Upload folder using huggingface_hub
Browse files- all_results.json +13 -0
- config.json +49 -0
- eval_results.json +8 -0
- model.safetensors +3 -0
- preprocessor_config.json +22 -0
- train_results.json +8 -0
- trainer_state.json +1855 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.0,
|
| 3 |
+
"eval_f1": 0.9760225669957687,
|
| 4 |
+
"eval_loss": 0.06743501126766205,
|
| 5 |
+
"eval_runtime": 16.9709,
|
| 6 |
+
"eval_samples_per_second": 215.663,
|
| 7 |
+
"eval_steps_per_second": 6.776,
|
| 8 |
+
"total_flos": 2.1045216318570566e+19,
|
| 9 |
+
"train_loss": 0.12371139868869004,
|
| 10 |
+
"train_runtime": 2341.9199,
|
| 11 |
+
"train_samples_per_second": 171.111,
|
| 12 |
+
"train_steps_per_second": 5.347
|
| 13 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/resnet-152",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ResNetForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"depths": [
|
| 7 |
+
3,
|
| 8 |
+
8,
|
| 9 |
+
36,
|
| 10 |
+
3
|
| 11 |
+
],
|
| 12 |
+
"downsample_in_bottleneck": false,
|
| 13 |
+
"downsample_in_first_stage": false,
|
| 14 |
+
"embedding_size": 64,
|
| 15 |
+
"hidden_act": "relu",
|
| 16 |
+
"hidden_sizes": [
|
| 17 |
+
256,
|
| 18 |
+
512,
|
| 19 |
+
1024,
|
| 20 |
+
2048
|
| 21 |
+
],
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "benign",
|
| 24 |
+
"1": "malignant"
|
| 25 |
+
},
|
| 26 |
+
"label2id": {
|
| 27 |
+
"benign": 0,
|
| 28 |
+
"malignant": 1
|
| 29 |
+
},
|
| 30 |
+
"layer_type": "bottleneck",
|
| 31 |
+
"model_type": "resnet",
|
| 32 |
+
"num_channels": 3,
|
| 33 |
+
"out_features": [
|
| 34 |
+
"stage4"
|
| 35 |
+
],
|
| 36 |
+
"out_indices": [
|
| 37 |
+
4
|
| 38 |
+
],
|
| 39 |
+
"problem_type": "single_label_classification",
|
| 40 |
+
"stage_names": [
|
| 41 |
+
"stem",
|
| 42 |
+
"stage1",
|
| 43 |
+
"stage2",
|
| 44 |
+
"stage3",
|
| 45 |
+
"stage4"
|
| 46 |
+
],
|
| 47 |
+
"torch_dtype": "float32",
|
| 48 |
+
"transformers_version": "4.48.2"
|
| 49 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.0,
|
| 3 |
+
"eval_f1": 0.9760225669957687,
|
| 4 |
+
"eval_loss": 0.06743501126766205,
|
| 5 |
+
"eval_runtime": 16.9709,
|
| 6 |
+
"eval_samples_per_second": 215.663,
|
| 7 |
+
"eval_steps_per_second": 6.776
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e248dadf16886387e947c42b544a5a07efdb7de3faed579f57858ec29b428b8
|
| 3 |
+
size 233321808
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_pct": 0.875,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.485,
|
| 8 |
+
0.456,
|
| 9 |
+
0.406
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ConvNextImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.229,
|
| 14 |
+
0.224,
|
| 15 |
+
0.225
|
| 16 |
+
],
|
| 17 |
+
"resample": 3,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"shortest_edge": 224
|
| 21 |
+
}
|
| 22 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.0,
|
| 3 |
+
"total_flos": 2.1045216318570566e+19,
|
| 4 |
+
"train_loss": 0.12371139868869004,
|
| 5 |
+
"train_runtime": 2341.9199,
|
| 6 |
+
"train_samples_per_second": 171.111,
|
| 7 |
+
"train_steps_per_second": 5.347
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1855 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9760225669957687,
|
| 3 |
+
"best_model_checkpoint": "melanoma-v3\\checkpoint-12523",
|
| 4 |
+
"epoch": 7.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 12523,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.027948574622694244,
|
| 13 |
+
"grad_norm": 9.020837783813477,
|
| 14 |
+
"learning_rate": 1.8355945730247407e-05,
|
| 15 |
+
"loss": 0.6846,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05589714924538849,
|
| 20 |
+
"grad_norm": 15.282539367675781,
|
| 21 |
+
"learning_rate": 3.790901835594573e-05,
|
| 22 |
+
"loss": 0.6027,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08384572386808273,
|
| 27 |
+
"grad_norm": 17.592166900634766,
|
| 28 |
+
"learning_rate": 5.78611332801277e-05,
|
| 29 |
+
"loss": 0.4504,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11179429849077697,
|
| 34 |
+
"grad_norm": 11.997602462768555,
|
| 35 |
+
"learning_rate": 7.781324820430965e-05,
|
| 36 |
+
"loss": 0.3523,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.13974287311347122,
|
| 41 |
+
"grad_norm": 16.85348129272461,
|
| 42 |
+
"learning_rate": 9.776536312849161e-05,
|
| 43 |
+
