
|
| 555 |
+
"step": 690
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.253886010362694,
|
| 559 |
+
"grad_norm": 6.0760722160339355,
|
| 560 |
+
"learning_rate": 1.5046296296296297e-05,
|
| 561 |
+
"loss": 1.2574,
|
| 562 |
+
"step": 700
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.357512953367876,
|
| 566 |
+
"grad_norm": 6.909919738769531,
|
| 567 |
+
"learning_rate": 1.4467592592592593e-05,
|
| 568 |
+
"loss": 1.2479,
|
| 569 |
+
"step": 710
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.461139896373057,
|
| 573 |
+
"grad_norm": 5.200183868408203,
|
| 574 |
+
"learning_rate": 1.388888888888889e-05,
|
| 575 |
+
"loss": 1.2062,
|
| 576 |
+
"step": 720
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 7.564766839378239,
|
| 580 |
+
"grad_norm": 6.757835865020752,
|
| 581 |
+
"learning_rate": 1.3310185185185186e-05,
|
| 582 |
+
"loss": 1.2496,
|
| 583 |
+
"step": 730
|
|
|
|
|
|
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 7.668393782383419,
|
| 587 |
+
"grad_norm": 5.478608131408691,
|
| 588 |
+
"learning_rate": 1.2731481481481482e-05,
|
| 589 |
+
"loss": 1.2493,
|
| 590 |
+
"step": 740
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 7.772020725388601,
|
| 594 |
+
"grad_norm": 7.153792858123779,
|
| 595 |
+
"learning_rate": 1.2152777777777779e-05,
|
| 596 |
+
"loss": 1.2315,
|
| 597 |
+
"step": 750
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 7.875647668393782,
|
| 601 |
+
"grad_norm": 5.429667949676514,
|
| 602 |
+
"learning_rate": 1.1574074074074075e-05,
|
| 603 |
+
"loss": 1.1949,
|
| 604 |
+
"step": 760
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 7.979274611398964,
|
| 608 |
+
"grad_norm": 5.698895454406738,
|
| 609 |
+
"learning_rate": 1.0995370370370372e-05,
|
| 610 |
+
"loss": 1.2091,
|
| 611 |
+
"step": 770
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 8.0,
|
| 615 |
+
"eval_accuracy": 0.6404845454982454,
|
| 616 |
+
"eval_loss": 1.1456378698349,
|
| 617 |
+
"eval_runtime": 3647.7317,
|
| 618 |
+
"eval_samples_per_second": 5.703,
|
| 619 |
+
"eval_steps_per_second": 0.089,
|
| 620 |
+
"step": 772
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 8.082901554404145,
|
| 624 |
+
"grad_norm": 5.858074188232422,
|
| 625 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 626 |
+
"loss": 1.2231,
|
| 627 |
+
"step": 780
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 8.186528497409327,
|
| 631 |
+
"grad_norm": 5.4128594398498535,
|
| 632 |
+
"learning_rate": 9.837962962962964e-06,
|
| 633 |
+
"loss": 1.1949,
|
| 634 |
+
"step": 790
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 8.290155440414507,
|
| 638 |
+
"grad_norm": 5.688471794128418,
|
| 639 |
+
"learning_rate": 9.259259259259259e-06,
|
| 640 |
+
"loss": 1.2229,
|
| 641 |
+
"step": 800
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.393782383419689,
|
| 645 |
+
"grad_norm": 5.991612434387207,
|
| 646 |
+
"learning_rate": 8.680555555555556e-06,
|
| 647 |
+
"loss": 1.2234,
|
| 648 |
+
"step": 810
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 8.49740932642487,
|
| 652 |
+
"grad_norm": 5.711306095123291,
|
| 653 |
+
"learning_rate": 8.101851851851852e-06,
|
| 654 |
+
"loss": 1.1823,
|
| 655 |
+
"step": 820
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 8.601036269430052,
|
| 659 |
+
"grad_norm": 6.6796441078186035,
|
| 660 |
+
"learning_rate": 7.523148148148148e-06,
|
| 661 |
+
"loss": 1.1891,
|
| 662 |
+
"step": 830
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 8.704663212435234,
|
| 666 |
+
"grad_norm": 6.775707721710205,
|
| 667 |
+
"learning_rate": 6.944444444444445e-06,
|
| 668 |
+
"loss": 1.1872,
|
| 669 |
+
"step": 840
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 8.808290155440414,
|
| 673 |
+
"grad_norm": 5.822583198547363,
|
| 674 |
+
"learning_rate": 6.365740740740741e-06,
