End of training
Browse files- README.md +3 -1
- all_results.json +9 -9
- eval_results.json +5 -5
- train_results.json +4 -4
- trainer_state.json +495 -68
README.md
CHANGED
|
@@ -5,6 +5,8 @@ license: apache-2.0
|
|
| 5 |
metrics:
|
| 6 |
- accuracy
|
| 7 |
tags:
|
|
|
|
|
|
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: vit-base-beans
|
|
@@ -16,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# vit-base-beans
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.8530
|
| 22 |
- Accuracy: 0.8045
|
|
|
|
| 5 |
metrics:
|
| 6 |
- accuracy
|
| 7 |
tags:
|
| 8 |
+
- image-classification
|
| 9 |
+
- vision
|
| 10 |
- generated_from_trainer
|
| 11 |
model-index:
|
| 12 |
- name: vit-base-beans
|
|
|
|
| 18 |
|
| 19 |
# vit-base-beans
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the beans dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
- Loss: 0.8530
|
| 24 |
- Accuracy: 0.8045
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 11.
|
| 6 |
-
"eval_samples_per_second": 11.
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
"total_flos": 4.0340206085566464e+17,
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.8045112781954887,
|
| 4 |
+
"eval_loss": 0.8529870510101318,
|
| 5 |
+
"eval_runtime": 11.4307,
|
| 6 |
+
"eval_samples_per_second": 11.635,
|
| 7 |
+
"eval_steps_per_second": 1.487,
|
| 8 |
"total_flos": 4.0340206085566464e+17,
|
| 9 |
+
"train_loss": 0.9484451132554275,
|
| 10 |
+
"train_runtime": 507.077,
|
| 11 |
+
"train_samples_per_second": 10.196,
|
| 12 |
+
"train_steps_per_second": 1.282
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 11.
|
| 6 |
-
"eval_samples_per_second": 11.
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.8045112781954887,
|
| 4 |
+
"eval_loss": 0.8529870510101318,
|
| 5 |
+
"eval_runtime": 11.4307,
|
| 6 |
+
"eval_samples_per_second": 11.635,
|
| 7 |
+
"eval_steps_per_second": 1.487
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
"total_flos": 4.0340206085566464e+17,
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
"total_flos": 4.0340206085566464e+17,
|
| 4 |
+
"train_loss": 0.9484451132554275,
|
| 5 |
+
"train_runtime": 507.077,
|
| 6 |
+
"train_samples_per_second": 10.196,
|
| 7 |
+
"train_steps_per_second": 1.282
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,98 +1,525 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./beans_lora_outputs/checkpoint-
|
| 4 |
"epoch": 5.0,
|
| 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":
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"eval_samples_per_second": 8.156,
|
| 17 |
-
"eval_steps_per_second": 0.123,
|
| 18 |
-
"step": 9
|
| 19 |
-
},
|
| 20 |
-
{
|
| 21 |
-
"epoch": 1.1111111111111112,
|
| 22 |
-
"grad_norm": 0.9401128888130188,
|
| 23 |
-
"learning_rate": 1.555555555555556e-05,
|
| 24 |
-
"loss": 1.0798,
|
| 25 |
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
-
"epoch":
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"eval_samples_per_second": 12.009,
|
| 33 |
-
"eval_steps_per_second": 0.181,
|
| 34 |
-
"step": 18
|
| 35 |
-
},
|
| 36 |
-
{
|
| 37 |
-
"epoch": 2.2222222222222223,
|
| 38 |
-
"grad_norm": 0.8580392003059387,
|
| 39 |
-
"learning_rate": 1.1111111111111113e-05,
|
| 40 |
-
"loss": 1.0728,
|
| 41 |
"step": 20
|
| 42 |
},
|
| 43 |
{
|
| 44 |
-
"epoch":
|
| 45 |
-
"
|
| 46 |
-
"
|
| 47 |
-
"
|
| 48 |
-
"eval_samples_per_second": 7.679,
|
| 49 |
-
"eval_steps_per_second": 0.115,
|
| 50 |
-
"step": 27
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"epoch": 3.3333333333333335,
|
| 54 |
-
"grad_norm": 0.9443564414978027,
|
| 55 |
-
"learning_rate": 6.666666666666667e-06,
|
| 56 |
-
"loss": 1.0719,
|
| 57 |
"step": 30
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch":
|
| 61 |
-
"
|
| 62 |
-
"
|
| 63 |
-
"
|
| 64 |
-
"eval_samples_per_second": 21.832,
|
| 65 |
-
"eval_steps_per_second": 0.328,
|
| 66 |
-
"step": 36
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 4.444444444444445,
|
| 70 |
-
"grad_norm": 0.8862206339836121,
|
| 71 |
-
"learning_rate": 2.222222222222222e-06,
|
| 72 |
-
"loss": 1.062,
|
| 73 |
"step": 40
|
| 74 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
{
|
| 76 |
"epoch": 5.0,
|
| 77 |
-
"eval_accuracy": 0.
