Commit
•
ea3005b
1
Parent(s):
1e069dc
End of training
Browse files- README.md +1 -1
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/Aug12_23-27-04_EMIMDGXA100GPU1/events.out.tfevents.1723520140.EMIMDGXA100GPU1.77542.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +840 -114
README.md
CHANGED
|
@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [microsoft/swinv2-base-patch4-window8-256](https://huggingface.co/microsoft/swinv2-base-patch4-window8-256) on the imagefolder dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
- Accuracy: 0.7241
|
| 37 |
|
| 38 |
## Model description
|
|
|
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [microsoft/swinv2-base-patch4-window8-256](https://huggingface.co/microsoft/swinv2-base-patch4-window8-256) on the imagefolder dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 0.6211
|
| 36 |
- Accuracy: 0.7241
|
| 37 |
|
| 38 |
## Model description
|
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": 1.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.734513274336283,
|
| 3 |
+
"eval_accuracy": 0.7241379310344828,
|
| 4 |
+
"eval_loss": 0.6211036443710327,
|
| 5 |
+
"eval_runtime": 1.0425,
|
| 6 |
+
"eval_samples_per_second": 27.817,
|
| 7 |
+
"eval_steps_per_second": 27.817,
|
| 8 |
+
"total_flos": 3.444219560381645e+17,
|
| 9 |
+
"train_loss": 0.5698859515644256,
|
| 10 |
+
"train_runtime": 500.495,
|
| 11 |
+
"train_samples_per_second": 6.773,
|
| 12 |
+
"train_steps_per_second": 1.678
|
| 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": 29.734513274336283,
|
| 3 |
+
"eval_accuracy": 0.7241379310344828,
|
| 4 |
+
"eval_loss": 0.6211036443710327,
|
| 5 |
+
"eval_runtime": 1.0425,
|
| 6 |
+
"eval_samples_per_second": 27.817,
|
| 7 |
+
"eval_steps_per_second": 27.817
|
| 8 |
}
|
runs/Aug12_23-27-04_EMIMDGXA100GPU1/events.out.tfevents.1723520140.EMIMDGXA100GPU1.77542.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3beef34e09f4d296cbf4a57508e0fb553b20cb3328ec7a2ad2ee536db738d2c5
|
| 3 |
+
size 411
|
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": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.734513274336283,
|
| 3 |
+
"total_flos": 3.444219560381645e+17,
|
| 4 |
+
"train_loss": 0.5698859515644256,
|
| 5 |
+
"train_runtime": 500.495,
|
| 6 |
+
"train_samples_per_second": 6.773,
|
| 7 |
+
"train_steps_per_second": 1.678
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,159 +1,885 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "swinv2-base-patch4-window8-256/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": 23.57,
|
| 18 |
-
"step": 6
|
| 19 |
},
|
| 20 |
{
|
| 21 |
-
"epoch":
|
| 22 |
-
"grad_norm":
|
| 23 |
-
"learning_rate":
|
| 24 |
-
"loss": 0.
|
| 25 |
-
"step":
|
| 26 |
},
|
| 27 |
{
|
| 28 |
-
"epoch":
|
| 29 |
-
"eval_accuracy": 0.
|
| 30 |
-
"eval_loss": 0.
|
| 31 |
-
"eval_runtime":
|
| 32 |
-
"eval_samples_per_second":
|
| 33 |
-
"eval_steps_per_second":
|
| 34 |
-
"step":
|
| 35 |
},
|
| 36 |
{
|
| 37 |
-
"epoch":
|
| 38 |
-
"
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
"eval_steps_per_second": 22.214,
|
| 43 |
-
"step": 18
|
| 44 |
},
|
| 45 |
{
|
| 46 |
-
"epoch":
|
| 47 |
-
"grad_norm":
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss": 0.
|
| 50 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 4.0,
|
| 54 |
-
"eval_accuracy": 0.
|
| 55 |
-
"eval_loss": 0.
|
| 56 |
-
"eval_runtime":
|
| 57 |
-
"eval_samples_per_second":
|
| 58 |
-
"eval_steps_per_second":
|
| 59 |
-
"step":
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 4.8,
|
| 63 |
-
"grad_norm": 7.068347930908203,
|
| 64 |
-
"learning_rate": 2.7777777777777783e-06,
|
| 65 |
-
"loss": 0.6313,
|
| 66 |
-
"step": 30
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 4.
|
| 70 |
-
"
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
| 74 |
-
"eval_steps_per_second": 21.711,
|
| 75 |
-
"step": 31
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"
|
| 80 |
-
"
|
| 81 |
-
"
|
| 82 |
-
"
|
| 83 |
-
"eval_steps_per_second": 19.648,
|
| 84 |
-
"step": 37
|
| 85 |
},
|
| 86 |
{
|
| 87 |
-
"epoch":
|
| 88 |
-
"grad_norm":
|
| 89 |
-
"learning_rate":
|
| 90 |
-
"loss": 0.
|
| 91 |
-
"step":
|
| 92 |
},
|
| 93 |
{
|
| 94 |
-
"epoch":
|
| 95 |
-
"eval_accuracy": 0.
|
| 96 |
-
"eval_loss": 0.
|
| 97 |
-
"eval_runtime":
|
| 98 |
-
"eval_samples_per_second":
|
| 99 |
-
"eval_steps_per_second":
|
| 100 |
-
"step":
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch":
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate": 9.
