Upload folder using huggingface_hub
Browse files- README.md +27 -17
- all_results.json +15 -18
- dr_results.json +8 -0
- model.safetensors +1 -1
- test_results.json +6 -7
- train_results.json +5 -7
- trainer_state.json +231 -85
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -6,19 +6,19 @@ tags:
|
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
-
- name:
|
| 10 |
results: []
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
|
| 16 |
-
#
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract](https://huggingface.co/microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -43,22 +43,32 @@ The following hyperparameters were used during training:
|
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
-
- num_epochs:
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
-
| Training Loss | Epoch | Step
|
| 51 |
-
|
| 52 |
-
| No log | 1.0 | 791
|
| 53 |
-
| 0.
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
|
| 63 |
|
| 64 |
### Framework versions
|
|
|
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
| 8 |
model-index:
|
| 9 |
+
- name: pubmedbert-abstract
|
| 10 |
results: []
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
|
| 16 |
+
# pubmedbert-abstract
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract](https://huggingface.co/microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.4618
|
| 21 |
+
- Accuracy: 0.9501
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
+
- num_epochs: 20
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 51 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
|
| 52 |
+
| No log | 1.0 | 791 | 0.2340 | 0.9386 |
|
| 53 |
+
| 0.1776 | 2.0 | 1582 | 0.2716 | 0.9419 |
|
| 54 |
+
| 0.0855 | 3.0 | 2373 | 0.2730 | 0.9431 |
|
| 55 |
+
| 0.0627 | 4.0 | 3164 | 0.3323 | 0.9382 |
|
| 56 |
+
| 0.0627 | 5.0 | 3955 | 0.3308 | 0.9451 |
|
| 57 |
+
| 0.0463 | 6.0 | 4746 | 0.3986 | 0.9412 |
|
| 58 |
+
| 0.0308 | 7.0 | 5537 | 0.4211 | 0.9419 |
|
| 59 |
+
| 0.0312 | 8.0 | 6328 | 0.3616 | 0.9437 |
|
| 60 |
+
| 0.0221 | 9.0 | 7119 | 0.4310 | 0.9396 |
|
| 61 |
+
| 0.0221 | 10.0 | 7910 | 0.4222 | 0.9438 |
|
| 62 |
+
| 0.0181 | 11.0 | 8701 | 0.4185 | 0.9445 |
|
| 63 |
+
| 0.0141 | 12.0 | 9492 | 0.4678 | 0.9456 |
|
| 64 |
+
| 0.0133 | 13.0 | 10283 | 0.4027 | 0.9503 |
|
| 65 |
+
| 0.0082 | 14.0 | 11074 | 0.4504 | 0.9473 |
|
| 66 |
+
| 0.0082 | 15.0 | 11865 | 0.4760 | 0.9505 |
|
| 67 |
+
| 0.0052 | 16.0 | 12656 | 0.4573 | 0.9449 |
|
| 68 |
+
| 0.0042 | 17.0 | 13447 | 0.4356 | 0.9522 |
|
| 69 |
+
| 0.0037 | 18.0 | 14238 | 0.4577 | 0.9487 |
|
| 70 |
+
| 0.0024 | 19.0 | 15029 | 0.4642 | 0.9493 |
|
| 71 |
+
| 0.0024 | 20.0 | 15820 | 0.4618 | 0.9501 |
|
| 72 |
|
| 73 |
|
| 74 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,20 +1,17 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
| 8 |
-
"
|
| 9 |
-
"
|
| 10 |
-
"
|
| 11 |
-
"
|
| 12 |
-
"
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"train_samples": 25296,
|
| 18 |
-
"train_samples_per_second": 622.025,
|
| 19 |
-
"train_steps_per_second": 2.434
|
| 20 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"dr_accuracy": 0.9990512333965844,
|
| 3 |
+
"dr_loss": 0.003322172211483121,
|
| 4 |
+
"dr_runtime": 38.9939,
|
| 5 |
+
"dr_samples_per_second": 648.718,
|
| 6 |
+
"dr_steps_per_second": 2.539,
|
| 7 |
+
"epoch": 20.0,
|
| 8 |
+
"test_accuracy": 0.9522393282015396,
|
| 9 |