|
| 78 |
-
"eval_loss":
|
| 79 |
-
"eval_runtime":
|
| 80 |
-
"eval_samples_per_second":
|
| 81 |
-
"eval_steps_per_second":
|
| 82 |
-
"step":
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 5.0,
|
| 86 |
-
"step":
|
| 87 |
"total_flos": 4.0340206085566464e+17,
|
| 88 |
-
"train_loss":
|
| 89 |
-
"train_runtime":
|
| 90 |
-
"train_samples_per_second":
|
| 91 |
-
"train_steps_per_second":
|
| 92 |
}
|
| 93 |
],
|
| 94 |
"logging_steps": 10,
|
| 95 |
-
"max_steps":
|
| 96 |
"num_input_tokens_seen": 0,
|
| 97 |
"num_train_epochs": 5,
|
| 98 |
"save_steps": 500,
|
|
@@ -109,7 +536,7 @@
|
|
| 109 |
}
|
| 110 |
},
|
| 111 |
"total_flos": 4.0340206085566464e+17,
|
| 112 |
-
"train_batch_size":
|
| 113 |
"trial_name": null,
|
| 114 |
"trial_params": null
|
| 115 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8529870510101318,
|
| 3 |
+
"best_model_checkpoint": "./beans_lora_outputs/checkpoint-650",
|
| 4 |
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 650,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.07692307692307693,
|
| 13 |
+
"grad_norm": 1.495912790298462,
|
| 14 |
+
"learning_rate": 1.9692307692307696e-05,
|
| 15 |
+
"loss": 1.0741,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.15384615384615385,
|
| 20 |
+
"grad_norm": 1.1474604606628418,
|
| 21 |
+
"learning_rate": 1.9384615384615386e-05,
|
| 22 |
+
"loss": 1.0879,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.23076923076923078,
|
| 27 |
+
"grad_norm": 1.6189851760864258,
|
| 28 |
+
"learning_rate": 1.907692307692308e-05,
|
| 29 |
+
"loss": 1.0802,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.3076923076923077,
|
| 34 |
+
"grad_norm": 2.2019317150115967,
|
| 35 |
+
"learning_rate": 1.876923076923077e-05,
|
| 36 |
+
"loss": 1.0493,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.38461538461538464,
|
| 41 |
+
"grad_norm": 1.3802038431167603,
|
| 42 |
+
"learning_rate": 1.8461538461538465e-05,
|
| 43 |
+
"loss": 1.0676,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.46153846153846156,
|
| 48 |
+
"grad_norm": 1.385358214378357,
|
| 49 |
+
"learning_rate": 1.8153846153846155e-05,
|
| 50 |
+
"loss": 1.0609,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.5384615384615384,
|
| 55 |
+
"grad_norm": 1.984493374824524,
|
| 56 |
+
"learning_rate": 1.784615384615385e-05,
|
| 57 |
+
"loss": 1.0451,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.6153846153846154,
|
| 62 |
+
"grad_norm": 1.7820314168930054,
|
| 63 |
+
"learning_rate": 1.753846153846154e-05,
|
| 64 |
+
"loss": 1.044,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.6923076923076923,
|
| 69 |
+
"grad_norm": 1.7298434972763062,
|
| 70 |
+
"learning_rate": 1.7230769230769234e-05,
|
| 71 |
+
"loss": 1.0523,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.7692307692307693,
|
| 76 |
+
"grad_norm": 2.324188470840454,
|
| 77 |
+
"learning_rate": 1.6923076923076924e-05,
|
| 78 |
+
"loss": 1.0571,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.8461538461538461,
|
| 83 |
+
"grad_norm": 1.5387742519378662,
|
| 84 |
+
"learning_rate": 1.6615384615384618e-05,
|
| 85 |
+
"loss": 1.0593,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9230769230769231,
|
| 90 |
+
"grad_norm": 1.8900723457336426,
|
| 91 |
+
"learning_rate": 1.630769230769231e-05,
|
| 92 |
+
"loss": 1.0329,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.0,