|
| 106 |
-
"loss": 0.
|
| 107 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 8.0,
|
| 111 |
-
"eval_accuracy": 0.
|
| 112 |
-
"eval_loss": 0.
|
| 113 |
-
"eval_runtime":
|
| 114 |
-
"eval_samples_per_second":
|
| 115 |
-
"eval_steps_per_second":
|
| 116 |
-
"step":
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 8.
|
| 120 |
-
"
|
| 121 |
-
"
|
| 122 |
-
"
|
| 123 |
-
"
|
| 124 |
-
|
| 125 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 126 |
},
|
| 127 |
{
|
| 128 |
-
"epoch":
|
| 129 |
-
"grad_norm":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 130 |
"learning_rate": 0.0,
|
| 131 |
-
"loss": 0.
|
| 132 |
-
"step":
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"eval_accuracy": 0.
|
| 137 |
-
"eval_loss": 0.
|
| 138 |
-
"eval_runtime":
|
| 139 |
-
"eval_samples_per_second":
|
| 140 |
-
"eval_steps_per_second":
|
| 141 |
-
"step":
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"step":
|
| 146 |
-
"total_flos":
|
| 147 |
-
"train_loss": 0.
|
| 148 |
-
"train_runtime":
|
| 149 |
-
"train_samples_per_second":
|
| 150 |
-
"train_steps_per_second": 1.
|
| 151 |
}
|
| 152 |
],
|
| 153 |
"logging_steps": 10,
|
| 154 |
-
"max_steps":
|
| 155 |
"num_input_tokens_seen": 0,
|
| 156 |
-
"num_train_epochs":
|
| 157 |
"save_steps": 500,
|
| 158 |
"stateful_callbacks": {
|
| 159 |
"TrainerControl": {
|
|
@@ -167,8 +893,8 @@
|
|
| 167 |
"attributes": {}
|
| 168 |
}
|
| 169 |
},
|
| 170 |
-
"total_flos":
|
| 171 |
-
"train_batch_size":
|
| 172 |
"trial_name": null,
|
| 173 |
"trial_params": null
|
| 174 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.7241379310344828,
|
| 3 |
+
"best_model_checkpoint": "swinv2-base-patch4-window8-256/checkpoint-28",
|
| 4 |
+
"epoch": 29.734513274336283,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 840,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.35398230088495575,
|
| 13 |
+
"grad_norm": 9.937736511230469,
|
| 14 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 15 |
+
"loss": 0.6341,
|
| 16 |
+
"step": 10
|
|
|
|
|
|
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.7079646017699115,
|
| 20 |
+
"grad_norm": 7.19003438949585,
|
| 21 |
+
"learning_rate": 2.380952380952381e-05,
|
| 22 |
+
"loss": 0.5428,
|
| 23 |
+
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.9911504424778761,
|
| 27 |
+
"eval_accuracy": 0.7241379310344828,
|
| 28 |
+
"eval_loss": 0.6211036443710327,
|
| 29 |
+
"eval_runtime": 1.0465,
|
| 30 |
+
"eval_samples_per_second": 27.71,
|
| 31 |
+
"eval_steps_per_second": 27.71,
|
| 32 |
+
"step": 28
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 1.0619469026548674,
|
| 36 |
+
"grad_norm": 9.382635116577148,
|
| 37 |
+
"learning_rate": 3.571428571428572e-05,
|
| 38 |
+
"loss": 0.5783,
|
| 39 |
+
"step": 30
|
|
|
|
|
|
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 1.415929203539823,
|
| 43 |
+
"grad_norm": 6.914736270904541,
|
| 44 |
+
"learning_rate": 4.761904761904762e-05,
|
| 45 |
+
"loss": 0.5546,
|
| 46 |
+
"step": 40
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.7699115044247788,
|
| 50 |
+
"grad_norm": 5.816603183746338,
|
| 51 |
+
"learning_rate": 5.9523809523809524e-05,
|
| 52 |
+
"loss": 0.6494,
|
| 53 |
+
"step": 50
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.9823008849557522,
|
| 57 |
+
"eval_accuracy": 0.7241379310344828,
|
| 58 |
+
"eval_loss": 0.6130104064941406,
|
| 59 |
+
"eval_runtime": 1.0448,
|
| 60 |
+
"eval_samples_per_second": 27.757,
|
| 61 |
+
"eval_steps_per_second": 27.757,
|
| 62 |
+
"step": 56
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 2.1238938053097347,
|
| 66 |
+
"grad_norm": 1.15414559841156,
|
| 67 |
+
"learning_rate": 7.142857142857143e-05,
|
| 68 |
+
"loss": 0.6304,
|
| 69 |
+
"step": 60
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.47787610619469,
|
| 73 |
+
"grad_norm": 3.671969175338745,
|
| 74 |
+
"learning_rate": 8.333333333333334e-05,
|
| 75 |
+
"loss": 0.6236,
|
| 76 |
+
"step": 70
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.831858407079646,
|
| 80 |
+
"grad_norm": 7.640573501586914,
|
| 81 |
+
"learning_rate": 9.523809523809524e-05,
|
| 82 |
+
"loss": 0.5752,
|
| 83 |
+
"step": 80
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 2.9734513274336285,
|
| 87 |
+
"eval_accuracy": 0.7241379310344828,
|
| 88 |
+
"eval_loss": 0.6845608949661255,
|
| 89 |
+
"eval_runtime": 1.0508,
|
| 90 |
+
