End of training
Browse files- README.md +5 -3
- all_results.json +13 -0
- config.json +1 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +543 -0
README.md
CHANGED
|
@@ -3,6 +3,8 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: google/vit-base-patch16-224-in21k
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
@@ -16,10 +18,10 @@ 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 |
-
-
|
| 22 |
-
-
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: google/vit-base-patch16-224-in21k
|
| 5 |
tags:
|
| 6 |
+
- image-classification
|
| 7 |
+
- vision
|
| 8 |
- generated_from_trainer
|
| 9 |
metrics:
|
| 10 |
- accuracy
|
|
|
|
| 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 AI-Lab-Makerere/beans dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.0655
|
| 24 |
+
- Accuracy: 0.9925
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9924812030075187,
|
| 4 |
+
"eval_loss": 0.06546766310930252,
|
| 5 |
+
"eval_runtime": 0.8127,
|
| 6 |
+
"eval_samples_per_second": 163.65,
|
| 7 |
+
"eval_steps_per_second": 20.918,
|
| 8 |
+
"total_flos": 4.006371770595533e+17,
|
| 9 |
+
"train_loss": 0.0,
|
| 10 |
+
"train_runtime": 0.1521,
|
| 11 |
+
"train_samples_per_second": 33994.24,
|
| 12 |
+
"train_steps_per_second": 4273.937
|
| 13 |
+
}
|
config.json
CHANGED
|
@@ -29,6 +29,7 @@
|
|
| 29 |
"patch_size": 16,
|
| 30 |
"pooler_act": "tanh",
|
| 31 |
"pooler_output_size": 768,
|
|
|
|
| 32 |
"qkv_bias": true,
|
| 33 |
"torch_dtype": "float32",
|
| 34 |
"transformers_version": "4.54.1"
|
|
|
|
| 29 |
"patch_size": 16,
|
| 30 |
"pooler_act": "tanh",
|
| 31 |
"pooler_output_size": 768,
|
| 32 |
+
"problem_type": "single_label_classification",
|
| 33 |
"qkv_bias": true,
|
| 34 |
"torch_dtype": "float32",
|
| 35 |
"transformers_version": "4.54.1"
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9924812030075187,
|
| 4 |
+
"eval_loss": 0.06546766310930252,
|
| 5 |
+
"eval_runtime": 0.8127,
|
| 6 |
+
"eval_samples_per_second": 163.65,
|
| 7 |
+
"eval_steps_per_second": 20.918
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 4.006371770595533e+17,
|
| 4 |
+
"train_loss": 0.0,
|
| 5 |
+
"train_runtime": 0.1521,
|
| 6 |
+
"train_samples_per_second": 33994.24,
|
| 7 |
+
"train_steps_per_second": 4273.937
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,543 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 520,
|
| 3 |
+
"best_metric": 0.06546766310930252,
|
| 4 |
+
"best_model_checkpoint": "./beans_outputs/checkpoint-520",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 650,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.07692307692307693,
|
| 14 |
+
"grad_norm": 2.198843479156494,
|
| 15 |
+
"learning_rate": 1.9723076923076924e-05,
|
| 16 |
+
"loss": 1.0245,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.15384615384615385,
|
| 21 |
+
"grad_norm": 1.917884111404419,
|
| 22 |
+
"learning_rate": 1.9415384615384615e-05,
|
| 23 |
+
"loss": 0.9454,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.23076923076923078,
|
| 28 |
+
"grad_norm": 2.078744411468506,
|
| 29 |
+
"learning_rate": 1.910769230769231e-05,
|
| 30 |
+
"loss": 0.8398,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.3076923076923077,
|
| 35 |
+
"grad_norm": 2.6160151958465576,
|
| 36 |
+
"learning_rate": 1.88e-05,
|
| 37 |
+
"loss": 0.6955,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.38461538461538464,
|
| 42 |
+
"grad_norm": 4.149175643920898,
|
| 43 |
+
"learning_rate": 1.8492307692307694e-05,
|
| 44 |
+
"loss": 0.661,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.46153846153846156,
|
| 49 |
+
"grad_norm": 2.1062347888946533,
|
| 50 |
+
"learning_rate": 1.8184615384615384e-05,
|
| 51 |
+
"loss": 0.5797,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.5384615384615384,
|
| 56 |
+
"grad_norm": 1.8043235540390015,
