🍻 cheers
Browse files- README.md +6 -3
- all_results.json +10 -10
- eval_results.json +5 -5
- runs/Mar29_18-51-27_X5C922065N/events.out.tfevents.1711735614.X5C922065N.77198.3 +3 -0
- train_results.json +5 -5
- trainer_state.json +499 -11
README.md
CHANGED
|
@@ -1,6 +1,9 @@
|
|
| 1 |
---
|
| 2 |
base_model: d071696/vit-finetune-scrap
|
| 3 |
tags:
|
|
|
|
|
|
|
|
|
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- arrow
|
|
@@ -13,7 +16,7 @@ model-index:
|
|
| 13 |
name: Image Classification
|
| 14 |
type: image-classification
|
| 15 |
dataset:
|
| 16 |
-
name:
|
| 17 |
type: arrow
|
| 18 |
config: default
|
| 19 |
split: train
|
|
@@ -29,9 +32,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 29 |
|
| 30 |
# vit-finetune-scrap
|
| 31 |
|
| 32 |
-
This model is a fine-tuned version of [d071696/vit-finetune-scrap](https://huggingface.co/d071696/vit-finetune-scrap) on the
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
-
- Loss: 0.
|
| 35 |
- Accuracy: 0.9550
|
| 36 |
|
| 37 |
## Model description
|
|
|
|
| 1 |
---
|
| 2 |
base_model: d071696/vit-finetune-scrap
|
| 3 |
tags:
|
| 4 |
+
- image-classification
|
| 5 |
+
- image-feature-extraction
|
| 6 |
+
- image-to-text
|
| 7 |
- generated_from_trainer
|
| 8 |
datasets:
|
| 9 |
- arrow
|
|
|
|
| 16 |
name: Image Classification
|
| 17 |
type: image-classification
|
| 18 |
dataset:
|
| 19 |
+
name: d071696/scraps1
|
| 20 |
type: arrow
|
| 21 |
config: default
|
| 22 |
split: train
|
|
|
|
| 32 |
|
| 33 |
# vit-finetune-scrap
|
| 34 |
|
| 35 |
+
This model is a fine-tuned version of [d071696/vit-finetune-scrap](https://huggingface.co/d071696/vit-finetune-scrap) on the d071696/scraps1 dataset.
|
| 36 |
It achieves the following results on the evaluation set:
|
| 37 |
+
- Loss: 0.1588
|
| 38 |
- Accuracy: 0.9550
|
| 39 |
|
| 40 |
## Model description
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 1.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.954983922829582,
|
| 4 |
+
"eval_loss": 0.15877817571163177,
|
| 5 |
+
"eval_runtime": 9.6813,
|
| 6 |
+
"eval_samples_per_second": 64.248,
|
| 7 |
+
"eval_steps_per_second": 8.057,
|
| 8 |
+
"total_flos": 7.703325099767808e+17,
|
| 9 |
+
"train_loss": 0.15572628828410345,
|
| 10 |
+
"train_runtime": 552.98,
|
| 11 |
+
"train_samples_per_second": 17.975,
|
| 12 |
+
"train_steps_per_second": 1.128
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.954983922829582,
|
| 4 |
+
"eval_loss": 0.15877817571163177,
|
| 5 |
+
"eval_runtime": 9.6813,
|
| 6 |
+
"eval_samples_per_second": 64.248,
|
| 7 |
+
"eval_steps_per_second": 8.057
|
| 8 |
}
|
runs/Mar29_18-51-27_X5C922065N/events.out.tfevents.1711735614.X5C922065N.77198.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32614bdfe35502a4fcafcda07b5ed6f2d23ec281eb707971215e1acd15091a41
|
| 3 |
+
size 734
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"total_flos": 7.703325099767808e+17,
|
| 4 |
+
"train_loss": 0.15572628828410345,
|
| 5 |
+
"train_runtime": 552.98,
|
| 6 |
+
"train_samples_per_second": 17.975,
|
| 7 |
+
"train_steps_per_second": 1.128
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,29 +1,517 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint":
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
| 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": 4.0,
|
| 13 |
-
"step":
|
| 14 |
-
"total_flos":
|
| 15 |
-
"train_loss":
|
| 16 |
-
"train_runtime":
|
| 17 |
-
"train_samples_per_second":
|
| 18 |
-
"train_steps_per_second": 1.