"eval_samples_per_second": 27.597,
|
| 91 |
+
"eval_steps_per_second": 27.597,
|
| 92 |
+
"step": 84
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 3.185840707964602,
|
| 96 |
+
"grad_norm": 1.3297613859176636,
|
| 97 |
+
"learning_rate": 9.920634920634922e-05,
|
| 98 |
+
"loss": 0.6231,
|
| 99 |
+
"step": 90
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 3.5398230088495577,
|
| 103 |
+
"grad_norm": 1.8182591199874878,
|
| 104 |
+
"learning_rate": 9.78835978835979e-05,
|
| 105 |
+
"loss": 0.6603,
|
| 106 |
+
"step": 100
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 3.893805309734513,
|
| 110 |
+
"grad_norm": 0.6220849752426147,
|
| 111 |
+
"learning_rate": 9.656084656084657e-05,
|
| 112 |
+
"loss": 0.7165,
|
| 113 |
+
"step": 110
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"epoch": 4.0,
|
| 117 |
+
"eval_accuracy": 0.7241379310344828,
|
| 118 |
+
"eval_loss": 0.964207112789154,
|
| 119 |
+
"eval_runtime": 2.4229,
|
| 120 |
+
"eval_samples_per_second": 11.969,
|
| 121 |
+
"eval_steps_per_second": 11.969,
|
| 122 |
+
"step": 113
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 4.247787610619469,
|
| 126 |
+
"grad_norm": 0.8275606036186218,
|
| 127 |
+
"learning_rate": 9.523809523809524e-05,
|
| 128 |
+
"loss": 0.6544,
|
| 129 |
+
"step": 120
|
|
|
|
|
|
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 4.601769911504425,
|
| 133 |
+
"grad_norm": 5.596276760101318,
|
| 134 |
+
"learning_rate": 9.391534391534393e-05,
|
| 135 |
+
"loss": 0.6842,
|
| 136 |
+
"step": 130
|
|
|
|
|
|
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 4.95575221238938,
|
| 140 |
+
"grad_norm": 0.6999258995056152,
|
| 141 |
+
"learning_rate": 9.25925925925926e-05,
|
| 142 |
+
"loss": 0.5699,
|
| 143 |
+
"step": 140
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 4.991150442477876,
|
| 147 |
+
"eval_accuracy": 0.7241379310344828,
|
| 148 |
+
"eval_loss": 0.6072261929512024,
|
| 149 |
+
"eval_runtime": 1.0484,
|
| 150 |
+
"eval_samples_per_second": 27.662,
|
| 151 |
+
"eval_steps_per_second": 27.662,
|
| 152 |
+
"step": 141
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 5.3097345132743365,
|
| 156 |
+
"grad_norm": 5.042980194091797,
|
| 157 |
+
"learning_rate": 9.126984126984128e-05,
|
| 158 |
+
"loss": 0.6453,
|
| 159 |
+
"step": 150
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 5.663716814159292,
|
| 163 |
+
"grad_norm": 2.5722317695617676,
|
| 164 |
+
"learning_rate": 8.994708994708995e-05,
|
| 165 |
+
"loss": 0.5517,
|
| 166 |
+
"step": 160
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 5.982300884955752,
|
| 170 |
+
"eval_accuracy": 0.7241379310344828,
|
| 171 |
+
"eval_loss": 0.6231208443641663,
|
| 172 |
+
"eval_runtime": 1.0298,
|
| 173 |
+
"eval_samples_per_second": 28.161,
|
| 174 |
+
"eval_steps_per_second": 28.161,
|
| 175 |
+
"step": 169
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 6.017699115044247,
|
| 179 |
+
"grad_norm": 0.5366746783256531,
|
| 180 |
+
"learning_rate": 8.862433862433864e-05,
|
| 181 |
+
"loss": 0.5733,
|
| 182 |
+
"step": 170
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 6.371681415929204,
|
| 186 |
+
"grad_norm": 0.9455431699752808,
|
| 187 |
+
"learning_rate": 8.730158730158731e-05,
|
| 188 |
+
"loss": 0.6246,
|
| 189 |
+
"step": 180
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 6.725663716814159,
|
| 193 |
+
"grad_norm": 1.3559240102767944,
|
| 194 |
+
"learning_rate": 8.597883597883598e-05,
|
| 195 |
+
"loss": 0.5268,
|
| 196 |
+
"step": 190
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 6.9734513274336285,
|
| 200 |
+
"eval_accuracy": 0.7241379310344828,
|
| 201 |
+
"eval_loss": 0.6098366379737854,
|
| 202 |
+
"eval_runtime": 1.0741,
|
| 203 |
+
"eval_samples_per_second": 27.0,
|
| 204 |
+
"eval_steps_per_second": 27.0,
|
| 205 |
+
"step": 197
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 7.079646017699115,
|
| 209 |
+
"grad_norm": 0.570971667766571,
|
| 210 |
+
"learning_rate": 8.465608465608466e-05,
|
| 211 |
+
"loss": 0.5397,
|
| 212 |
+
"step": 200
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 7.433628318584071,
|
| 216 |
+
"grad_norm": 1.1077288389205933,
|
| 217 |
+
"learning_rate": 8.333333333333334e-05,
|
| 218 |
+
"loss": 0.4539,
|
| 219 |
+
"step": 210
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 7.787610619469026,
|
| 223 |
+
"grad_norm": 1.5305235385894775,
|
| 224 |
+