"loss": 0.2612,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.16769144773616546,
|
| 48 |
+
"grad_norm": 5.394933700561523,
|
| 49 |
+
"learning_rate": 0.00011771747805267359,
|
| 50 |
+
"loss": 0.2493,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.1956400223588597,
|
| 55 |
+
"grad_norm": 15.290493965148926,
|
| 56 |
+
"learning_rate": 0.00013766959297685554,
|
| 57 |
+
"loss": 0.2092,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.22358859698155395,
|
| 62 |
+
"grad_norm": 15.310522079467773,
|
| 63 |
+
"learning_rate": 0.00015762170790103753,
|
| 64 |
+
"loss": 0.1995,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2515371716042482,
|
| 69 |
+
"grad_norm": 20.243497848510742,
|
| 70 |
+
"learning_rate": 0.00017757382282521946,
|
| 71 |
+
"loss": 0.2205,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.27948574622694244,
|
| 76 |
+
"grad_norm": 5.062504291534424,
|
| 77 |
+
"learning_rate": 0.00019752593774940145,
|
| 78 |
+
"loss": 0.2124,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3074343208496367,
|
| 83 |
+
"grad_norm": 5.295910358428955,
|
| 84 |
+
"learning_rate": 0.00021747805267358338,
|
| 85 |
+
"loss": 0.2178,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3353828954723309,
|
| 90 |
+
"grad_norm": 1.7261399030685425,
|
| 91 |
+
"learning_rate": 0.00023743016759776537,
|
| 92 |
+
"loss": 0.1946,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.36333147009502514,
|
| 97 |
+
"grad_norm": 5.8339338302612305,
|
| 98 |
+
"learning_rate": 0.00025738228252194736,
|
| 99 |
+
"loss": 0.2097,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3912800447177194,
|
| 104 |
+
"grad_norm": 2.142676830291748,
|
| 105 |
+
"learning_rate": 0.0002773343974461293,
|
| 106 |
+
"loss": 0.2165,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.41922861934041367,
|
| 111 |
+
"grad_norm": 2.970517873764038,
|
| 112 |
+
"learning_rate": 0.00029728651237031123,
|
| 113 |
+
"loss": 0.2051,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4471771939631079,
|
| 118 |
+
"grad_norm": 4.551942348480225,
|
| 119 |
+
"learning_rate": 0.0003172386272944932,
|
| 120 |
+
"loss": 0.2288,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.47512576858580213,
|
| 125 |
+
"grad_norm": 5.390867710113525,
|
| 126 |
+
"learning_rate": 0.0003371907422186752,
|
| 127 |
+
"loss": 0.2091,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5030743432084964,
|
| 132 |
+
"grad_norm": 1.3094195127487183,
|
| 133 |
+
"learning_rate": 0.00035714285714285714,
|
| 134 |
+
"loss": 0.2302,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5310229178311906,
|
| 139 |
+
"grad_norm": 2.8628153800964355,
|
| 140 |
+
"learning_rate": 0.00037709497206703913,
|
| 141 |
+
"loss": 0.2244,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5589714924538849,
|
| 146 |
+
"grad_norm": 2.0258989334106445,
|
| 147 |
+
"learning_rate": 0.00039704708699122107,
|
| 148 |
+
"loss": 0.1956,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5869200670765791,
|
| 153 |
+
"grad_norm": 2.2154078483581543,
|
| 154 |
+
"learning_rate": 0.000416999201915403,
|
| 155 |
+
"loss": 0.2345,
|
| 156 |
+
"step": 1050
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6148686416992734,
|
| 160 |
+
"grad_norm": 1.6475298404693604,
|
| 161 |
+
"learning_rate": 0.000436951316839585,
|
| 162 |
+
"loss": 0.2137,
|
| 163 |
+
"step": 1100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6428172163219675,
|
| 167 |
+
"grad_norm": 2.1503453254699707,
|
| 168 |
+
"learning_rate": 0.000456903431763767,
|
| 169 |
+
"loss": 0.2324,
|
| 170 |
+
"step": 1150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6707657909446618,
|
| 174 |
+
"grad_norm": 1.5340570211410522,
|
| 175 |
+
"learning_rate": 0.00047685554668794897,
|
| 176 |
+
"loss": 0.2288,
|
| 177 |
+
"step": 1200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6987143655673561,
|
| 181 |
+
"grad_norm": 2.333324432373047,
|
| 182 |
+
"learning_rate": 0.0004968076616121309,
|
| 183 |
+
"loss": 0.2061,
|
| 184 |
+
"step": 1250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7266629401900503,
|
| 188 |
+
"grad_norm": 1.712575912475586,
|
| 189 |
+
"learning_rate": 0.000498136645962733,
|
| 190 |
+
"loss": 0.2189,
|
| 191 |
+
"step": 1300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7546115148127446,
|
| 195 |
+