|
| 19 |
}
|
| 20 |
],
|
| 21 |
"logging_steps": 10,
|
| 22 |
-
"max_steps":
|
| 23 |
"num_input_tokens_seen": 0,
|
| 24 |
"num_train_epochs": 4,
|
| 25 |
"save_steps": 100,
|
| 26 |
-
"total_flos":
|
| 27 |
"train_batch_size": 16,
|
| 28 |
"trial_name": null,
|
| 29 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.15877817571163177,
|
| 3 |
+
"best_model_checkpoint": "./vit-finetune-scrap/checkpoint-300",
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 624,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06,
|
| 13 |
+
"grad_norm": 2.869983673095703,
|
| 14 |
+
"learning_rate": 0.00019679487179487178,
|
| 15 |
+
"loss": 1.9747,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.13,
|
| 20 |
+
"grad_norm": 2.9758315086364746,
|
| 21 |
+
"learning_rate": 0.0001935897435897436,
|
| 22 |
+
"loss": 1.209,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.19,
|
| 27 |
+
"grad_norm": 3.3387157917022705,
|
| 28 |
+
"learning_rate": 0.00019038461538461538,
|
| 29 |
+
"loss": 0.7205,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.26,
|
| 34 |
+
"grad_norm": 2.921093463897705,
|
| 35 |
+
"learning_rate": 0.0001871794871794872,
|
| 36 |
+
"loss": 0.4159,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.32,
|
| 41 |
+
"grad_norm": 2.4197134971618652,
|
| 42 |
+
"learning_rate": 0.00018397435897435897,
|
| 43 |
+
"loss": 0.3879,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.38,
|
| 48 |
+
"grad_norm": 2.498006582260132,
|
| 49 |
+
"learning_rate": 0.00018076923076923077,
|
| 50 |
+
"loss": 0.2695,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.45,
|
| 55 |
+
"grad_norm": 0.9913655519485474,
|
| 56 |
+
"learning_rate": 0.00017756410256410257,
|
| 57 |
+
"loss": 0.33,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.51,
|
| 62 |
+
"grad_norm": 5.167428493499756,
|
| 63 |
+
"learning_rate": 0.00017435897435897436,
|
| 64 |
+
"loss": 0.2374,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.58,
|
| 69 |
+
"grad_norm": 5.158258438110352,
|
| 70 |
+
"learning_rate": 0.00017115384615384616,
|
| 71 |
+
"loss": 0.2531,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.64,
|
| 76 |
+
"grad_norm": 2.1700403690338135,
|
| 77 |
+
"learning_rate": 0.00016794871794871796,
|
| 78 |
+
"loss": 0.1672,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.64,
|
| 83 |
+
"eval_accuracy": 0.9485530546623794,
|
| 84 |
+
"eval_loss": 0.22496841847896576,
|
| 85 |
+
"eval_runtime": 11.2945,
|
| 86 |
+
"eval_samples_per_second": 55.071,
|
| 87 |
+
"eval_steps_per_second": 6.906,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.71,
|
| 92 |
+
"grad_norm": 4.994268894195557,
|
| 93 |
+
"learning_rate": 0.00016474358974358976,
|
| 94 |
+
"loss": 0.1319,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.77,
|
| 99 |
+
"grad_norm": 2.2666022777557373,
|
| 100 |
+
"learning_rate": 0.00016153846153846155,
|
| 101 |
+
"loss": 0.283,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.83,
|
| 106 |
+
"grad_norm": 3.1912319660186768,
|
| 107 |
+
"learning_rate": 0.00015833333333333332,
|
| 108 |
+
"loss": 0.1666,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.9,
|
| 113 |
+
"grad_norm": 3.1358578205108643,
|
| 114 |
+
"learning_rate": 0.00015512820512820515,
|
| 115 |
+
"loss": 0.2819,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.96,
|
| 120 |
+
"grad_norm": 2.050241470336914,
|
| 121 |
+
"learning_rate": 0.00015192307692307692,
|
| 122 |
+
"loss": 0.2874,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.03,
|
| 127 |
+
"grad_norm": 5.193918704986572,
|
| 128 |
+
"learning_rate": 0.00014871794871794872,
|
| 129 |
+
"loss": 0.2103,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.09,
|
| 134 |
+