"learning_rate": 8.201058201058202e-05,
|
| 225 |
+
"loss": 0.672,
|
| 226 |
+
"step": 220
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 8.0,
|
| 230 |
+
"eval_accuracy": 0.7241379310344828,
|
| 231 |
+
"eval_loss": 0.5890592932701111,
|
| 232 |
+
"eval_runtime": 1.1176,
|
| 233 |
+
"eval_samples_per_second": 25.948,
|
| 234 |
+
"eval_steps_per_second": 25.948,
|
| 235 |
+
"step": 226
|
| 236 |
},
|
| 237 |
{
|
| 238 |
+
"epoch": 8.141592920353983,
|
| 239 |
+
"grad_norm": 1.5520473718643188,
|
| 240 |
+
"learning_rate": 8.068783068783069e-05,
|
| 241 |
+
"loss": 0.5567,
|
| 242 |
+
"step": 230
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 8.495575221238939,
|
| 246 |
+
"grad_norm": 1.5060292482376099,
|
| 247 |
+
"learning_rate": 7.936507936507937e-05,
|
| 248 |
+
"loss": 0.6923,
|
| 249 |
+
"step": 240
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 8.849557522123893,
|
| 253 |
+
"grad_norm": 0.2911463677883148,
|
| 254 |
+
"learning_rate": 7.804232804232805e-05,
|
| 255 |
+
"loss": 0.5448,
|
| 256 |
+
"step": 250
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 8.991150442477876,
|
| 260 |
+
"eval_accuracy": 0.7241379310344828,
|
| 261 |
+
"eval_loss": 0.6022756695747375,
|
| 262 |
+
"eval_runtime": 1.0313,
|
| 263 |
+
"eval_samples_per_second": 28.12,
|
| 264 |
+
"eval_steps_per_second": 28.12,
|
| 265 |
+
"step": 254
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 9.20353982300885,
|
| 269 |
+
"grad_norm": 0.38871487975120544,
|
| 270 |
+
"learning_rate": 7.671957671957673e-05,
|
| 271 |
+
"loss": 0.4555,
|
| 272 |
+
"step": 260
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 9.557522123893806,
|
| 276 |
+
"grad_norm": 1.8789817094802856,
|
| 277 |
+
"learning_rate": 7.53968253968254e-05,
|
| 278 |
+
"loss": 0.7061,
|
| 279 |
+
"step": 270
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 9.91150442477876,
|
| 283 |
+
"grad_norm": 1.5270930528640747,
|
| 284 |
+
"learning_rate": 7.407407407407407e-05,
|
| 285 |
+
"loss": 0.555,
|
| 286 |
+
"step": 280
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 9.982300884955752,
|
| 290 |
+
"eval_accuracy": 0.7241379310344828,
|
| 291 |
+
"eval_loss": 0.5916692614555359,
|
| 292 |
+
"eval_runtime": 1.0352,
|
| 293 |
+
"eval_samples_per_second": 28.013,
|
| 294 |
+
"eval_steps_per_second": 28.013,
|
| 295 |
+
"step": 282
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 10.265486725663717,
|
| 299 |
+
"grad_norm": 1.4813481569290161,
|
| 300 |
+
"learning_rate": 7.275132275132276e-05,
|
| 301 |
+
"loss": 0.5704,
|
| 302 |
+
"step": 290
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 10.619469026548673,
|
| 306 |
+
"grad_norm": 1.4199880361557007,
|
| 307 |
+
"learning_rate": 7.142857142857143e-05,
|
| 308 |
+
"loss": 0.5333,
|
| 309 |
+
"step": 300
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 10.973451327433628,
|
| 313 |
+
"grad_norm": 1.5007575750350952,
|
| 314 |
+
"learning_rate": 7.010582010582011e-05,
|
| 315 |
+
"loss": 0.5818,
|
| 316 |
+
"step": 310
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 10.973451327433628,
|
| 320 |
+
"eval_accuracy": 0.7241379310344828,
|
| 321 |
+
"eval_loss": 0.5940178632736206,
|
| 322 |
+
"eval_runtime": 1.0438,
|
| 323 |
+
"eval_samples_per_second": 27.784,
|
| 324 |
+
"eval_steps_per_second": 27.784,
|
| 325 |
+
"step": 310
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 11.327433628318584,
|
| 329 |
+
"grad_norm": 0.39495939016342163,
|
| 330 |
+
"learning_rate": 6.878306878306878e-05,
|
| 331 |
+
"loss": 0.5062,
|
| 332 |
+
"step": 320
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 11.68141592920354,
|
| 336 |
+
"grad_norm": 0.27960801124572754,
|
| 337 |
+
"learning_rate": 6.746031746031747e-05,
|
| 338 |
+
"loss": 0.6556,
|
| 339 |
+
"step": 330
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 12.0,
|
| 343 |
+
"eval_accuracy": 0.7241379310344828,
|
| 344 |
+
"eval_loss": 0.5965593457221985,
|
| 345 |
+
"eval_runtime": 1.0472,
|
| 346 |
+
"eval_samples_per_second": 27.693,
|
| 347 |
+
"eval_steps_per_second": 27.693,
|
| 348 |
+
"step": 339
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 12.035398230088495,
|
| 352 |
+
"grad_norm": 1.464916467666626,
|
| 353 |
+
"learning_rate": 6.613756613756614e-05,
|
| 354 |
+
"loss": 0.4407,
|
| 355 |
+
"step": 340
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 12.389380530973451,