+
"test_loss": 0.43564239144325256,
|
| 10 |
+
"test_runtime": 9.413,
|
| 11 |
+
"test_samples_per_second": 607.247,
|
| 12 |
+
"test_steps_per_second": 2.443,
|
| 13 |
+
"train_loss": 0.03330213655864846,
|
| 14 |
+
"train_runtime": 3337.005,
|
| 15 |
+
"train_samples_per_second": 151.609,
|
| 16 |
+
"train_steps_per_second": 4.741
|
|
|
|
|
|
|
|
|
|
| 17 |
}
|
dr_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dr_accuracy": 0.9990512333965844,
|
| 3 |
+
"dr_loss": 0.003322172211483121,
|
| 4 |
+
"dr_runtime": 38.9939,
|
| 5 |
+
"dr_samples_per_second": 648.718,
|
| 6 |
+
"dr_steps_per_second": 2.539,
|
| 7 |
+
"epoch": 20.0
|
| 8 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 437967876
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8177fdfdfd82ee15073e43ce61ec4066744df2c00a56772f510f614e05b00a41
|
| 3 |
size 437967876
|
test_results.json
CHANGED
|
@@ -1,9 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
| 8 |
-
"test_steps_per_second": 2.321
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"test_accuracy": 0.9522393282015396,
|
| 4 |
+
"test_loss": 0.43564239144325256,
|
| 5 |
+
"test_runtime": 9.413,
|
| 6 |
+
"test_samples_per_second": 607.247,
|
| 7 |
+
"test_steps_per_second": 2.443
|
|
|
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"train_samples_per_second": 622.025,
|
| 8 |
-
"train_steps_per_second": 2.434
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"train_loss": 0.03330213655864846,
|
| 4 |
+
"train_runtime": 3337.005,
|
| 5 |
+
"train_samples_per_second": 151.609,
|
| 6 |
+
"train_steps_per_second": 4.741
|
|
|
|
|
|
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,168 +1,314 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_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": 1.0,
|
| 13 |
-
"eval_accuracy": 0.
|
| 14 |
-
"eval_loss": 0.
|
| 15 |
-
"eval_runtime":
|
| 16 |
-
"eval_samples_per_second":
|
| 17 |
-
"eval_steps_per_second":
|
| 18 |
"step": 791
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 1.26,
|
| 22 |
-
"grad_norm":
|
| 23 |
-
"learning_rate": 4.
|
| 24 |
-
"loss": 0.
|
| 25 |
"step": 1000
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2.0,
|
| 29 |
-
"eval_accuracy": 0.
|
| 30 |
-
"eval_loss": 0.
|
| 31 |
-
"eval_runtime":
|
| 32 |
-
"eval_samples_per_second":
|
| 33 |
-
"eval_steps_per_second":
|
| 34 |
"step": 1582
|
| 35 |
},
|
| 36 |
{
|
| 37 |
"epoch": 2.53,
|
| 38 |
-
"grad_norm":
|
| 39 |
-
"learning_rate":
|
| 40 |
-
"loss": 0.
|
| 41 |
"step": 2000
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 3.0,
|
| 45 |
-
"eval_accuracy": 0.
|
| 46 |
-
"eval_loss": 0.
|
| 47 |
-
"eval_runtime":
|
| 48 |
-
"eval_samples_per_second":
|
| 49 |
-
"eval_steps_per_second": 2.
|
| 50 |
"step": 2373
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 3.79,
|
| 54 |
-
"grad_norm": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 3000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 4.0,
|
| 61 |
-
"eval_accuracy": 0.
|
| 62 |
-
"eval_loss": 0.
|
| 63 |
-
"eval_runtime":
|
| 64 |
-
"eval_samples_per_second":
|
| 65 |
-
"eval_steps_per_second": 2.
|
| 66 |
"step": 3164
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 5.0,
|
| 70 |
-
"eval_accuracy": 0.
|
| 71 |
-
"eval_loss": 0.
|
| 72 |
-
"eval_runtime":
|
| 73 |
-
"eval_samples_per_second":
|
| 74 |
-
"eval_steps_per_second": 2.