|
| 97 |
+
"grad_norm": 2.8200511932373047,
|
| 98 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 99 |
+
"loss": 1.0326,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.0,
|
| 104 |
+
"eval_accuracy": 0.6090225563909775,
|
| 105 |
+
"eval_loss": 1.0319362878799438,
|
| 106 |
+
"eval_runtime": 6.8087,
|
| 107 |
+
"eval_samples_per_second": 19.534,
|
| 108 |
+
"eval_steps_per_second": 2.497,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.0769230769230769,
|
| 113 |
+
"grad_norm": 1.4958809614181519,
|
| 114 |
+
"learning_rate": 1.5692307692307693e-05,
|
| 115 |
+
"loss": 1.0263,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.1538461538461537,
|
| 120 |
+
"grad_norm": 1.5726135969161987,
|
| 121 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 122 |
+
"loss": 1.0135,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.2307692307692308,
|
| 127 |
+
"grad_norm": 1.3866033554077148,
|
| 128 |
+
"learning_rate": 1.5076923076923078e-05,
|
| 129 |
+
"loss": 1.0212,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.3076923076923077,
|
| 134 |
+
"grad_norm": 1.6613730192184448,
|
| 135 |
+
"learning_rate": 1.4769230769230772e-05,
|
| 136 |
+
"loss": 1.0116,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.3846153846153846,
|
| 141 |
+
"grad_norm": 1.5437824726104736,
|
| 142 |
+
"learning_rate": 1.4461538461538462e-05,
|
| 143 |
+
"loss": 1.0009,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.4615384615384617,
|
| 148 |
+
"grad_norm": 1.5443662405014038,
|
| 149 |
+
"learning_rate": 1.4153846153846156e-05,
|
| 150 |
+
"loss": 1.0036,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.5384615384615383,
|
| 155 |
+
"grad_norm": 1.3017466068267822,
|
| 156 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 157 |
+
"loss": 1.0066,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.6153846153846154,
|
| 162 |
+
"grad_norm": 1.4007813930511475,
|
| 163 |
+
"learning_rate": 1.353846153846154e-05,
|
| 164 |
+
"loss": 0.9954,
|
| 165 |
+
"step": 210
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.6923076923076923,
|
| 169 |
+
"grad_norm": 2.7631592750549316,
|
| 170 |
+
"learning_rate": 1.3230769230769231e-05,
|
| 171 |
+
"loss": 0.9775,
|
| 172 |
+
"step": 220
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.7692307692307692,
|
| 176 |
+
"grad_norm": 1.6825937032699585,
|
| 177 |
+
"learning_rate": 1.2923076923076925e-05,
|
| 178 |
+
"loss": 0.9898,
|
| 179 |
+
"step": 230
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.8461538461538463,
|
| 183 |
+
"grad_norm": 1.4360311031341553,
|
| 184 |
+
"learning_rate": 1.2615384615384616e-05,
|
| 185 |
+
"loss": 1.0017,
|
| 186 |
+
"step": 240
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.9230769230769231,
|
| 190 |
+
"grad_norm": 1.6261993646621704,
|
| 191 |
+
"learning_rate": 1.230769230769231e-05,
|
| 192 |
+
"loss": 0.9741,
|
| 193 |
+
"step": 250
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.0,
|
| 197 |
+
"grad_norm": 2.090559720993042,
|
| 198 |
+
"learning_rate": 1.2e-05,
|
| 199 |
+
"loss": 0.9721,
|
| 200 |
+
"step": 260
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.0,
|
| 204 |
+
"eval_accuracy": 0.7293233082706767,
|
| 205 |
+
"eval_loss": 0.9698564410209656,
|
| 206 |
+
"eval_runtime": 11.0803,
|
| 207 |
+
"eval_samples_per_second": 12.003,
|
| 208 |
+