|
| 359 |
+
"grad_norm": 0.8306865096092224,
|
| 360 |
+
"learning_rate": 6.481481481481482e-05,
|
| 361 |
+
"loss": 0.5427,
|
| 362 |
+
"step": 350
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 12.743362831858407,
|
| 366 |
+
"grad_norm": 1.798261284828186,
|
| 367 |
+
"learning_rate": 6.349206349206349e-05,
|
| 368 |
+
"loss": 0.716,
|
| 369 |
+
"step": 360
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 12.991150442477876,
|
| 373 |
+
"eval_accuracy": 0.7241379310344828,
|
| 374 |
+
"eval_loss": 0.5903679132461548,
|
| 375 |
+
"eval_runtime": 1.0433,
|
| 376 |
+
"eval_samples_per_second": 27.795,
|
| 377 |
+
"eval_steps_per_second": 27.795,
|
| 378 |
+
"step": 367
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 13.097345132743364,
|
| 382 |
+
"grad_norm": 0.575097918510437,
|
| 383 |
+
"learning_rate": 6.216931216931218e-05,
|
| 384 |
+
"loss": 0.4514,
|
| 385 |
+
"step": 370
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 13.451327433628318,
|
| 389 |
+
"grad_norm": 1.393051028251648,
|
| 390 |
+
"learning_rate": 6.084656084656085e-05,
|
| 391 |
+
"loss": 0.5387,
|
| 392 |
+
"step": 380
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 13.805309734513274,
|
| 396 |
+
"grad_norm": 2.5438356399536133,
|
| 397 |
+
"learning_rate": 5.9523809523809524e-05,
|
| 398 |
+
"loss": 0.6104,
|
| 399 |
+
"step": 390
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 13.982300884955752,
|
| 403 |
+
"eval_accuracy": 0.7241379310344828,
|
| 404 |
+
"eval_loss": 0.5938333868980408,
|
| 405 |
+
"eval_runtime": 1.0518,
|
| 406 |
+
"eval_samples_per_second": 27.571,
|
| 407 |
+
"eval_steps_per_second": 27.571,
|
| 408 |
+
"step": 395
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 14.15929203539823,
|
| 412 |
+
"grad_norm": 1.7508420944213867,
|
| 413 |
+
"learning_rate": 5.82010582010582e-05,
|
| 414 |
+
"loss": 0.5812,
|
| 415 |
+
"step": 400
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 14.513274336283185,
|
| 419 |
+
"grad_norm": 0.19824433326721191,
|
| 420 |
+
"learning_rate": 5.6878306878306885e-05,
|
| 421 |
+
"loss": 0.5678,
|
| 422 |
+
"step": 410
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 14.867256637168142,
|
| 426 |
+
"grad_norm": 0.09461919218301773,
|
| 427 |
+
"learning_rate": 5.555555555555556e-05,
|
| 428 |
+
"loss": 0.5046,
|
| 429 |
+
"step": 420
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 14.973451327433628,
|
| 433 |
+
"eval_accuracy": 0.7241379310344828,
|
| 434 |
+
"eval_loss": 0.5921294689178467,
|
| 435 |
+
"eval_runtime": 1.0718,
|
| 436 |
+
"eval_samples_per_second": 27.057,
|
| 437 |
+
"eval_steps_per_second": 27.057,
|
| 438 |
+
"step": 423
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 15.221238938053098,
|
| 442 |
+
"grad_norm": 0.46888694167137146,
|
| 443 |
+
"learning_rate": 5.423280423280423e-05,
|
| 444 |
+
"loss": 0.5344,
|
| 445 |
+
"step": 430
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 15.575221238938052,
|
| 449 |
+
"grad_norm": 2.1239452362060547,
|
| 450 |
+
"learning_rate": 5.291005291005291e-05,
|
| 451 |
+
"loss": 0.4711,
|
| 452 |
+
"step": 440
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 15.929203539823009,
|
| 456 |
+
"grad_norm": 2.082629442214966,
|
| 457 |
+
"learning_rate": 5.158730158730159e-05,
|
| 458 |
+
"loss": 0.5871,
|
| 459 |
+
"step": 450
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 16.0,
|
| 463 |
+
"eval_accuracy": 0.7241379310344828,
|
| 464 |
+
"eval_loss": 0.6027411818504333,
|
| 465 |
+
"eval_runtime": 1.0332,
|
| 466 |
+
"eval_samples_per_second": 28.069,
|
| 467 |
+
"eval_steps_per_second": 28.069,
|
| 468 |
+
"step": 452
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 16.283185840707965,
|
| 472 |
+
"grad_norm": 3.216160535812378,
|
| 473 |
+
"learning_rate": 5.026455026455027e-05,
|
| 474 |
+
"loss": 0.7275,
|
| 475 |
+
"step": 460
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 16.63716814159292,
|
| 479 |
+
"grad_norm": 0.5754966139793396,
|
| 480 |
+
"learning_rate": 4.894179894179895e-05,
|
| 481 |
+
"loss": 0.503,
|
| 482 |
+
"step": 470
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 16.991150442477878,
|
| 486 |
+
"grad_norm": 1.4574097394943237,
|
| 487 |
+
"learning_rate": 4.761904761904762e-05,
|
| 488 |
+
"loss": 0.5222,
|
| 489 |
+
"step": 480
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 16.991150442477878,
|