|
| 75 |
"step": 3955
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 5.06,
|
| 79 |
-
"grad_norm": 0.
|
| 80 |
-
"learning_rate":
|
| 81 |
-
"loss": 0.
|
| 82 |
"step": 4000
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 6.0,
|
| 86 |
-
"eval_accuracy": 0.
|
| 87 |
-
"eval_loss": 0.
|
| 88 |
-
"eval_runtime":
|
| 89 |
-
"eval_samples_per_second":
|
| 90 |
-
"eval_steps_per_second": 2.
|
| 91 |
"step": 4746
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 6.32,
|
| 95 |
-
"grad_norm":
|
| 96 |
-
"learning_rate":
|
| 97 |
-
"loss": 0.
|
| 98 |
"step": 5000
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 7.0,
|
| 102 |
-
"eval_accuracy": 0.
|
| 103 |
-
"eval_loss": 0.
|
| 104 |
-
"eval_runtime":
|
| 105 |
-
"eval_samples_per_second":
|
| 106 |
-
"eval_steps_per_second": 2.
|
| 107 |
"step": 5537
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 7.59,
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 6000
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 8.0,
|
| 118 |
-
"eval_accuracy": 0.
|
| 119 |
-
"eval_loss": 0.
|
| 120 |
-
"eval_runtime":
|
| 121 |
-
"eval_samples_per_second":
|
| 122 |
-
"eval_steps_per_second": 2.
|
| 123 |
"step": 6328
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 8.85,
|
| 127 |
-
"grad_norm": 0.
|
| 128 |
-
"learning_rate":
|
| 129 |
-
"loss": 0.
|
| 130 |
"step": 7000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 9.0,
|
| 134 |
-
"eval_accuracy": 0.
|
| 135 |
-
"eval_loss": 0.
|
| 136 |
-
"eval_runtime":
|
| 137 |
-
"eval_samples_per_second":
|
| 138 |
-
"eval_steps_per_second": 2.
|
| 139 |
"step": 7119
|
| 140 |
},
|
| 141 |
{
|
| 142 |
"epoch": 10.0,
|
| 143 |
-
"eval_accuracy": 0.
|
| 144 |
-
"eval_loss": 0.
|
| 145 |
-
"eval_runtime":
|
| 146 |
-
"eval_samples_per_second":
|
| 147 |
-
"eval_steps_per_second": 2.
|
| 148 |
"step": 7910
|
| 149 |
},
|
| 150 |
{
|
| 151 |
-
"epoch": 10.
|
| 152 |
-
"
|
| 153 |
-
"
|
| 154 |
-
"
|
| 155 |
-
"
|
| 156 |
-
|
| 157 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
}
|
| 159 |
],
|
| 160 |
"logging_steps": 1000,
|
| 161 |
-
"max_steps":
|
| 162 |
"num_input_tokens_seen": 0,
|
| 163 |
-
"num_train_epochs":
|
| 164 |
"save_steps": 500,
|
| 165 |
-
"total_flos":
|
| 166 |
"train_batch_size": 32,
|
| 167 |
"trial_name": null,
|
| 168 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9522393282015396,
|
| 3 |
+
"best_model_checkpoint": "../../checkpoint/ddi/pubmedbert-abstract/checkpoint-13447",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 15820,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 1.0,
|
| 13 |
+
"eval_accuracy": 0.938593421973408,
|
| 14 |
+
"eval_loss": 0.23401664197444916,
|
| 15 |
+
"eval_runtime": 9.3525,
|
| 16 |
+
"eval_samples_per_second": 611.174,
|
| 17 |
+
"eval_steps_per_second": 2.459,
|
| 18 |
"step": 791
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 1.26,
|
| 22 |
+
"grad_norm": 22.48932647705078,
|
| 23 |
+
"learning_rate": 4.683944374209861e-05,
|
| 24 |
+
"loss": 0.1776,
|
| 25 |
"step": 1000
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2.0,
|
| 29 |
+
"eval_accuracy": 0.9419174247725682,