|
| 57 |
+
"learning_rate": 1.7876923076923078e-05,
|
| 58 |
+
"loss": 0.5192,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.6153846153846154,
|
| 63 |
+
"grad_norm": 1.217357873916626,
|
| 64 |
+
"learning_rate": 1.7569230769230772e-05,
|
| 65 |
+
"loss": 0.3928,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6923076923076923,
|
| 70 |
+
"grad_norm": 2.427462577819824,
|
| 71 |
+
"learning_rate": 1.7261538461538463e-05,
|
| 72 |
+
"loss": 0.3534,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.7692307692307693,
|
| 77 |
+
"grad_norm": 3.0030345916748047,
|
| 78 |
+
"learning_rate": 1.6953846153846156e-05,
|
| 79 |
+
"loss": 0.3683,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.8461538461538461,
|
| 84 |
+
"grad_norm": 2.384122371673584,
|
| 85 |
+
"learning_rate": 1.6646153846153847e-05,
|
| 86 |
+
"loss": 0.3326,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.9230769230769231,
|
| 91 |
+
"grad_norm": 2.4714367389678955,
|
| 92 |
+
"learning_rate": 1.633846153846154e-05,
|
| 93 |
+
"loss": 0.2446,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.0,
|
| 98 |
+
"grad_norm": 7.0579328536987305,
|
| 99 |
+
"learning_rate": 1.603076923076923e-05,
|
| 100 |
+
"loss": 0.281,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.0,
|
| 105 |
+
"eval_accuracy": 0.9624060150375939,
|
| 106 |
+
"eval_loss": 0.21590378880500793,
|
| 107 |
+
"eval_runtime": 0.6008,
|
| 108 |
+
"eval_samples_per_second": 221.361,
|
| 109 |
+
"eval_steps_per_second": 28.294,
|
| 110 |
+
"step": 130
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.0769230769230769,
|
| 114 |
+
"grad_norm": 1.5222105979919434,
|
| 115 |
+
"learning_rate": 1.5723076923076926e-05,
|
| 116 |
+
"loss": 0.1984,
|
| 117 |
+
"step": 140
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.1538461538461537,
|
| 121 |
+
"grad_norm": 3.3074021339416504,
|
| 122 |
+
"learning_rate": 1.5415384615384616e-05,
|
| 123 |
+
"loss": 0.2224,
|
| 124 |
+
"step": 150
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.2307692307692308,
|
| 128 |
+
"grad_norm": 1.5715149641036987,
|
| 129 |
+
"learning_rate": 1.510769230769231e-05,
|
| 130 |
+
"loss": 0.2032,
|
| 131 |
+
"step": 160
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.3076923076923077,
|
| 135 |
+
"grad_norm": 0.5117770433425903,
|
| 136 |
+
"learning_rate": 1.48e-05,
|
| 137 |
+
"loss": 0.2197,
|
| 138 |
+
"step": 170
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.3846153846153846,
|
| 142 |
+
"grad_norm": 0.6276758909225464,
|
| 143 |
+
"learning_rate": 1.4492307692307695e-05,
|
| 144 |
+
"loss": 0.209,
|
| 145 |
+
"step": 180
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 1.4615384615384617,
|
| 149 |
+
"grad_norm": 2.73740553855896,
|
| 150 |
+
"learning_rate": 1.4184615384615385e-05,
|
| 151 |
+
"loss": 0.1603,
|
| 152 |
+
"step": 190
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.5384615384615383,
|
| 156 |
+
"grad_norm": 0.4705379605293274,
|
| 157 |
+
"learning_rate": 1.3876923076923079e-05,
|
| 158 |
+
"loss": 0.2545,
|
| 159 |
+
"step": 200
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.6153846153846154,
|
| 163 |
+
"grad_norm": 0.6391307711601257,
|
| 164 |
+
"learning_rate": 1.356923076923077e-05,
|
| 165 |
+
"loss": 0.1793,
|
| 166 |
+
"step": 210
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.6923076923076923,
|
| 170 |
+
"grad_norm": 3.0284461975097656,
|
| 171 |
+
"learning_rate": 1.3261538461538464e-05,
|
| 172 |
+
"loss": 0.2168,
|
| 173 |
+
"step": 220
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.7692307692307692,
|
| 177 |
+
"grad_norm": 2.021752119064331,
|
| 178 |
+
"learning_rate": 1.2953846153846154e-05,
|
| 179 |
+
"loss": 0.1622,
|
| 180 |
+
"step": 230
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 1.8461538461538463,
|