"grad_norm": 0.405056893825531,
|
| 135 |
+
"learning_rate": 0.00014551282051282051,
|
| 136 |
+
"loss": 0.1818,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.15,
|
| 141 |
+
"grad_norm": 11.979884147644043,
|
| 142 |
+
"learning_rate": 0.0001423076923076923,
|
| 143 |
+
"loss": 0.1626,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.22,
|
| 148 |
+
"grad_norm": 0.3927968442440033,
|
| 149 |
+
"learning_rate": 0.0001391025641025641,
|
| 150 |
+
"loss": 0.0824,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.28,
|
| 155 |
+
"grad_norm": 0.11850783228874207,
|
| 156 |
+
"learning_rate": 0.0001358974358974359,
|
| 157 |
+
"loss": 0.1277,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.28,
|
| 162 |
+
"eval_accuracy": 0.9372990353697749,
|
| 163 |
+
"eval_loss": 0.24667073786258698,
|
| 164 |
+
"eval_runtime": 76.2101,
|
| 165 |
+
"eval_samples_per_second": 8.162,
|
| 166 |
+
"eval_steps_per_second": 1.023,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.35,
|
| 171 |
+
"grad_norm": 0.8870755434036255,
|
| 172 |
+
"learning_rate": 0.0001326923076923077,
|
| 173 |
+
"loss": 0.1308,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.41,
|
| 178 |
+
"grad_norm": 0.12857934832572937,
|
| 179 |
+
"learning_rate": 0.0001294871794871795,
|
| 180 |
+
"loss": 0.1047,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.47,
|
| 185 |
+
"grad_norm": 0.11967829614877701,
|
| 186 |
+
"learning_rate": 0.00012628205128205127,
|
| 187 |
+
"loss": 0.0523,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.54,
|
| 192 |
+
"grad_norm": 1.8435248136520386,
|
| 193 |
+
"learning_rate": 0.0001230769230769231,
|
| 194 |
+
"loss": 0.089,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 1.6,
|
| 199 |
+
"grad_norm": 0.07049544900655746,
|
| 200 |
+
"learning_rate": 0.00011987179487179487,
|
| 201 |
+
"loss": 0.0651,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 1.67,
|
| 206 |
+
"grad_norm": 7.795147895812988,
|
| 207 |
+
"learning_rate": 0.00011666666666666668,
|
| 208 |
+
"loss": 0.0827,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.73,
|
| 213 |
+
"grad_norm": 0.06726662069559097,
|
| 214 |
+
"learning_rate": 0.00011346153846153846,
|
| 215 |
+
"loss": 0.1727,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.79,
|
| 220 |
+
"grad_norm": 4.7732672691345215,
|
| 221 |
+
"learning_rate": 0.00011025641025641027,
|
| 222 |
+
"loss": 0.0867,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.86,
|
| 227 |
+
"grad_norm": 0.08257576823234558,
|
| 228 |
+
"learning_rate": 0.00010705128205128206,
|
| 229 |
+
"loss": 0.0349,
|
| 230 |
+
"step": 290
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.92,
|
| 234 |
+
"grad_norm": 0.15855157375335693,
|
| 235 |
+
"learning_rate": 0.00010384615384615386,
|
| 236 |
+
"loss": 0.0253,
|
| 237 |
+
"step": 300
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.92,
|
| 241 |
+
"eval_accuracy": 0.954983922829582,
|
| 242 |
+
"eval_loss": 0.15877817571163177,
|
| 243 |
+
"eval_runtime": 9.849,
|
| 244 |
+
"eval_samples_per_second": 63.153,
|
| 245 |
+
"eval_steps_per_second": 7.92,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.99,
|
| 250 |
+
"grad_norm": 11.21109676361084,
|
| 251 |
+
"learning_rate": 0.00010064102564102564,
|
| 252 |
+
"loss": 0.0988,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 2.05,
|
| 257 |
+
"grad_norm": 0.06449388712644577,
|
| 258 |
+
"learning_rate": 9.743589743589744e-05,
|
| 259 |
+
"loss": 0.0666,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.12,
|
| 264 |
+
"grad_norm": 0.988405168056488,
|
| 265 |
+
"learning_rate": 9.423076923076924e-05,
|
| 266 |
+
"loss": 0.0295,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.18,