"eval_steps_per_second": 1.534,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.076923076923077,
|
| 213 |
+
"grad_norm": 1.219846487045288,
|
| 214 |
+
"learning_rate": 1.1692307692307694e-05,
|
| 215 |
+
"loss": 0.9583,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.1538461538461537,
|
| 220 |
+
"grad_norm": 1.85811448097229,
|
| 221 |
+
"learning_rate": 1.1384615384615385e-05,
|
| 222 |
+
"loss": 0.9665,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.230769230769231,
|
| 227 |
+
"grad_norm": 1.8803926706314087,
|
| 228 |
+
"learning_rate": 1.1076923076923079e-05,
|
| 229 |
+
"loss": 0.9529,
|
| 230 |
+
"step": 290
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.3076923076923075,
|
| 234 |
+
"grad_norm": 1.4712238311767578,
|
| 235 |
+
"learning_rate": 1.076923076923077e-05,
|
| 236 |
+
"loss": 0.9423,
|
| 237 |
+
"step": 300
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.3846153846153846,
|
| 241 |
+
"grad_norm": 1.5016744136810303,
|
| 242 |
+
"learning_rate": 1.0461538461538463e-05,
|
| 243 |
+
"loss": 0.9554,
|
| 244 |
+
"step": 310
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.4615384615384617,
|
| 248 |
+
"grad_norm": 1.317639946937561,
|
| 249 |
+
"learning_rate": 1.0153846153846154e-05,
|
| 250 |
+
"loss": 0.9414,
|
| 251 |
+
"step": 320
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.5384615384615383,
|
| 255 |
+
"grad_norm": 1.2138745784759521,
|
| 256 |
+
"learning_rate": 9.846153846153848e-06,
|
| 257 |
+
"loss": 0.9417,
|
| 258 |
+
"step": 330
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.6153846153846154,
|
| 262 |
+
"grad_norm": 1.5316927433013916,
|
| 263 |
+
"learning_rate": 9.53846153846154e-06,
|
| 264 |
+
"loss": 0.9303,
|
| 265 |
+
"step": 340
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.6923076923076925,
|
| 269 |
+
"grad_norm": 1.4810699224472046,
|
| 270 |
+
"learning_rate": 9.230769230769232e-06,
|
| 271 |
+
"loss": 0.9361,
|
| 272 |
+
"step": 350
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.769230769230769,
|
| 276 |
+
"grad_norm": 1.3046553134918213,
|
| 277 |
+
"learning_rate": 8.923076923076925e-06,
|
| 278 |
+
"loss": 0.9262,
|
| 279 |
+
"step": 360
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.8461538461538463,
|
| 283 |
+
"grad_norm": 1.4417724609375,
|
| 284 |
+
"learning_rate": 8.615384615384617e-06,
|
| 285 |
+
"loss": 0.9237,
|
| 286 |
+
"step": 370
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 2.9230769230769234,
|
| 290 |
+
"grad_norm": 1.4357845783233643,
|
| 291 |
+
"learning_rate": 8.307692307692309e-06,
|
| 292 |
+
"loss": 0.9172,
|
| 293 |
+
"step": 380
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 3.0,
|
| 297 |
+
"grad_norm": 2.6237688064575195,
|
| 298 |
+
"learning_rate": 8.000000000000001e-06,
|
| 299 |
+
"loss": 0.8948,
|
| 300 |
+
"step": 390
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 3.0,
|
| 304 |
+
"eval_accuracy": 0.7819548872180451,
|
| 305 |
+
"eval_loss": 0.905978798866272,
|
| 306 |
+
"eval_runtime": 11.2154,
|
| 307 |
+
"eval_samples_per_second": 11.859,
|
| 308 |
+
"eval_steps_per_second": 1.516,
|
| 309 |
+
"step": 390
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.076923076923077,
|
| 313 |
+
"grad_norm": 1.473265290260315,
|
| 314 |
+
"learning_rate": 7.692307692307694e-06,
|
| 315 |
+
"loss": 0.9283,
|
| 316 |
+