| 184 |
+
"grad_norm": 5.118215560913086,
|
| 185 |
+
"learning_rate": 1.2646153846153848e-05,
|
| 186 |
+
"loss": 0.2609,
|
| 187 |
+
"step": 240
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 1.9230769230769231,
|
| 191 |
+
"grad_norm": 0.380063533782959,
|
| 192 |
+
"learning_rate": 1.2338461538461539e-05,
|
| 193 |
+
"loss": 0.137,
|
| 194 |
+
"step": 250
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.0,
|
| 198 |
+
"grad_norm": 0.34861981868743896,
|
| 199 |
+
"learning_rate": 1.2030769230769233e-05,
|
| 200 |
+
"loss": 0.1296,
|
| 201 |
+
"step": 260
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 2.0,
|
| 205 |
+
"eval_accuracy": 0.9699248120300752,
|
| 206 |
+
"eval_loss": 0.13036151230335236,
|
| 207 |
+
"eval_runtime": 0.607,
|
| 208 |
+
"eval_samples_per_second": 219.099,
|
| 209 |
+
"eval_steps_per_second": 28.005,
|
| 210 |
+
"step": 260
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 2.076923076923077,
|
| 214 |
+
"grad_norm": 0.48520398139953613,
|
| 215 |
+
"learning_rate": 1.1723076923076923e-05,
|
| 216 |
+
"loss": 0.0894,
|
| 217 |
+
"step": 270
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.1538461538461537,
|
| 221 |
+
"grad_norm": 0.3357318639755249,
|
| 222 |
+
"learning_rate": 1.1415384615384617e-05,
|
| 223 |
+
"loss": 0.1421,
|
| 224 |
+
"step": 280
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.230769230769231,
|
| 228 |
+
"grad_norm": 7.024319648742676,
|
| 229 |
+
"learning_rate": 1.1107692307692308e-05,
|
| 230 |
+
"loss": 0.1226,
|
| 231 |
+
"step": 290
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 2.3076923076923075,
|
| 235 |
+
"grad_norm": 8.50007152557373,
|
| 236 |
+
"learning_rate": 1.0800000000000002e-05,
|
| 237 |
+
"loss": 0.1355,
|
| 238 |
+
"step": 300
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 2.3846153846153846,
|
| 242 |
+
"grad_norm": 4.990291118621826,
|
| 243 |
+
"learning_rate": 1.0492307692307692e-05,
|
| 244 |
+
"loss": 0.1761,
|
| 245 |
+
"step": 310
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 2.4615384615384617,
|
| 249 |
+
"grad_norm": 0.3203364908695221,
|
| 250 |
+
"learning_rate": 1.0184615384615386e-05,
|
| 251 |
+
"loss": 0.1472,
|
| 252 |
+
"step": 320
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 2.5384615384615383,
|
| 256 |
+
"grad_norm": 6.037258148193359,
|
| 257 |
+
"learning_rate": 9.876923076923077e-06,
|
| 258 |
+
"loss": 0.1877,
|
| 259 |
+
"step": 330
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.6153846153846154,
|
| 263 |
+
"grad_norm": 1.1843386888504028,
|
| 264 |
+
"learning_rate": 9.569230769230769e-06,
|
| 265 |
+
"loss": 0.0712,
|
| 266 |
+
"step": 340
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.6923076923076925,
|
| 270 |
+
"grad_norm": 0.35253170132637024,
|
| 271 |
+
"learning_rate": 9.261538461538461e-06,
|
| 272 |
+
"loss": 0.1313,
|
| 273 |
+
"step": 350
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.769230769230769,
|
| 277 |
+
"grad_norm": 9.631648063659668,
|
| 278 |
+
"learning_rate": 8.953846153846153e-06,
|
| 279 |
+
"loss": 0.1164,
|
| 280 |
+
"step": 360
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 2.8461538461538463,
|
| 284 |
+
"grad_norm": 0.31370919942855835,
|
| 285 |
+
"learning_rate": 8.646153846153846e-06,
|
| 286 |
+
"loss": 0.1535,
|
| 287 |
+
"step": 370
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 2.9230769230769234,
|
| 291 |
+
"grad_norm": 1.1416504383087158,
|
| 292 |
+
"learning_rate": 8.338461538461538e-06,
|
| 293 |
+
"loss": 0.1321,
|
| 294 |
+
"step": 380
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 3.0,
|
| 298 |
+
"grad_norm": 0.2953951358795166,
|
| 299 |
+
"learning_rate": 8.03076923076923e-06,
|
| 300 |
+
"loss": 0.1373,
|
| 301 |
+
"step": 390
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 3.0,