"grad_norm": 1.3168742656707764,
|
| 196 |
+
"learning_rate": 0.0004959183673469388,
|
| 197 |
+
"loss": 0.1978,
|
| 198 |
+
"step": 1350
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7825600894354388,
|
| 202 |
+
"grad_norm": 4.792356967926025,
|
| 203 |
+
"learning_rate": 0.0004937000887311446,
|
| 204 |
+
"loss": 0.2333,
|
| 205 |
+
"step": 1400
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.810508664058133,
|
| 209 |
+
"grad_norm": 1.0282349586486816,
|
| 210 |
+
"learning_rate": 0.0004914818101153504,
|
| 211 |
+
"loss": 0.23,
|
| 212 |
+
"step": 1450
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.8384572386808273,
|
| 216 |
+
"grad_norm": 0.6257219910621643,
|
| 217 |
+
"learning_rate": 0.0004892635314995564,
|
| 218 |
+
"loss": 0.2082,
|
| 219 |
+
"step": 1500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8664058133035215,
|
| 223 |
+
"grad_norm": 1.1651033163070679,
|
| 224 |
+
"learning_rate": 0.00048704525288376224,
|
| 225 |
+
"loss": 0.1976,
|
| 226 |
+
"step": 1550
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8943543879262158,
|
| 230 |
+
"grad_norm": 1.0239338874816895,
|
| 231 |
+
"learning_rate": 0.00048482697426796806,
|
| 232 |
+
"loss": 0.1945,
|
| 233 |
+
"step": 1600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.92230296254891,
|
| 237 |
+
"grad_norm": 1.616534948348999,
|
| 238 |
+
"learning_rate": 0.0004826086956521739,
|
| 239 |
+
"loss": 0.1801,
|
| 240 |
+
"step": 1650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.9502515371716043,
|
| 244 |
+
"grad_norm": 1.4642572402954102,
|
| 245 |
+
"learning_rate": 0.00048039041703637974,
|
| 246 |
+
"loss": 0.1888,
|
| 247 |
+
"step": 1700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9782001117942984,
|
| 251 |
+
"grad_norm": 5.763888359069824,
|
| 252 |
+
"learning_rate": 0.00047817213842058566,
|
| 253 |
+
"loss": 0.1798,
|
| 254 |
+
"step": 1750
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0,
|
| 258 |
+
"eval_f1": 0.8859060402684564,
|
| 259 |
+
"eval_loss": 0.2657495439052582,
|
| 260 |
+
"eval_runtime": 17.3445,
|
| 261 |
+
"eval_samples_per_second": 211.017,
|
| 262 |
+
"eval_steps_per_second": 6.63,
|
| 263 |
+
"step": 1789
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.0061486864169928,
|
| 267 |
+
"grad_norm": 0.4531048536300659,
|
| 268 |
+
"learning_rate": 0.0004759538598047915,
|
| 269 |
+
"loss": 0.1844,
|
| 270 |
+
"step": 1800
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.034097261039687,
|
| 274 |
+
"grad_norm": 1.737192153930664,
|
| 275 |
+
"learning_rate": 0.00047373558118899735,
|
| 276 |
+
"loss": 0.2026,
|
| 277 |
+
"step": 1850
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.0620458356623812,
|
| 281 |
+
"grad_norm": 1.7833003997802734,
|
| 282 |
+
"learning_rate": 0.00047151730257320317,
|
| 283 |
+
"loss": 0.1724,
|
| 284 |
+
"step": 1900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.0899944102850754,
|
| 288 |
+
"grad_norm": 1.5082063674926758,
|
| 289 |
+
"learning_rate": 0.0004692990239574091,
|
| 290 |
+
"loss": 0.1891,
|
| 291 |
+
"step": 1950
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.1179429849077698,
|
| 295 |
+
"grad_norm": 1.3961833715438843,
|
| 296 |
+
"learning_rate": 0.00046708074534161496,
|
| 297 |
+
"loss": 0.1737,
|
| 298 |
+
"step": 2000
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.145891559530464,
|
| 302 |
+
"grad_norm": 0.9128915667533875,
|
| 303 |
+
"learning_rate": 0.0004648624667258208,
|
| 304 |
+
"loss": 0.1584,
|
| 305 |
+
"step": 2050
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.1738401341531581,
|
| 309 |
+
"grad_norm": 1.8712306022644043,
|
| 310 |
+
"learning_rate": 0.00046264418811002664,
|
| 311 |
+
"loss": 0.1765,
|
| 312 |
+
"step": 2100
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.2017887087758523,
|
| 316 |
+
"grad_norm": 1.0604172945022583,
|
| 317 |
+
"learning_rate": 0.00046042590949423246,
|
| 318 |
+
"loss": 0.2011,
|
| 319 |
+
"step": 2150
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.2297372833985467,
|
| 323 |
+
"grad_norm": 2.5036442279815674,
|
| 324 |
+
"learning_rate": 0.0004582076308784384,
|
| 325 |
+
"loss": 0.1873,
|
| 326 |
+
"step": 2200
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.2576858580212409,
|
| 330 |
+
"grad_norm": 2.0800774097442627,
|
| 331 |
+