"step": 400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.1538461538461537,
|
| 320 |
+
"grad_norm": 1.2494803667068481,
|
| 321 |
+
"learning_rate": 7.384615384615386e-06,
|
| 322 |
+
"loss": 0.9173,
|
| 323 |
+
"step": 410
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.230769230769231,
|
| 327 |
+
"grad_norm": 1.5195587873458862,
|
| 328 |
+
"learning_rate": 7.076923076923078e-06,
|
| 329 |
+
"loss": 0.916,
|
| 330 |
+
"step": 420
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.3076923076923075,
|
| 334 |
+
"grad_norm": 2.023468255996704,
|
| 335 |
+
"learning_rate": 6.76923076923077e-06,
|
| 336 |
+
"loss": 0.8841,
|
| 337 |
+
"step": 430
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.3846153846153846,
|
| 341 |
+
"grad_norm": 1.7898038625717163,
|
| 342 |
+
"learning_rate": 6.461538461538463e-06,
|
| 343 |
+
"loss": 0.9006,
|
| 344 |
+
"step": 440
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.4615384615384617,
|
| 348 |
+
"grad_norm": 1.475136160850525,
|
| 349 |
+
"learning_rate": 6.153846153846155e-06,
|
| 350 |
+
"loss": 0.8768,
|
| 351 |
+
"step": 450
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.5384615384615383,
|
| 355 |
+
"grad_norm": 1.7361855506896973,
|
| 356 |
+
"learning_rate": 5.846153846153847e-06,
|
| 357 |
+
"loss": 0.8955,
|
| 358 |
+
"step": 460
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.6153846153846154,
|
| 362 |
+
"grad_norm": 1.3542256355285645,
|
| 363 |
+
"learning_rate": 5.538461538461539e-06,
|
| 364 |
+
"loss": 0.8628,
|
| 365 |
+
"step": 470
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.6923076923076925,
|
| 369 |
+
"grad_norm": 2.064852714538574,
|
| 370 |
+
"learning_rate": 5.230769230769232e-06,
|
| 371 |
+
"loss": 0.8895,
|
| 372 |
+
"step": 480
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.769230769230769,
|
| 376 |
+
"grad_norm": 1.4197514057159424,
|
| 377 |
+
"learning_rate": 4.923076923076924e-06,
|
| 378 |
+
"loss": 0.8612,
|
| 379 |
+
"step": 490
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.8461538461538463,
|
| 383 |
+
"grad_norm": 1.2602348327636719,
|
| 384 |
+
"learning_rate": 4.615384615384616e-06,
|
| 385 |
+
"loss": 0.8698,
|
| 386 |
+
"step": 500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.9230769230769234,
|
| 390 |
+
"grad_norm": 1.2403748035430908,
|
| 391 |
+
"learning_rate": 4.307692307692308e-06,
|
| 392 |
+
"loss": 0.8712,
|
| 393 |
+
"step": 510
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 4.0,
|
| 397 |
+
"grad_norm": 3.1204898357391357,
|
| 398 |
+
"learning_rate": 4.000000000000001e-06,
|
| 399 |
+
"loss": 0.8524,
|
| 400 |
+
"step": 520
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 4.0,
|
| 404 |
+
"eval_accuracy": 0.8045112781954887,
|
| 405 |
+
"eval_loss": 0.8660793304443359,
|
| 406 |
+
"eval_runtime": 1.2917,
|
| 407 |
+
"eval_samples_per_second": 102.962,
|
| 408 |
+
"eval_steps_per_second": 13.161,
|
| 409 |
+
"step": 520
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.076923076923077,
|
| 413 |
+
"grad_norm": 2.3789713382720947,
|
| 414 |
+
"learning_rate": 3.692307692307693e-06,
|
| 415 |
+
"loss": 0.8624,
|
| 416 |
+
"step": 530
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.153846153846154,
|
| 420 |
+
"grad_norm": 1.572312831878662,
|
| 421 |
+
"learning_rate": 3.384615384615385e-06,
|
| 422 |
+
"loss": 0.8991,
|
| 423 |
+
"step": 540