|
| 675 |
+
"loss": 1.1953,
|
| 676 |
+
"step": 850
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 8.911917098445596,
|
| 680 |
+
"grad_norm": 4.7059478759765625,
|
| 681 |
+
"learning_rate": 5.787037037037038e-06,
|
| 682 |
+
"loss": 1.2191,
|
| 683 |
+
"step": 860
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 8.994818652849741,
|
| 687 |
+
"eval_accuracy": 0.6431283949430371,
|
| 688 |
+
"eval_loss": 1.1358870267868042,
|
| 689 |
+
"eval_runtime": 3765.2145,
|
| 690 |
+
"eval_samples_per_second": 5.525,
|
| 691 |
+
"eval_steps_per_second": 0.087,
|
| 692 |
+
"step": 868
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 9.015544041450777,
|
| 696 |
+
"grad_norm": 6.1211371421813965,
|
| 697 |
+
"learning_rate": 5.208333333333334e-06,
|
| 698 |
+
"loss": 1.2054,
|
| 699 |
+
"step": 870
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 9.119170984455959,
|
| 703 |
+
"grad_norm": 6.023680686950684,
|
| 704 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 705 |
+
"loss": 1.2076,
|
| 706 |
+
"step": 880
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 9.22279792746114,
|
| 710 |
+
"grad_norm": 5.6407389640808105,
|
| 711 |
+
"learning_rate": 4.050925925925926e-06,
|
| 712 |
+
"loss": 1.2433,
|
| 713 |
+
"step": 890
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 9.32642487046632,
|
| 717 |
+
"grad_norm": 5.213710308074951,
|
| 718 |
+
"learning_rate": 3.4722222222222224e-06,
|
| 719 |
+
"loss": 1.2494,
|
| 720 |
+
"step": 900
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 9.430051813471502,
|
| 724 |
+
"grad_norm": 5.261003494262695,
|
| 725 |
+
"learning_rate": 2.893518518518519e-06,
|
| 726 |
+
"loss": 1.196,
|
| 727 |
+
"step": 910
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 9.533678756476684,
|
| 731 |
+
"grad_norm": 5.6940813064575195,
|
| 732 |
+
"learning_rate": 2.3148148148148148e-06,
|
| 733 |
+
"loss": 1.1802,
|
| 734 |
+
"step": 920
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 9.637305699481866,
|
| 738 |
+
"grad_norm": 5.532364368438721,
|
| 739 |
+
"learning_rate": 1.7361111111111112e-06,
|
| 740 |
+
"loss": 1.1724,
|
| 741 |
+
"step": 930
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 9.740932642487046,
|
| 745 |
+
"grad_norm": 7.711543083190918,
|
| 746 |
+
"learning_rate": 1.1574074074074074e-06,
|
| 747 |
+
"loss": 1.18,
|
| 748 |
+
"step": 940
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 9.844559585492227,
|
| 752 |
+
"grad_norm": 5.626430034637451,
|
| 753 |
+
"learning_rate": 5.787037037037037e-07,
|
| 754 |
+
"loss": 1.1592,
|
| 755 |
+
"step": 950
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 9.94818652849741,
|
| 759 |
+
"grad_norm": 6.389472007751465,
|
| 760 |
"learning_rate": 0.0,
|
| 761 |
+
"loss": 1.1983,
|
| 762 |
+
"step": 960
|
| 763 |
},
|
| 764 |
{
|
| 765 |
+
"epoch": 9.94818652849741,
|
| 766 |
+
"eval_accuracy": 0.6431283949430371,
|
| 767 |
+
"eval_loss": 1.138728141784668,
|
| 768 |
+
"eval_runtime": 3898.9781,
|
| 769 |
+
"eval_samples_per_second": 5.336,
|
| 770 |
+
"eval_steps_per_second": 0.084,
|
| 771 |
+
"step": 960
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 9.94818652849741,
|
| 775 |
+
"step": 960,
|
| 776 |
+
"total_flos": 6.10407050089078e+18,
|
| 777 |
+
"train_loss": 1.4552634547154109,
|
| 778 |
+
"train_runtime": 78972.9917,
|
| 779 |
+
"train_samples_per_second": 3.124,
|
| 780 |
"train_steps_per_second": 0.012
|
| 781 |
}
|
| 782 |
],
|
| 783 |
"logging_steps": 10,
|
| 784 |
+
"max_steps": 960,
|
| 785 |
"num_input_tokens_seen": 0,
|
| 786 |
"num_train_epochs": 10,
|
| 787 |
"save_steps": 500,
|
|
|
|
| 797 |
"attributes": {}
|
| 798 |
}
|
| 799 |
},
|
| 800 |
+
"total_flos": 6.10407050089078e+18,
|
| 801 |
"train_batch_size": 64,
|
| 802 |
"trial_name": null,
|
| 803 |
"trial_params": null
|