"learning_rate": 0.0004559893522626442,
|
| 332 |
+
"loss": 0.1795,
|
| 333 |
+
"step": 2250
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.2856344326439353,
|
| 337 |
+
"grad_norm": 1.3409812450408936,
|
| 338 |
+
"learning_rate": 0.00045377107364685007,
|
| 339 |
+
"loss": 0.1708,
|
| 340 |
+
"step": 2300
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.3135830072666295,
|
| 344 |
+
"grad_norm": 1.0831228494644165,
|
| 345 |
+
"learning_rate": 0.0004515527950310559,
|
| 346 |
+
"loss": 0.1738,
|
| 347 |
+
"step": 2350
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.3415315818893236,
|
| 351 |
+
"grad_norm": 0.5806594491004944,
|
| 352 |
+
"learning_rate": 0.00044933451641526175,
|
| 353 |
+
"loss": 0.1659,
|
| 354 |
+
"step": 2400
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.3694801565120178,
|
| 358 |
+
"grad_norm": 0.9568463563919067,
|
| 359 |
+
"learning_rate": 0.0004471162377994676,
|
| 360 |
+
"loss": 0.1621,
|
| 361 |
+
"step": 2450
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.397428731134712,
|
| 365 |
+
"grad_norm": 0.8637075424194336,
|
| 366 |
+
"learning_rate": 0.0004448979591836735,
|
| 367 |
+
"loss": 0.2007,
|
| 368 |
+
"step": 2500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.4253773057574064,
|
| 372 |
+
"grad_norm": 1.1704365015029907,
|
| 373 |
+
"learning_rate": 0.0004426796805678793,
|
| 374 |
+
"loss": 0.1556,
|
| 375 |
+
"step": 2550
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.4533258803801006,
|
| 379 |
+
"grad_norm": 0.8730977773666382,
|
| 380 |
+
"learning_rate": 0.0004404614019520852,
|
| 381 |
+
"loss": 0.1741,
|
| 382 |
+
"step": 2600
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.481274455002795,
|
| 386 |
+
"grad_norm": 0.7711090445518494,
|
| 387 |
+
"learning_rate": 0.00043824312333629105,
|
| 388 |
+
"loss": 0.1628,
|
| 389 |
+
"step": 2650
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.5092230296254892,
|
| 393 |
+
"grad_norm": 0.6016321182250977,
|
| 394 |
+
"learning_rate": 0.0004360248447204969,
|
| 395 |
+
"loss": 0.1393,
|
| 396 |
+
"step": 2700
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.5371716042481833,
|
| 400 |
+
"grad_norm": 0.9047902822494507,
|
| 401 |
+
"learning_rate": 0.00043380656610470273,
|
| 402 |
+
"loss": 0.1607,
|
| 403 |
+
"step": 2750
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.5651201788708775,
|
| 407 |
+
"grad_norm": 1.0270776748657227,
|
| 408 |
+
"learning_rate": 0.0004315882874889086,
|
| 409 |
+
"loss": 0.1736,
|
| 410 |
+
"step": 2800
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.5930687534935717,
|
| 414 |
+
"grad_norm": 0.7460580468177795,
|
| 415 |
+
"learning_rate": 0.0004293700088731145,
|
| 416 |
+
"loss": 0.1627,
|
| 417 |
+
"step": 2850
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.621017328116266,
|
| 421 |
+
"grad_norm": 0.8114007115364075,
|
| 422 |
+
"learning_rate": 0.00042715173025732034,
|
| 423 |
+
"loss": 0.1756,
|
| 424 |
+
"step": 2900
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.6489659027389603,
|
| 428 |
+
"grad_norm": 0.7761090397834778,
|
| 429 |
+
"learning_rate": 0.0004249334516415262,
|
| 430 |
+
"loss": 0.1491,
|
| 431 |
+
"step": 2950
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.6769144773616547,
|
| 435 |
+
"grad_norm": 0.7432196140289307,
|
| 436 |
+
"learning_rate": 0.00042271517302573203,
|
| 437 |
+
"loss": 0.1663,
|
| 438 |
+
"step": 3000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.7048630519843488,
|
| 442 |
+
"grad_norm": 0.5656224489212036,
|
| 443 |
+
"learning_rate": 0.0004204968944099379,
|
| 444 |
+
"loss": 0.1808,
|
| 445 |
+
"step": 3050
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.732811626607043,
|
| 449 |
+
"grad_norm": 1.192177653312683,
|
| 450 |
+
"learning_rate": 0.00041827861579414377,
|
| 451 |
+
"loss": 0.1605,
|
| 452 |
+
"step": 3100
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.7607602012297372,
|
| 456 |
+
"grad_norm": 0.8815297484397888,
|
| 457 |
+
"learning_rate": 0.00041606033717834964,
|
| 458 |
+
"loss": 0.1595,
|
| 459 |
+
"step": 3150
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.7887087758524314,
|
| 463 |
+
"grad_norm": 0.7288667559623718,
|
| 464 |
+
"learning_rate": 0.00041384205856255545,
|
| 465 |
+
"loss": 0.1529,
|
| 466 |
+
"step": 3200