|
| 30 |
+
"eval_loss": 0.27159374952316284,
|
| 31 |
+
"eval_runtime": 9.4015,
|
| 32 |
+
"eval_samples_per_second": 607.985,
|
| 33 |
+
"eval_steps_per_second": 2.446,
|
| 34 |
"step": 1582
|
| 35 |
},
|
| 36 |
{
|
| 37 |
"epoch": 2.53,
|
| 38 |
+
"grad_norm": 0.30297374725341797,
|
| 39 |
+
"learning_rate": 4.367888748419722e-05,
|
| 40 |
+
"loss": 0.0855,
|
| 41 |
"step": 2000
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 3.0,
|
| 45 |
+
"eval_accuracy": 0.9431420573827851,
|
| 46 |
+
"eval_loss": 0.27298399806022644,
|
| 47 |
+
"eval_runtime": 9.4208,
|
| 48 |
+
"eval_samples_per_second": 606.741,
|
| 49 |
+
"eval_steps_per_second": 2.441,
|
| 50 |
"step": 2373
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 3.79,
|
| 54 |
+
"grad_norm": 0.08143208175897598,
|
| 55 |
+
"learning_rate": 4.051833122629583e-05,
|
| 56 |
+
"loss": 0.0627,
|
| 57 |
"step": 3000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 4.0,
|
| 61 |
+
"eval_accuracy": 0.9382435269419174,
|
| 62 |
+
"eval_loss": 0.33226683735847473,
|
| 63 |
+
"eval_runtime": 9.4056,
|
| 64 |
+
"eval_samples_per_second": 607.725,
|
| 65 |
+
"eval_steps_per_second": 2.445,
|
| 66 |
"step": 3164
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 5.0,
|
| 70 |
+
"eval_accuracy": 0.9450664800559833,
|
| 71 |
+
"eval_loss": 0.3307989239692688,
|
| 72 |
+
"eval_runtime": 9.3945,
|
| 73 |
+
"eval_samples_per_second": 608.444,
|
| 74 |
+
"eval_steps_per_second": 2.448,
|
| 75 |
"step": 3955
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 5.06,
|
| 79 |
+
"grad_norm": 0.012913365848362446,
|
| 80 |
+
"learning_rate": 3.735777496839444e-05,
|
| 81 |
+
"loss": 0.0463,
|
| 82 |
"step": 4000
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 6.0,
|
| 86 |
+
"eval_accuracy": 0.9412176347095871,
|
| 87 |
+
"eval_loss": 0.39864641427993774,
|
| 88 |
+
"eval_runtime": 9.3742,
|
| 89 |
+
"eval_samples_per_second": 609.757,
|
| 90 |
+
"eval_steps_per_second": 2.454,
|
| 91 |
"step": 4746
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 6.32,
|
| 95 |
+
"grad_norm": 2.0644302368164062,
|
| 96 |
+
"learning_rate": 3.419721871049305e-05,
|
| 97 |
+
"loss": 0.0308,
|
| 98 |
"step": 5000
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 7.0,
|
| 102 |
+
"eval_accuracy": 0.9419174247725682,
|
| 103 |
+
"eval_loss": 0.42111822962760925,
|
| 104 |
+
"eval_runtime": 9.3868,
|
| 105 |
+
"eval_samples_per_second": 608.939,
|
| 106 |
+
"eval_steps_per_second": 2.45,
|
| 107 |
"step": 5537
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 7.59,
|
| 111 |
+
"grad_norm": 30.287784576416016,
|
| 112 |
+
"learning_rate": 3.1036662452591655e-05,
|
| 113 |
+
"loss": 0.0312,
|
| 114 |
"step": 6000
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 8.0,
|
| 118 |
+
"eval_accuracy": 0.943666899930021,
|
| 119 |
+
"eval_loss": 0.36164581775665283,
|
| 120 |
+
"eval_runtime": 9.3894,
|
| 121 |
+
"eval_samples_per_second": 608.769,
|
| 122 |
+
"eval_steps_per_second": 2.45,
|
| 123 |
"step": 6328
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 8.85,
|
| 127 |
+
"grad_norm": 0.03705460578203201,
|