|
| 305 |
+
"eval_accuracy": 0.9774436090225563,
|
| 306 |
+
"eval_loss": 0.09720253199338913,
|
| 307 |
+
"eval_runtime": 0.5922,
|
| 308 |
+
"eval_samples_per_second": 224.569,
|
| 309 |
+
"eval_steps_per_second": 28.704,
|
| 310 |
+
"step": 390
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 3.076923076923077,
|
| 314 |
+
"grad_norm": 2.0271453857421875,
|
| 315 |
+
"learning_rate": 7.723076923076924e-06,
|
| 316 |
+
"loss": 0.1345,
|
| 317 |
+
"step": 400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 3.1538461538461537,
|
| 321 |
+
"grad_norm": 7.004184722900391,
|
| 322 |
+
"learning_rate": 7.4153846153846164e-06,
|
| 323 |
+
"loss": 0.1496,
|
| 324 |
+
"step": 410
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 3.230769230769231,
|
| 328 |
+
"grad_norm": 9.225290298461914,
|
| 329 |
+
"learning_rate": 7.107692307692309e-06,
|
| 330 |
+
"loss": 0.1388,
|
| 331 |
+
"step": 420
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.3076923076923075,
|
| 335 |
+
"grad_norm": 0.2332003116607666,
|
| 336 |
+
"learning_rate": 6.800000000000001e-06,
|
| 337 |
+
"loss": 0.0772,
|
| 338 |
+
"step": 430
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.3846153846153846,
|
| 342 |
+
"grad_norm": 3.141547918319702,
|
| 343 |
+
"learning_rate": 6.492307692307693e-06,
|
| 344 |
+
"loss": 0.1145,
|
| 345 |
+
"step": 440
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.4615384615384617,
|
| 349 |
+
"grad_norm": 12.893486022949219,
|
| 350 |
+
"learning_rate": 6.1846153846153855e-06,
|
| 351 |
+
"loss": 0.0818,
|
| 352 |
+
"step": 450
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.5384615384615383,
|
| 356 |
+
"grad_norm": 5.630913257598877,
|
| 357 |
+
"learning_rate": 5.876923076923078e-06,
|
| 358 |
+
"loss": 0.0948,
|
| 359 |
+
"step": 460
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.6153846153846154,
|
| 363 |
+
"grad_norm": 7.8553266525268555,
|
| 364 |
+
"learning_rate": 5.56923076923077e-06,
|
| 365 |
+
"loss": 0.0719,
|
| 366 |
+
"step": 470
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.6923076923076925,
|
| 370 |
+
"grad_norm": 0.21685053408145905,
|
| 371 |
+
"learning_rate": 5.261538461538462e-06,
|
| 372 |
+
"loss": 0.1,
|
| 373 |
+
"step": 480
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 3.769230769230769,
|
| 377 |
+
"grad_norm": 0.21081575751304626,
|
| 378 |
+
"learning_rate": 4.9538461538461545e-06,
|
| 379 |
+
"loss": 0.069,
|
| 380 |
+
"step": 490
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.8461538461538463,
|
| 384 |
+
"grad_norm": 0.8678991198539734,
|
| 385 |
+
"learning_rate": 4.646153846153847e-06,
|
| 386 |
+
"loss": 0.1374,
|
| 387 |
+
"step": 500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.9230769230769234,
|
| 391 |
+
"grad_norm": 0.20829035341739655,
|
| 392 |
+
"learning_rate": 4.338461538461539e-06,
|
| 393 |
+
"loss": 0.1311,
|
| 394 |
+
"step": 510
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 4.0,
|
| 398 |
+
"grad_norm": 0.30338263511657715,
|
| 399 |
+
"learning_rate": 4.030769230769231e-06,
|
| 400 |
+
"loss": 0.0845,
|
| 401 |
+
"step": 520
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 4.0,
|
| 405 |
+
"eval_accuracy": 0.9924812030075187,
|
| 406 |
+
"eval_loss": 0.06546766310930252,
|
| 407 |
+
"eval_runtime": 0.5826,
|
| 408 |
+
"eval_samples_per_second": 228.276,
|
| 409 |
+
"eval_steps_per_second": 29.178,
|
| 410 |
+
"step": 520
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.076923076923077,
|
| 414 |
+
"grad_norm": 7.523567199707031,
|
| 415 |
+
"learning_rate": 3.723076923076923e-06,
|
| 416 |
+
"loss": 0.0997,
|
| 417 |
+
"step": 530
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 4.153846153846154,
|
| 421 |
+
"grad_norm": 1.7107905149459839,
|
| 422 |
+
"learning_rate": 3.4153846153846154e-06,