|
| 271 |
+
"grad_norm": 0.06675518304109573,
|
| 272 |
+
"learning_rate": 9.102564102564103e-05,
|
| 273 |
+
"loss": 0.018,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.24,
|
| 278 |
+
"grad_norm": 0.08486536890268326,
|
| 279 |
+
"learning_rate": 8.782051282051283e-05,
|
| 280 |
+
"loss": 0.0714,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.31,
|
| 285 |
+
"grad_norm": 0.05260853096842766,
|
| 286 |
+
"learning_rate": 8.461538461538461e-05,
|
| 287 |
+
"loss": 0.0354,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.37,
|
| 292 |
+
"grad_norm": 0.053938668221235275,
|
| 293 |
+
"learning_rate": 8.141025641025641e-05,
|
| 294 |
+
"loss": 0.0548,
|
| 295 |
+
"step": 370
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.44,
|
| 299 |
+
"grad_norm": 0.06470278650522232,
|
| 300 |
+
"learning_rate": 7.820512820512821e-05,
|
| 301 |
+
"loss": 0.0162,
|
| 302 |
+
"step": 380
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.5,
|
| 306 |
+
"grad_norm": 0.0850602388381958,
|
| 307 |
+
"learning_rate": 7.500000000000001e-05,
|
| 308 |
+
"loss": 0.033,
|
| 309 |
+
"step": 390
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.56,
|
| 313 |
+
"grad_norm": 0.04342366382479668,
|
| 314 |
+
"learning_rate": 7.17948717948718e-05,
|
| 315 |
+
"loss": 0.0224,
|
| 316 |
+
"step": 400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 2.56,
|
| 320 |
+
"eval_accuracy": 0.9533762057877814,
|
| 321 |
+
"eval_loss": 0.16905710101127625,
|
| 322 |
+
"eval_runtime": 9.8491,
|
| 323 |
+
"eval_samples_per_second": 63.153,
|
| 324 |
+
"eval_steps_per_second": 7.92,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.63,
|
| 329 |
+
"grad_norm": 0.05912560597062111,
|
| 330 |
+
"learning_rate": 6.858974358974359e-05,
|
| 331 |
+
"loss": 0.0503,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.69,
|
| 336 |
+
"grad_norm": 0.04359501227736473,
|
| 337 |
+
"learning_rate": 6.538461538461539e-05,
|
| 338 |
+
"loss": 0.0537,
|
| 339 |
+
"step": 420
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.76,
|
| 343 |
+
"grad_norm": 0.0935799852013588,
|
| 344 |
+
"learning_rate": 6.217948717948718e-05,
|
| 345 |
+
"loss": 0.0145,
|
| 346 |
+
"step": 430
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.82,
|
| 350 |
+
"grad_norm": 0.05057013779878616,
|
| 351 |
+
"learning_rate": 5.897435897435898e-05,
|
| 352 |
+
"loss": 0.0132,
|
| 353 |
+
"step": 440
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.88,
|
| 357 |
+
"grad_norm": 0.4015754461288452,
|
| 358 |
+
"learning_rate": 5.576923076923077e-05,
|
| 359 |
+
"loss": 0.0382,
|
| 360 |
+
"step": 450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.95,
|
| 364 |
+
"grad_norm": 0.03421848267316818,
|
| 365 |
+
"learning_rate": 5.256410256410257e-05,
|
| 366 |
+
"loss": 0.0123,
|
| 367 |
+
"step": 460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 3.01,
|
| 371 |
+
"grad_norm": 0.043205052614212036,
|
| 372 |
+
"learning_rate": 4.935897435897436e-05,
|
| 373 |
+
"loss": 0.0417,
|
| 374 |
+
"step": 470
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 3.08,
|
| 378 |
+
"grad_norm": 0.03611929342150688,
|
| 379 |
+
"learning_rate": 4.615384615384616e-05,
|
| 380 |
+
"loss": 0.0125,
|
| 381 |
+
"step": 480
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 3.14,
|
| 385 |
+
"grad_norm": 0.04078483581542969,
|
| 386 |
+
"learning_rate": 4.294871794871795e-05,
|
| 387 |
+
"loss": 0.0193,
|
| 388 |
+
"step": 490
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 3.21,
|
| 392 |
+
"grad_norm": 0.043482307344675064,
|
| 393 |
+
"learning_rate": 3.974358974358974e-05,
|
| 394 |
+
"loss": 0.0321,
|
| 395 |
+
"step": 500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.21,