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.8166573504751258,
|
| 470 |
+
"grad_norm": 0.41470789909362793,
|
| 471 |
+
"learning_rate": 0.0004116237799467613,
|
| 472 |
+
"loss": 0.1661,
|
| 473 |
+
"step": 3250
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.8446059250978202,
|
| 477 |
+
"grad_norm": 0.5929239392280579,
|
| 478 |
+
"learning_rate": 0.00040940550133096714,
|
| 479 |
+
"loss": 0.1658,
|
| 480 |
+
"step": 3300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.8725544997205144,
|
| 484 |
+
"grad_norm": 0.5527342557907104,
|
| 485 |
+
"learning_rate": 0.00040718722271517306,
|
| 486 |
+
"loss": 0.1686,
|
| 487 |
+
"step": 3350
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.9005030743432085,
|
| 491 |
+
"grad_norm": 0.7223379611968994,
|
| 492 |
+
"learning_rate": 0.0004049689440993789,
|
| 493 |
+
"loss": 0.1616,
|
| 494 |
+
"step": 3400
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.9284516489659027,
|
| 498 |
+
"grad_norm": 1.6212414503097534,
|
| 499 |
+
"learning_rate": 0.00040275066548358475,
|
| 500 |
+
"loss": 0.1455,
|
| 501 |
+
"step": 3450
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.9564002235885969,
|
| 505 |
+
"grad_norm": 0.5442690849304199,
|
| 506 |
+
"learning_rate": 0.00040053238686779056,
|
| 507 |
+
"loss": 0.1666,
|
| 508 |
+
"step": 3500
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.984348798211291,
|
| 512 |
+
"grad_norm": 0.4307806193828583,
|
| 513 |
+
"learning_rate": 0.00039831410825199643,
|
| 514 |
+
"loss": 0.1634,
|
| 515 |
+
"step": 3550
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.0,
|
| 519 |
+
"eval_f1": 0.8839446782922429,
|
| 520 |
+
"eval_loss": 0.26385268568992615,
|
| 521 |
+
"eval_runtime": 15.9608,
|
| 522 |
+
"eval_samples_per_second": 229.312,
|
| 523 |
+
"eval_steps_per_second": 7.205,
|
| 524 |
+
"step": 3578
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.0122973728339857,
|
| 528 |
+
"grad_norm": 1.0025759935379028,
|
| 529 |
+
"learning_rate": 0.00039609582963620236,
|
| 530 |
+
"loss": 0.16,
|
| 531 |
+
"step": 3600
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.04024594745668,
|
| 535 |
+
"grad_norm": 0.7586847543716431,
|
| 536 |
+
"learning_rate": 0.00039387755102040817,
|
| 537 |
+
"loss": 0.18,
|
| 538 |
+
"step": 3650
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.068194522079374,
|
| 542 |
+
"grad_norm": 1.3456194400787354,
|
| 543 |
+
"learning_rate": 0.00039165927240461404,
|
| 544 |
+
"loss": 0.1344,
|
| 545 |
+
"step": 3700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.096143096702068,
|
| 549 |
+
"grad_norm": 0.5425747632980347,
|
| 550 |
+
"learning_rate": 0.00038944099378881986,
|
| 551 |
+
"loss": 0.1386,
|
| 552 |
+
"step": 3750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.1240916713247624,
|
| 556 |
+
"grad_norm": 0.8643823266029358,
|
| 557 |
+
"learning_rate": 0.0003872227151730258,
|
| 558 |
+
"loss": 0.1609,
|
| 559 |
+
"step": 3800
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.1520402459474566,
|
| 563 |
+
"grad_norm": 0.6121143698692322,
|
| 564 |
+
"learning_rate": 0.0003850044365572316,
|
| 565 |
+
"loss": 0.1433,
|
| 566 |
+
"step": 3850
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.1799888205701508,
|
| 570 |
+
"grad_norm": 1.2089327573776245,
|
| 571 |
+
"learning_rate": 0.00038278615794143747,
|
| 572 |
+
"loss": 0.1635,
|
| 573 |
+
"step": 3900
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.207937395192845,
|
| 577 |
+
"grad_norm": 0.845815122127533,
|
| 578 |
+
"learning_rate": 0.0003805678793256433,
|
| 579 |
+
"loss": 0.1633,
|
| 580 |
+
"step": 3950
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.2358859698155396,
|
| 584 |
+
"grad_norm": 0.8113688230514526,
|
| 585 |
+
"learning_rate": 0.00037834960070984915,
|
| 586 |
+
"loss": 0.1497,
|
| 587 |
+
"step": 4000
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.2638345444382337,
|
| 591 |
+
"grad_norm": 1.681497573852539,
|
| 592 |
+
"learning_rate": 0.000376131322094055,
|
| 593 |
+
"loss": 0.1402,
|
| 594 |
+
"step": 4050
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.291783119060928,
|
| 598 |
+
"grad_norm": 0.9330260157585144,
|
| 599 |
+
"learning_rate": 0.0003739130434782609,
|
| 600 |
+
"loss": 0.1561,
|
| 601 |
+
"step": 4100
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.319731693683622,