| 493 |
+
"eval_accuracy": 0.7241379310344828,
|
| 494 |
+
"eval_loss": 0.5921464562416077,
|
| 495 |
+
"eval_runtime": 1.0314,
|
| 496 |
+
"eval_samples_per_second": 28.118,
|
| 497 |
+
"eval_steps_per_second": 28.118,
|
| 498 |
+
"step": 480
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 17.345132743362832,
|
| 502 |
+
"grad_norm": 0.36778560280799866,
|
| 503 |
+
"learning_rate": 4.62962962962963e-05,
|
| 504 |
+
"loss": 0.5548,
|
| 505 |
+
"step": 490
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 17.699115044247787,
|
| 509 |
+
"grad_norm": 1.3566818237304688,
|
| 510 |
+
"learning_rate": 4.4973544973544974e-05,
|
| 511 |
+
"loss": 0.5511,
|
| 512 |
+
"step": 500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 17.98230088495575,
|
| 516 |
+
"eval_accuracy": 0.7241379310344828,
|
| 517 |
+
"eval_loss": 0.5947627425193787,
|
| 518 |
+
"eval_runtime": 1.0877,
|
| 519 |
+
"eval_samples_per_second": 26.661,
|
| 520 |
+
"eval_steps_per_second": 26.661,
|
| 521 |
+
"step": 508
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 18.053097345132745,
|
| 525 |
+
"grad_norm": 3.3262369632720947,
|
| 526 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 527 |
+
"loss": 0.6192,
|
| 528 |
+
"step": 510
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 18.4070796460177,
|
| 532 |
+
"grad_norm": 0.4368630051612854,
|
| 533 |
+
"learning_rate": 4.232804232804233e-05,
|
| 534 |
+
"loss": 0.4436,
|
| 535 |
+
"step": 520
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 18.761061946902654,
|
| 539 |
+
"grad_norm": 0.6899360418319702,
|
| 540 |
+
"learning_rate": 4.100529100529101e-05,
|
| 541 |
+
"loss": 0.6394,
|
| 542 |
+
"step": 530
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 18.97345132743363,
|
| 546 |
+
"eval_accuracy": 0.7241379310344828,
|
| 547 |
+
"eval_loss": 0.5969327688217163,
|
| 548 |
+
"eval_runtime": 1.0482,
|
| 549 |
+
"eval_samples_per_second": 27.667,
|
| 550 |
+
"eval_steps_per_second": 27.667,
|
| 551 |
+
"step": 536
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 19.115044247787612,
|
| 555 |
+
"grad_norm": 0.5160775780677795,
|
| 556 |
+
"learning_rate": 3.968253968253968e-05,
|
| 557 |
+
"loss": 0.5015,
|
| 558 |
+
"step": 540
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 19.469026548672566,
|
| 562 |
+
"grad_norm": 1.3984384536743164,
|
| 563 |
+
"learning_rate": 3.835978835978836e-05,
|
| 564 |
+
"loss": 0.6334,
|
| 565 |
+
"step": 550
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 19.82300884955752,
|
| 569 |
+
"grad_norm": 0.24300755560398102,
|
| 570 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 571 |
+
"loss": 0.566,
|
| 572 |
+
"step": 560
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 20.0,
|
| 576 |
+
"eval_accuracy": 0.7241379310344828,
|
| 577 |
+
"eval_loss": 0.6005056500434875,
|
| 578 |
+
"eval_runtime": 1.0644,
|
| 579 |
+
"eval_samples_per_second": 27.247,
|
| 580 |
+
"eval_steps_per_second": 27.247,
|
| 581 |
+
"step": 565
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 20.17699115044248,
|
| 585 |
+
"grad_norm": 1.3175139427185059,
|
| 586 |
+
"learning_rate": 3.571428571428572e-05,
|
| 587 |
+
"loss": 0.4341,
|
| 588 |
+
"step": 570
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 20.530973451327434,
|
| 592 |
+
"grad_norm": 1.367430567741394,
|
| 593 |
+
"learning_rate": 3.439153439153439e-05,
|
| 594 |
+
"loss": 0.5373,
|
| 595 |
+
"step": 580
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 20.884955752212388,
|
| 599 |
+
"grad_norm": 0.3279801905155182,
|
| 600 |
+
"learning_rate": 3.306878306878307e-05,
|
| 601 |
+
"loss": 0.6032,
|
| 602 |
+
"step": 590
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 20.991150442477878,
|
| 606 |
+
"eval_accuracy": 0.7241379310344828,
|
| 607 |
+
"eval_loss": 0.5968325138092041,
|
| 608 |
+
"eval_runtime": 1.0696,
|
| 609 |
+
"eval_samples_per_second": 27.113,
|
| 610 |
+
"eval_steps_per_second": 27.113,
|
| 611 |
+
"step": 593
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 21.238938053097346,
|
| 615 |
+
"grad_norm": 3.318119764328003,
|
| 616 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 617 |
+
"loss": 0.5679,
|
| 618 |
+
"step": 600
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 21.5929203539823,
|
| 622 |
+
"grad_norm": 1.7342548370361328,
|
| 623 |
+
"learning_rate": 3.0423280423280425e-05,
|
| 624 |
+
"loss": 0.6877,