| 128 |
+
"learning_rate": 2.7876106194690264e-05,
|
| 129 |
+
"loss": 0.0221,
|
| 130 |
"step": 7000
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 9.0,
|
| 134 |
+
"eval_accuracy": 0.9396431070678797,
|
| 135 |
+
"eval_loss": 0.4309725761413574,
|
| 136 |
+
"eval_runtime": 9.4203,
|
| 137 |
+
"eval_samples_per_second": 606.777,
|
| 138 |
+
"eval_steps_per_second": 2.442,
|
| 139 |
"step": 7119
|
| 140 |
},
|
| 141 |
{
|
| 142 |
"epoch": 10.0,
|
| 143 |
+
"eval_accuracy": 0.9438418474457663,
|
| 144 |
+
"eval_loss": 0.4221705198287964,
|
| 145 |
+
"eval_runtime": 9.4325,
|
| 146 |
+
"eval_samples_per_second": 605.989,
|
| 147 |
+
"eval_steps_per_second": 2.438,
|
| 148 |
"step": 7910
|
| 149 |
},
|
| 150 |
{
|
| 151 |
+
"epoch": 10.11,
|
| 152 |
+
"grad_norm": 0.11050642281770706,
|
| 153 |
+
"learning_rate": 2.4715549936788876e-05,
|
| 154 |
+
"loss": 0.0181,
|
| 155 |
+
"step": 8000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 11.0,
|
| 159 |
+
"eval_accuracy": 0.9445416375087474,
|
| 160 |
+
"eval_loss": 0.4184616804122925,
|
| 161 |
+
"eval_runtime": 9.4176,
|
| 162 |
+
"eval_samples_per_second": 606.948,
|
| 163 |
+
"eval_steps_per_second": 2.442,
|
| 164 |
+
"step": 8701
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 11.38,
|
| 168 |
+
"grad_norm": 0.03253033012151718,
|
| 169 |
+
"learning_rate": 2.1554993678887485e-05,
|
| 170 |
+
"loss": 0.0141,
|
| 171 |
+
"step": 9000
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 12.0,
|
| 175 |
+
"eval_accuracy": 0.945591322603219,
|
| 176 |
+
"eval_loss": 0.46782681345939636,
|
| 177 |
+
"eval_runtime": 9.4133,
|
| 178 |
+
"eval_samples_per_second": 607.226,
|
| 179 |
+
"eval_steps_per_second": 2.443,
|
| 180 |
+
"step": 9492
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 12.64,
|
| 184 |
+
"grad_norm": 0.005740019958466291,
|
| 185 |
+
"learning_rate": 1.8394437420986094e-05,
|
| 186 |
+
"loss": 0.0133,
|
| 187 |
+
"step": 10000
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 13.0,
|
| 191 |
+
"eval_accuracy": 0.9503149055283415,
|
| 192 |
+
"eval_loss": 0.40269356966018677,
|
| 193 |
+
"eval_runtime": 9.4053,
|
| 194 |
+
"eval_samples_per_second": 607.74,
|
| 195 |
+
"eval_steps_per_second": 2.445,
|
| 196 |
+
"step": 10283
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 13.91,
|
| 200 |
+
"grad_norm": 0.0024239453487098217,
|
| 201 |
+
"learning_rate": 1.5233881163084704e-05,
|
| 202 |
+
"loss": 0.0082,
|
| 203 |
+
"step": 11000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 14.0,
|
| 207 |
+
"eval_accuracy": 0.9473407977606718,
|
| 208 |
+
"eval_loss": 0.45041143894195557,
|
| 209 |
+
"eval_runtime": 9.639,
|
| 210 |
+
"eval_samples_per_second": 593.006,
|
| 211 |
+
"eval_steps_per_second": 2.386,
|
| 212 |
+
"step": 11074
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 15.0,
|
| 216 |
+
"eval_accuracy": 0.9504898530440867,
|
| 217 |
+
"eval_loss": 0.47598323225975037,
|
| 218 |
+
"eval_runtime": 9.4326,
|
| 219 |
+
"eval_samples_per_second": 605.982,
|
| 220 |
+
"eval_steps_per_second": 2.438,
|
| 221 |
+
"step": 11865
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 15.17,