|
| 605 |
+
"grad_norm": 1.3728777170181274,
|
| 606 |
+
"learning_rate": 0.0003716947648624667,
|
| 607 |
+
"loss": 0.1228,
|
| 608 |
+
"step": 4150
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.3476802683063163,
|
| 612 |
+
"grad_norm": 1.523483395576477,
|
| 613 |
+
"learning_rate": 0.0003694764862466726,
|
| 614 |
+
"loss": 0.1327,
|
| 615 |
+
"step": 4200
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.3756288429290104,
|
| 619 |
+
"grad_norm": 0.6407974362373352,
|
| 620 |
+
"learning_rate": 0.00036725820763087845,
|
| 621 |
+
"loss": 0.1499,
|
| 622 |
+
"step": 4250
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.4035774175517046,
|
| 626 |
+
"grad_norm": 0.5848202109336853,
|
| 627 |
+
"learning_rate": 0.0003650399290150843,
|
| 628 |
+
"loss": 0.1082,
|
| 629 |
+
"step": 4300
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.4315259921743992,
|
| 633 |
+
"grad_norm": 0.21722891926765442,
|
| 634 |
+
"learning_rate": 0.0003628216503992902,
|
| 635 |
+
"loss": 0.1498,
|
| 636 |
+
"step": 4350
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.4594745667970934,
|
| 640 |
+
"grad_norm": 0.7241097092628479,
|
| 641 |
+
"learning_rate": 0.000360603371783496,
|
| 642 |
+
"loss": 0.1354,
|
| 643 |
+
"step": 4400
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.4874231414197876,
|
| 647 |
+
"grad_norm": 1.2298479080200195,
|
| 648 |
+
"learning_rate": 0.00035838509316770187,
|
| 649 |
+
"loss": 0.1409,
|
| 650 |
+
"step": 4450
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.5153717160424818,
|
| 654 |
+
"grad_norm": 0.6272071003913879,
|
| 655 |
+
"learning_rate": 0.00035616681455190774,
|
| 656 |
+
"loss": 0.1358,
|
| 657 |
+
"step": 4500
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.543320290665176,
|
| 661 |
+
"grad_norm": 0.9323145151138306,
|
| 662 |
+
"learning_rate": 0.0003539485359361136,
|
| 663 |
+
"loss": 0.1316,
|
| 664 |
+
"step": 4550
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.5712688652878706,
|
| 668 |
+
"grad_norm": 0.44958174228668213,
|
| 669 |
+
"learning_rate": 0.0003517302573203194,
|
| 670 |
+
"loss": 0.1283,
|
| 671 |
+
"step": 4600
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.5992174399105643,
|
| 675 |
+
"grad_norm": 0.30582261085510254,
|
| 676 |
+
"learning_rate": 0.0003495119787045253,
|
| 677 |
+
"loss": 0.1225,
|
| 678 |
+
"step": 4650
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.627166014533259,
|
| 682 |
+
"grad_norm": 0.9854662418365479,
|
| 683 |
+
"learning_rate": 0.00034729370008873117,
|
| 684 |
+
"loss": 0.1527,
|
| 685 |
+
"step": 4700
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.655114589155953,
|
| 689 |
+
"grad_norm": 0.3222338557243347,
|
| 690 |
+
"learning_rate": 0.00034507542147293703,
|
| 691 |
+
"loss": 0.1485,
|
| 692 |
+
"step": 4750
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.6830631637786473,
|
| 696 |
+
"grad_norm": 0.7984679341316223,
|
| 697 |
+
"learning_rate": 0.00034285714285714285,
|
| 698 |
+
"loss": 0.15,
|
| 699 |
+
"step": 4800
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.7110117384013415,
|
| 703 |
+
"grad_norm": 0.27581000328063965,
|
| 704 |
+
"learning_rate": 0.0003406388642413487,
|
| 705 |
+
"loss": 0.1423,
|
| 706 |
+
"step": 4850
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.7389603130240356,
|
| 710 |
+
"grad_norm": 0.36392349004745483,
|
| 711 |
+
"learning_rate": 0.00033842058562555454,
|
| 712 |
+
"loss": 0.12,
|
| 713 |
+
"step": 4900
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.7669088876467303,
|
| 717 |
+
"grad_norm": 0.8806673884391785,
|
| 718 |
+
"learning_rate": 0.00033620230700976046,
|
| 719 |
+
"loss": 0.1353,
|
| 720 |
+
"step": 4950
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.794857462269424,
|
| 724 |
+
"grad_norm": 0.4529147446155548,
|
| 725 |
+
"learning_rate": 0.0003339840283939663,
|
| 726 |
+
"loss": 0.1524,
|
| 727 |
+
"step": 5000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.8228060368921186,
|
| 731 |
+
"grad_norm": 0.7661668658256531,
|
| 732 |
+
"learning_rate": 0.00033176574977817214,
|
| 733 |
+
"loss": 0.1431,
|
| 734 |
+
"step": 5050
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.850754611514813,
|
| 738 |
+
"grad_norm": 0.7224089503288269,
|
| 739 |
+
"learning_rate": 0.000329547471162378,