|
| 625 |
+
"step": 610
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 21.946902654867255,
|
| 629 |
+
"grad_norm": 0.20895572006702423,
|
| 630 |
+
"learning_rate": 2.91005291005291e-05,
|
| 631 |
+
"loss": 0.4824,
|
| 632 |
+
"step": 620
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 21.98230088495575,
|
| 636 |
+
"eval_accuracy": 0.7241379310344828,
|
| 637 |
+
"eval_loss": 0.5933610200881958,
|
| 638 |
+
"eval_runtime": 1.0375,
|
| 639 |
+
"eval_samples_per_second": 27.951,
|
| 640 |
+
"eval_steps_per_second": 27.951,
|
| 641 |
+
"step": 621
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 22.300884955752213,
|
| 645 |
+
"grad_norm": 0.5702241659164429,
|
| 646 |
+
"learning_rate": 2.777777777777778e-05,
|
| 647 |
+
"loss": 0.5076,
|
| 648 |
+
"step": 630
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 22.654867256637168,
|
| 652 |
+
"grad_norm": 1.405441164970398,
|
| 653 |
+
"learning_rate": 2.6455026455026456e-05,
|
| 654 |
+
"loss": 0.4975,
|
| 655 |
+
"step": 640
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 22.97345132743363,
|
| 659 |
+
"eval_accuracy": 0.7241379310344828,
|
| 660 |
+
"eval_loss": 0.5979170799255371,
|
| 661 |
+
"eval_runtime": 1.0599,
|
| 662 |
+
"eval_samples_per_second": 27.362,
|
| 663 |
+
"eval_steps_per_second": 27.362,
|
| 664 |
+
"step": 649
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 23.008849557522122,
|
| 668 |
+
"grad_norm": 1.4034713506698608,
|
| 669 |
+
"learning_rate": 2.5132275132275137e-05,
|
| 670 |
+
"loss": 0.5977,
|
| 671 |
+
"step": 650
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 23.36283185840708,
|
| 675 |
+
"grad_norm": 1.3884029388427734,
|
| 676 |
+
"learning_rate": 2.380952380952381e-05,
|
| 677 |
+
"loss": 0.4975,
|
| 678 |
+
"step": 660
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 23.716814159292035,
|
| 682 |
+
"grad_norm": 0.6450229287147522,
|
| 683 |
+
"learning_rate": 2.2486772486772487e-05,
|
| 684 |
+
"loss": 0.4976,
|
| 685 |
+
"step": 670
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 24.0,
|
| 689 |
+
"eval_accuracy": 0.7241379310344828,
|
| 690 |
+
"eval_loss": 0.603380560874939,
|
| 691 |
+
"eval_runtime": 1.0349,
|
| 692 |
+
"eval_samples_per_second": 28.023,
|
| 693 |
+
"eval_steps_per_second": 28.023,
|
| 694 |
+
"step": 678
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 24.07079646017699,
|
| 698 |
+
"grad_norm": 0.3577538728713989,
|
| 699 |
+
"learning_rate": 2.1164021164021164e-05,
|
| 700 |
+
"loss": 0.7906,
|
| 701 |
+
"step": 680
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 24.424778761061948,
|
| 705 |
+
"grad_norm": 0.5209062099456787,
|
| 706 |
+
"learning_rate": 1.984126984126984e-05,
|
| 707 |
+
"loss": 0.4288,
|
| 708 |
+
"step": 690
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 24.778761061946902,
|
| 712 |
+
"grad_norm": 1.9989219903945923,
|
| 713 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 714 |
+
"loss": 0.5355,
|
| 715 |
+
"step": 700
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 24.991150442477878,
|
| 719 |
+
"eval_accuracy": 0.7241379310344828,
|
| 720 |
+
"eval_loss": 0.6033223867416382,
|
| 721 |
+
"eval_runtime": 1.0457,
|
| 722 |
+
"eval_samples_per_second": 27.733,
|
| 723 |
+
"eval_steps_per_second": 27.733,
|
| 724 |
+
"step": 706
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 25.13274336283186,
|
| 728 |
+
"grad_norm": 1.9387425184249878,
|
| 729 |
+
"learning_rate": 1.7195767195767195e-05,
|
| 730 |
+
"loss": 0.6319,
|
| 731 |
+
"step": 710
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 25.486725663716815,
|
| 735 |
+
"grad_norm": 0.4894324541091919,
|
| 736 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 737 |
+
"loss": 0.5932,
|
| 738 |
+
"step": 720
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 25.84070796460177,
|
| 742 |
+
"grad_norm": 0.26004230976104736,
|
| 743 |
+
"learning_rate": 1.455026455026455e-05,
|
| 744 |
+
"loss": 0.4323,
|
| 745 |
+
"step": 730
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 25.98230088495575,
|
| 749 |
+
"eval_accuracy": 0.7241379310344828,
|
| 750 |
+
"eval_loss": 0.601476788520813,
|
| 751 |
+
"eval_runtime": 1.0397,
|
| 752 |
+
"eval_samples_per_second": 27.892,
|
| 753 |
+
"eval_steps_per_second": 27.892,
|
| 754 |
+
"step": 734
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 26.194690265486727,
|
| 758 |
+