|
| 399 |
+
"eval_accuracy": 0.9565916398713826,
|
| 400 |
+
"eval_loss": 0.17511475086212158,
|
| 401 |
+
"eval_runtime": 9.8922,
|
| 402 |
+
"eval_samples_per_second": 62.878,
|
| 403 |
+
"eval_steps_per_second": 7.885,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.27,
|
| 408 |
+
"grad_norm": 0.03763527050614357,
|
| 409 |
+
"learning_rate": 3.653846153846154e-05,
|
| 410 |
+
"loss": 0.0118,
|
| 411 |
+
"step": 510
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 3.33,
|
| 415 |
+
"grad_norm": 0.05929262936115265,
|
| 416 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 417 |
+
"loss": 0.0636,
|
| 418 |
+
"step": 520
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 3.4,
|
| 422 |
+
"grad_norm": 0.039751436561346054,
|
| 423 |
+
"learning_rate": 3.012820512820513e-05,
|
| 424 |
+
"loss": 0.0105,
|
| 425 |
+
"step": 530
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.46,
|
| 429 |
+
"grad_norm": 0.03735564276576042,
|
| 430 |
+
"learning_rate": 2.6923076923076923e-05,
|
| 431 |
+
"loss": 0.0495,
|
| 432 |
+
"step": 540
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.53,
|
| 436 |
+
"grad_norm": 0.051983293145895004,
|
| 437 |
+
"learning_rate": 2.3717948717948718e-05,
|
| 438 |
+
"loss": 0.011,
|
| 439 |
+
"step": 550
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 3.59,
|
| 443 |
+
"grad_norm": 0.034572675824165344,
|
| 444 |
+
"learning_rate": 2.0512820512820512e-05,
|
| 445 |
+
"loss": 0.0109,
|
| 446 |
+
"step": 560
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.65,
|
| 450 |
+
"grad_norm": 0.04169879108667374,
|
| 451 |
+
"learning_rate": 1.730769230769231e-05,
|
| 452 |
+
"loss": 0.0108,
|
| 453 |
+
"step": 570
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.72,
|
| 457 |
+
"grad_norm": 0.032876156270504,
|
| 458 |
+
"learning_rate": 1.4102564102564104e-05,
|
| 459 |
+
"loss": 0.0104,
|
| 460 |
+
"step": 580
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 3.78,
|
| 464 |
+
"grad_norm": 0.03522384166717529,
|
| 465 |
+
"learning_rate": 1.0897435897435898e-05,
|
| 466 |
+
"loss": 0.0109,
|
| 467 |
+
"step": 590
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 3.85,
|
| 471 |
+
"grad_norm": 0.11409874260425568,
|
| 472 |
+
"learning_rate": 7.692307692307694e-06,
|
| 473 |
+
"loss": 0.0112,
|
| 474 |
+
"step": 600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.85,
|
| 478 |
+
"eval_accuracy": 0.954983922829582,
|
| 479 |
+
"eval_loss": 0.18050101399421692,
|
| 480 |
+
"eval_runtime": 9.8888,
|
| 481 |
+
"eval_samples_per_second": 62.899,
|
| 482 |
+
"eval_steps_per_second": 7.888,
|
| 483 |
+
"step": 600
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 3.91,
|
| 487 |
+
"grad_norm": 0.04359051212668419,
|
| 488 |
+
"learning_rate": 4.487179487179488e-06,
|
| 489 |
+
"loss": 0.0109,
|
| 490 |
+
"step": 610
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 3.97,
|
| 494 |
+
"grad_norm": 0.03071708232164383,
|
| 495 |
+
"learning_rate": 1.282051282051282e-06,
|
| 496 |
+
"loss": 0.0429,
|
| 497 |
+
"step": 620
|
| 498 |
+
},
|
| 499 |
{
|
| 500 |
"epoch": 4.0,
|
| 501 |
+
"step": 624,
|
| 502 |
+
"total_flos": 7.703325099767808e+17,
|
| 503 |
+
"train_loss": 0.15572628828410345,
|
| 504 |
+
"train_runtime": 552.98,
|
| 505 |
+
"train_samples_per_second": 17.975,
|
| 506 |
+
"train_steps_per_second": 1.128
|
| 507 |
}
|
| 508 |
],
|
| 509 |
"logging_steps": 10,
|
| 510 |
+
"max_steps": 624,
|
| 511 |
"num_input_tokens_seen": 0,
|
| 512 |
"num_train_epochs": 4,
|
| 513 |
"save_steps": 100,
|
| 514 |
+
"total_flos": 7.703325099767808e+17,
|
| 515 |
"train_batch_size": 16,
|
| 516 |
"trial_name": null,
|
| 517 |
"trial_params": null
|