|
| 740 |
+
"loss": 0.1328,
|
| 741 |
+
"step": 5100
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.878703186137507,
|
| 745 |
+
"grad_norm": 0.7597375512123108,
|
| 746 |
+
"learning_rate": 0.00032732919254658383,
|
| 747 |
+
"loss": 0.1322,
|
| 748 |
+
"step": 5150
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.906651760760201,
|
| 752 |
+
"grad_norm": 1.273445725440979,
|
| 753 |
+
"learning_rate": 0.00032511091393078975,
|
| 754 |
+
"loss": 0.1234,
|
| 755 |
+
"step": 5200
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.9346003353828953,
|
| 759 |
+
"grad_norm": 1.428731918334961,
|
| 760 |
+
"learning_rate": 0.00032289263531499557,
|
| 761 |
+
"loss": 0.1476,
|
| 762 |
+
"step": 5250
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.96254891000559,
|
| 766 |
+
"grad_norm": 0.4086410701274872,
|
| 767 |
+
"learning_rate": 0.00032067435669920144,
|
| 768 |
+
"loss": 0.1288,
|
| 769 |
+
"step": 5300
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.9904974846282837,
|
| 773 |
+
"grad_norm": 1.4084666967391968,
|
| 774 |
+
"learning_rate": 0.00031845607808340725,
|
| 775 |
+
"loss": 0.1403,
|
| 776 |
+
"step": 5350
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 3.0,
|
| 780 |
+
"eval_f1": 0.922771152754117,
|
| 781 |
+
"eval_loss": 0.1881607472896576,
|
| 782 |
+
"eval_runtime": 15.5665,
|
| 783 |
+
"eval_samples_per_second": 235.12,
|
| 784 |
+
"eval_steps_per_second": 7.388,
|
| 785 |
+
"step": 5367
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.0184460592509783,
|
| 789 |
+
"grad_norm": 1.1583423614501953,
|
| 790 |
+
"learning_rate": 0.0003162377994676132,
|
| 791 |
+
"loss": 0.1348,
|
| 792 |
+
"step": 5400
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.0463946338736725,
|
| 796 |
+
"grad_norm": 0.5038284063339233,
|
| 797 |
+
"learning_rate": 0.000314019520851819,
|
| 798 |
+
"loss": 0.1332,
|
| 799 |
+
"step": 5450
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.0743432084963667,
|
| 803 |
+
"grad_norm": 1.139427900314331,
|
| 804 |
+
"learning_rate": 0.00031180124223602486,
|
| 805 |
+
"loss": 0.1071,
|
| 806 |
+
"step": 5500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.102291783119061,
|
| 810 |
+
"grad_norm": 0.632847011089325,
|
| 811 |
+
"learning_rate": 0.0003095829636202307,
|
| 812 |
+
"loss": 0.129,
|
| 813 |
+
"step": 5550
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.130240357741755,
|
| 817 |
+
"grad_norm": 0.9477543830871582,
|
| 818 |
+
"learning_rate": 0.00030736468500443655,
|
| 819 |
+
"loss": 0.1296,
|
| 820 |
+
"step": 5600
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.1581889323644496,
|
| 824 |
+
"grad_norm": 0.7813945412635803,
|
| 825 |
+
"learning_rate": 0.0003051464063886424,
|
| 826 |
+
"loss": 0.1356,
|
| 827 |
+
"step": 5650
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.186137506987144,
|
| 831 |
+
"grad_norm": 1.0547813177108765,
|
| 832 |
+
"learning_rate": 0.0003029281277728483,
|
| 833 |
+
"loss": 0.1223,
|
| 834 |
+
"step": 5700
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.214086081609838,
|
| 838 |
+
"grad_norm": 0.8783809542655945,
|
| 839 |
+
"learning_rate": 0.0003007098491570541,
|
| 840 |
+
"loss": 0.1086,
|
| 841 |
+
"step": 5750
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.242034656232532,
|
| 845 |
+
"grad_norm": 1.7557278871536255,
|
| 846 |
+
"learning_rate": 0.00029849157054126,
|
| 847 |
+
"loss": 0.1176,
|
| 848 |
+
"step": 5800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.2699832308552264,
|
| 852 |
+
"grad_norm": 0.8250858783721924,
|
| 853 |
+
"learning_rate": 0.0002962732919254659,
|
| 854 |
+
"loss": 0.1244,
|
| 855 |
+
"step": 5850
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.2979318054779205,
|
| 859 |
+
"grad_norm": 0.5830790996551514,
|
| 860 |
+
"learning_rate": 0.0002940550133096717,
|
| 861 |
+
"loss": 0.1255,
|
| 862 |
+
"step": 5900
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.3258803801006147,
|
| 866 |
+
"grad_norm": 0.39319124817848206,
|
| 867 |
+
"learning_rate": 0.0002918367346938776,
|
| 868 |
+
"loss": 0.1145,
|
| 869 |
+
"step": 5950
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.3538289547233093,
|
| 873 |
+
"grad_norm": 0.7558562755584717,
|
| 874 |
+
"learning_rate": 0.0002896184560780834,
|
| 875 |
+
"loss": 0.1449,
|
| 876 |
+