"grad_norm": 1.3441038131713867,
|
| 759 |
+
"learning_rate": 1.3227513227513228e-05,
|
| 760 |
+
"loss": 0.6053,
|
| 761 |
+
"step": 740
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 26.548672566371682,
|
| 765 |
+
"grad_norm": 1.3334237337112427,
|
| 766 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 767 |
+
"loss": 0.5,
|
| 768 |
+
"step": 750
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 26.902654867256636,
|
| 772 |
+
"grad_norm": 0.3579448461532593,
|
| 773 |
+
"learning_rate": 1.0582010582010582e-05,
|
| 774 |
+
"loss": 0.5579,
|
| 775 |
+
"step": 760
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 26.97345132743363,
|
| 779 |
+
"eval_accuracy": 0.7241379310344828,
|
| 780 |
+
"eval_loss": 0.6042701601982117,
|
| 781 |
+
"eval_runtime": 1.0854,
|
| 782 |
+
"eval_samples_per_second": 26.718,
|
| 783 |
+
"eval_steps_per_second": 26.718,
|
| 784 |
+
"step": 762
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 27.256637168141594,
|
| 788 |
+
"grad_norm": 1.311233639717102,
|
| 789 |
+
"learning_rate": 9.259259259259259e-06,
|
| 790 |
+
"loss": 0.5794,
|
| 791 |
+
"step": 770
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 27.61061946902655,
|
| 795 |
+
"grad_norm": 1.3221209049224854,
|
| 796 |
+
"learning_rate": 7.936507936507936e-06,
|
| 797 |
+
"loss": 0.494,
|
| 798 |
+
"step": 780
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 27.964601769911503,
|
| 802 |
+
"grad_norm": 0.31315815448760986,
|
| 803 |
+
"learning_rate": 6.613756613756614e-06,
|
| 804 |
+
"loss": 0.5639,
|
| 805 |
+
"step": 790
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 28.0,
|
| 809 |
+
"eval_accuracy": 0.7241379310344828,
|
| 810 |
+
"eval_loss": 0.6023350358009338,
|
| 811 |
+
"eval_runtime": 1.0409,
|
| 812 |
+
"eval_samples_per_second": 27.86,
|
| 813 |
+
"eval_steps_per_second": 27.86,
|
| 814 |
+
"step": 791
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 28.31858407079646,
|
| 818 |
+
"grad_norm": 0.7090272307395935,
|
| 819 |
+
"learning_rate": 5.291005291005291e-06,
|
| 820 |
+
"loss": 0.5711,
|
| 821 |
+
"step": 800
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 28.672566371681416,
|
| 825 |
+
"grad_norm": 1.377335548400879,
|
| 826 |
+
"learning_rate": 3.968253968253968e-06,
|
| 827 |
+
"loss": 0.5595,
|
| 828 |
+
"step": 810
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 28.991150442477878,
|
| 832 |
+
"eval_accuracy": 0.7241379310344828,
|
| 833 |
+
"eval_loss": 0.5995790958404541,
|
| 834 |
+
"eval_runtime": 1.0359,
|
| 835 |
+
"eval_samples_per_second": 27.995,
|
| 836 |
+
"eval_steps_per_second": 27.995,
|
| 837 |
+
"step": 819
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 29.02654867256637,
|
| 841 |
+
"grad_norm": 1.294968605041504,
|
| 842 |
+
"learning_rate": 2.6455026455026455e-06,
|
| 843 |
+
"loss": 0.5339,
|
| 844 |
+
"step": 820
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 29.38053097345133,
|
| 848 |
+
"grad_norm": 1.297398328781128,
|
| 849 |
+
"learning_rate": 1.3227513227513228e-06,
|
| 850 |
+
"loss": 0.6959,
|
| 851 |
+
"step": 830
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 29.734513274336283,
|
| 855 |
+
"grad_norm": 1.3724257946014404,
|
| 856 |
"learning_rate": 0.0,
|
| 857 |
+
"loss": 0.4372,
|
| 858 |
+
"step": 840
|
| 859 |
},
|
| 860 |
{
|
| 861 |
+
"epoch": 29.734513274336283,
|
| 862 |
+
"eval_accuracy": 0.7241379310344828,
|
| 863 |
+
"eval_loss": 0.599529504776001,
|
| 864 |
+
"eval_runtime": 1.0824,
|
| 865 |
+
"eval_samples_per_second": 26.793,
|
| 866 |
+
"eval_steps_per_second": 26.793,
|
| 867 |
+
"step": 840
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 29.734513274336283,
|
| 871 |
+
"step": 840,
|
| 872 |
+
"total_flos": 3.444219560381645e+17,
|
| 873 |
+
"train_loss": 0.5698859515644256,
|
| 874 |
+
"train_runtime": 500.495,
|
| 875 |
+
"train_samples_per_second": 6.773,
|
| 876 |
+
"train_steps_per_second": 1.678
|
| 877 |
}
|
| 878 |
],
|
| 879 |
"logging_steps": 10,
|
| 880 |
+
"max_steps": 840,
|
| 881 |
"num_input_tokens_seen": 0,
|
| 882 |
+
"num_train_epochs": 30,
|
| 883 |
"save_steps": 500,
|
| 884 |
"stateful_callbacks": {
|
| 885 |
"TrainerControl": {
|
|
|
|
| 893 |
"attributes": {}
|
| 894 |
}
|
| 895 |
},
|
| 896 |
+
"total_flos": 3.444219560381645e+17,
|
| 897 |
+
"train_batch_size": 1,
|
| 898 |
"trial_name": null,
|
| 899 |
"trial_params": null
|
| 900 |
}
|