|
| 225 |
+
"grad_norm": 0.0034840325824916363,
|
| 226 |
+
"learning_rate": 1.2073324905183313e-05,
|
| 227 |
+
"loss": 0.0052,
|
| 228 |
+
"step": 12000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 16.0,
|
| 232 |
+
"eval_accuracy": 0.9448915325402379,
|
| 233 |
+
"eval_loss": 0.45725104212760925,
|
| 234 |
+
"eval_runtime": 9.4366,
|
| 235 |
+
"eval_samples_per_second": 605.728,
|
| 236 |
+
"eval_steps_per_second": 2.437,
|
| 237 |
+
"step": 12656
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 16.43,
|
| 241 |
+
"grad_norm": 0.4831530451774597,
|
| 242 |
+
"learning_rate": 8.912768647281922e-06,
|
| 243 |
+
"loss": 0.0042,
|
| 244 |
+
"step": 13000
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 17.0,
|
| 248 |
+
"eval_accuracy": 0.9522393282015396,
|
| 249 |
+
"eval_loss": 0.43564239144325256,
|
| 250 |
+
"eval_runtime": 9.4029,
|
| 251 |
+
"eval_samples_per_second": 607.9,
|
| 252 |
+
"eval_steps_per_second": 2.446,
|
| 253 |
+
"step": 13447
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 17.7,
|
| 257 |
+
"grad_norm": 0.0005753316800110042,
|
| 258 |
+
"learning_rate": 5.752212389380531e-06,
|
| 259 |
+
"loss": 0.0037,
|
| 260 |
+
"step": 14000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 18.0,
|
| 264 |
+
"eval_accuracy": 0.948740377886634,
|
| 265 |
+
"eval_loss": 0.45767056941986084,
|
| 266 |
+
"eval_runtime": 9.3872,
|
| 267 |
+
"eval_samples_per_second": 608.913,
|
| 268 |
+
"eval_steps_per_second": 2.45,
|
| 269 |
+
"step": 14238
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 18.96,
|
| 273 |
+
"grad_norm": 0.0019946701359003782,
|
| 274 |
+
"learning_rate": 2.59165613147914e-06,
|
| 275 |
+
"loss": 0.0024,
|
| 276 |
+
"step": 15000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 19.0,
|
| 280 |
+
"eval_accuracy": 0.9492652204338698,
|
| 281 |
+
"eval_loss": 0.4641525149345398,
|
| 282 |
+
"eval_runtime": 9.4075,
|
| 283 |
+
"eval_samples_per_second": 607.599,
|
| 284 |
+
"eval_steps_per_second": 2.445,
|
| 285 |
+
"step": 15029
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 20.0,
|
| 289 |
+
"eval_accuracy": 0.9501399580125962,
|
| 290 |
+
"eval_loss": 0.46178367733955383,
|
| 291 |
+
"eval_runtime": 9.3958,
|
| 292 |
+
"eval_samples_per_second": 608.356,
|
| 293 |
+
"eval_steps_per_second": 2.448,
|
| 294 |
+
"step": 15820
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 20.0,
|
| 298 |
+
"step": 15820,
|
| 299 |
+
"total_flos": 3.327918265884672e+16,
|
| 300 |
+
"train_loss": 0.03330213655864846,
|
| 301 |
+
"train_runtime": 3337.005,
|
| 302 |
+
"train_samples_per_second": 151.609,
|
| 303 |
+
"train_steps_per_second": 4.741
|
| 304 |
}
|
| 305 |
],
|
| 306 |
"logging_steps": 1000,
|
| 307 |
+
"max_steps": 15820,
|
| 308 |
"num_input_tokens_seen": 0,
|
| 309 |
+
"num_train_epochs": 20,
|
| 310 |
"save_steps": 500,
|
| 311 |
+
"total_flos": 3.327918265884672e+16,
|
| 312 |
"train_batch_size": 32,
|
| 313 |
"trial_name": null,
|
| 314 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea63ec5bf6fd89b86ae59f8fb9e574688ea61151666e13315351c52df7380d6c
|
| 3 |
+
size 4984
|