|
| 423 |
+
"loss": 0.1668,
|
| 424 |
+
"step": 540
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 4.230769230769231,
|
| 428 |
+
"grad_norm": 0.3665030300617218,
|
| 429 |
+
"learning_rate": 3.1076923076923076e-06,
|
| 430 |
+
"loss": 0.1545,
|
| 431 |
+
"step": 550
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 4.3076923076923075,
|
| 435 |
+
"grad_norm": 1.344637155532837,
|
| 436 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 437 |
+
"loss": 0.0503,
|
| 438 |
+
"step": 560
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 4.384615384615385,
|
| 442 |
+
"grad_norm": 0.27009326219558716,
|
| 443 |
+
"learning_rate": 2.4923076923076926e-06,
|
| 444 |
+
"loss": 0.135,
|
| 445 |
+
"step": 570
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 4.461538461538462,
|
| 449 |
+
"grad_norm": 0.2649720311164856,
|
| 450 |
+
"learning_rate": 2.184615384615385e-06,
|
| 451 |
+
"loss": 0.083,
|
| 452 |
+
"step": 580
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 4.538461538461538,
|
| 456 |
+
"grad_norm": 9.79839038848877,
|
| 457 |
+
"learning_rate": 1.876923076923077e-06,
|
| 458 |
+
"loss": 0.072,
|
| 459 |
+
"step": 590
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 4.615384615384615,
|
| 463 |
+
"grad_norm": 2.8558285236358643,
|
| 464 |
+
"learning_rate": 1.5692307692307693e-06,
|
| 465 |
+
"loss": 0.0905,
|
| 466 |
+
"step": 600
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.6923076923076925,
|
| 470 |
+
"grad_norm": 5.410094738006592,
|
| 471 |
+
"learning_rate": 1.2615384615384616e-06,
|
| 472 |
+
"loss": 0.0874,
|
| 473 |
+
"step": 610
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 4.769230769230769,
|
| 477 |
+
"grad_norm": 0.26254087686538696,
|
| 478 |
+
"learning_rate": 9.53846153846154e-07,
|
| 479 |
+
"loss": 0.0866,
|
| 480 |
+
"step": 620
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 4.846153846153846,
|
| 484 |
+
"grad_norm": 0.2139281928539276,
|
| 485 |
+
"learning_rate": 6.461538461538462e-07,
|
| 486 |
+
"loss": 0.0662,
|
| 487 |
+
"step": 630
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 4.923076923076923,
|
| 491 |
+
"grad_norm": 0.23087060451507568,
|
| 492 |
+
"learning_rate": 3.3846153846153845e-07,
|
| 493 |
+
"loss": 0.0822,
|
| 494 |
+
"step": 640
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 5.0,
|
| 498 |
+
"grad_norm": 0.618672788143158,
|
| 499 |
+
"learning_rate": 3.076923076923077e-08,
|
| 500 |
+
"loss": 0.1147,
|
| 501 |
+
"step": 650
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 5.0,
|
| 505 |
+
"eval_accuracy": 0.9774436090225563,
|
| 506 |
+
"eval_loss": 0.08272561430931091,
|
| 507 |
+
"eval_runtime": 0.6349,
|
| 508 |
+
"eval_samples_per_second": 209.467,
|
| 509 |
+
"eval_steps_per_second": 26.774,
|
| 510 |
+
"step": 650
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 5.0,
|
| 514 |
+
"step": 650,
|
| 515 |
+
"total_flos": 4.006371770595533e+17,
|
| 516 |
+
"train_loss": 0.0,
|
| 517 |
+
"train_runtime": 0.1521,
|
| 518 |
+
"train_samples_per_second": 33994.24,
|
| 519 |
+
"train_steps_per_second": 4273.937
|
| 520 |
+
}
|
| 521 |
+
],
|
| 522 |
+
"logging_steps": 10,
|
| 523 |
+
"max_steps": 650,
|
| 524 |
+
"num_input_tokens_seen": 0,
|
| 525 |
+
"num_train_epochs": 5,
|
| 526 |
+
"save_steps": 500,
|
| 527 |
+
"stateful_callbacks": {
|
| 528 |
+
"TrainerControl": {
|
| 529 |
+
"args": {
|
| 530 |
+
"should_epoch_stop": false,
|
| 531 |
+
"should_evaluate": false,
|
| 532 |
+
"should_log": false,
|
| 533 |
+
"should_save": true,
|
| 534 |
+
"should_training_stop": true
|
| 535 |
+
},
|
| 536 |
+
"attributes": {}
|
| 537 |
+
}
|
| 538 |
+
},
|
| 539 |
+
"total_flos": 4.006371770595533e+17,
|
| 540 |
+
"train_batch_size": 8,
|
| 541 |
+
"trial_name": null,
|
| 542 |
+
"trial_params": null
|
| 543 |
+
}
|