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.230769230769231,
|
| 427 |
+
"grad_norm": 2.145324945449829,
|
| 428 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 429 |
+
"loss": 0.8963,
|
| 430 |
+
"step": 550
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.3076923076923075,
|
| 434 |
+
"grad_norm": 1.2909295558929443,
|
| 435 |
+
"learning_rate": 2.7692307692307697e-06,
|
| 436 |
+
"loss": 0.8629,
|
| 437 |
+
"step": 560
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 4.384615384615385,
|
| 441 |
+
"grad_norm": 1.4236739873886108,
|
| 442 |
+
"learning_rate": 2.461538461538462e-06,
|
| 443 |
+
"loss": 0.8597,
|
| 444 |
+
"step": 570
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.461538461538462,
|
| 448 |
+
"grad_norm": 1.638335943222046,
|
| 449 |
+
"learning_rate": 2.153846153846154e-06,
|
| 450 |
+
"loss": 0.833,
|
| 451 |
+
"step": 580
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 4.538461538461538,
|
| 455 |
+
"grad_norm": 1.5316523313522339,
|
| 456 |
+
"learning_rate": 1.8461538461538465e-06,
|
| 457 |
+
"loss": 0.8521,
|
| 458 |
+
"step": 590
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.615384615384615,
|
| 462 |
+
"grad_norm": 1.2549222707748413,
|
| 463 |
+
"learning_rate": 1.5384615384615387e-06,
|
| 464 |
+
"loss": 0.8741,
|
| 465 |
+
"step": 600
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 4.6923076923076925,
|
| 469 |
+
"grad_norm": 2.406038761138916,
|
| 470 |
+
"learning_rate": 1.230769230769231e-06,
|
| 471 |
+
"loss": 0.8801,
|
| 472 |
+
"step": 610
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 4.769230769230769,
|
| 476 |
+
"grad_norm": 1.3158031702041626,
|
| 477 |
+
"learning_rate": 9.230769230769232e-07,
|
| 478 |
+
"loss": 0.829,
|
| 479 |
+
"step": 620
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 4.846153846153846,
|
| 483 |
+
"grad_norm": 1.4627388715744019,
|
| 484 |
+
"learning_rate": 6.153846153846155e-07,
|
| 485 |
+
"loss": 0.8492,
|
| 486 |
+
"step": 630
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 4.923076923076923,
|
| 490 |
+
"grad_norm": 2.0194759368896484,
|
| 491 |
+
"learning_rate": 3.0769230769230774e-07,
|
| 492 |
+
"loss": 0.8351,
|
| 493 |
+
"step": 640
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 5.0,
|
| 497 |
+
"grad_norm": 3.150322914123535,
|
| 498 |
+
"learning_rate": 0.0,
|
| 499 |
+
"loss": 0.866,
|
| 500 |
+
"step": 650
|
| 501 |
+
},
|
| 502 |
{
|
| 503 |
"epoch": 5.0,
|
| 504 |
+
"eval_accuracy": 0.8045112781954887,
|
| 505 |
+
"eval_loss": 0.8529870510101318,
|
| 506 |
+
"eval_runtime": 5.7804,
|
| 507 |
+
"eval_samples_per_second": 23.009,
|
| 508 |
+
"eval_steps_per_second": 2.941,
|
| 509 |
+
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 5.0,
|
| 513 |
+
"step": 650,
|
| 514 |
"total_flos": 4.0340206085566464e+17,
|
| 515 |
+
"train_loss": 0.9484451132554275,
|
| 516 |
+
"train_runtime": 507.077,
|
| 517 |
+
"train_samples_per_second": 10.196,
|
| 518 |
+
"train_steps_per_second": 1.282
|
| 519 |
}
|
| 520 |
],
|
| 521 |
"logging_steps": 10,
|
| 522 |
+
"max_steps": 650,
|
| 523 |
"num_input_tokens_seen": 0,
|
| 524 |
"num_train_epochs": 5,
|
| 525 |
"save_steps": 500,
|
|
|
|
| 536 |
}
|
| 537 |
},
|
| 538 |
"total_flos": 4.0340206085566464e+17,
|
| 539 |
+
"train_batch_size": 8,
|
| 540 |
"trial_name": null,
|
| 541 |
"trial_params": null
|
| 542 |
}
|