"step": 6000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.3817775293460035,
|
| 880 |
+
"grad_norm": 0.686841607093811,
|
| 881 |
+
"learning_rate": 0.00028740017746228927,
|
| 882 |
+
"loss": 0.1269,
|
| 883 |
+
"step": 6050
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.4097261039686977,
|
| 887 |
+
"grad_norm": 0.3600720167160034,
|
| 888 |
+
"learning_rate": 0.00028518189884649514,
|
| 889 |
+
"loss": 0.1095,
|
| 890 |
+
"step": 6100
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.437674678591392,
|
| 894 |
+
"grad_norm": 0.38374069333076477,
|
| 895 |
+
"learning_rate": 0.000282963620230701,
|
| 896 |
+
"loss": 0.11,
|
| 897 |
+
"step": 6150
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.465623253214086,
|
| 901 |
+
"grad_norm": 0.736842691898346,
|
| 902 |
+
"learning_rate": 0.0002807453416149068,
|
| 903 |
+
"loss": 0.1099,
|
| 904 |
+
"step": 6200
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.49357182783678,
|
| 908 |
+
"grad_norm": 0.21129830181598663,
|
| 909 |
+
"learning_rate": 0.0002785270629991127,
|
| 910 |
+
"loss": 0.1291,
|
| 911 |
+
"step": 6250
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.5215204024594744,
|
| 915 |
+
"grad_norm": 0.6699479222297668,
|
| 916 |
+
"learning_rate": 0.0002763087843833185,
|
| 917 |
+
"loss": 0.1018,
|
| 918 |
+
"step": 6300
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.549468977082169,
|
| 922 |
+
"grad_norm": 1.1297788619995117,
|
| 923 |
+
"learning_rate": 0.00027409050576752443,
|
| 924 |
+
"loss": 0.0986,
|
| 925 |
+
"step": 6350
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.577417551704863,
|
| 929 |
+
"grad_norm": 0.7633982300758362,
|
| 930 |
+
"learning_rate": 0.00027187222715173025,
|
| 931 |
+
"loss": 0.1249,
|
| 932 |
+
"step": 6400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.6053661263275574,
|
| 936 |
+
"grad_norm": 1.3231158256530762,
|
| 937 |
+
"learning_rate": 0.0002696539485359361,
|
| 938 |
+
"loss": 0.1178,
|
| 939 |
+
"step": 6450
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.6333147009502516,
|
| 943 |
+
"grad_norm": 0.5314428806304932,
|
| 944 |
+
"learning_rate": 0.00026743566992014193,
|
| 945 |
+
"loss": 0.1094,
|
| 946 |
+
"step": 6500
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.6612632755729457,
|
| 950 |
+
"grad_norm": 0.3701280355453491,
|
| 951 |
+
"learning_rate": 0.00026521739130434786,
|
| 952 |
+
"loss": 0.12,
|
| 953 |
+
"step": 6550
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.68921185019564,
|
| 957 |
+
"grad_norm": 0.6473478674888611,
|
| 958 |
+
"learning_rate": 0.00026299911268855367,
|
| 959 |
+
"loss": 0.106,
|
| 960 |
+
"step": 6600
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.717160424818334,
|
| 964 |
+
"grad_norm": 0.4988769590854645,
|
| 965 |
+
"learning_rate": 0.00026078083407275954,
|
| 966 |
+
"loss": 0.1396,
|
| 967 |
+
"step": 6650
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.7451089994410287,
|
| 971 |
+
"grad_norm": 1.0782082080841064,
|
| 972 |
+
"learning_rate": 0.0002585625554569654,
|
| 973 |
+
"loss": 0.1144,
|
| 974 |
+
"step": 6700
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.773057574063723,
|
| 978 |
+
"grad_norm": 0.846603274345398,
|
| 979 |
+
"learning_rate": 0.00025634427684117123,
|
| 980 |
+
"loss": 0.1142,
|
| 981 |
+
"step": 6750
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 3.801006148686417,
|
| 985 |
+
"grad_norm": 1.2522852420806885,
|
| 986 |
+
"learning_rate": 0.00025412599822537715,
|
| 987 |
+
"loss": 0.1086,
|
| 988 |
+
"step": 6800
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 3.8289547233091112,
|
| 992 |
+
"grad_norm": 0.30715683102607727,
|
| 993 |
+
"learning_rate": 0.00025190771960958297,
|
| 994 |
+
"loss": 0.1023,
|
| 995 |
+
"step": 6850
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.8569032979318054,
|
| 999 |
+
"grad_norm": 1.1746628284454346,
|
| 1000 |
+
"learning_rate": 0.00024968944099378884,
|
| 1001 |
+
"loss": 0.1182,
|
| 1002 |
+
"step": 6900
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 3.8848518725544996,
|
| 1006 |
+
"grad_norm": 2.2346272468566895,
|
| 1007 |
+
"learning_rate": 0.0002474711623779947,
|
| 1008 |
+
"loss": 0.1207,
|
| 1009 |
+
"step": 6950
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 3.9128004471771938,
|
| 1013 |
+
"grad_norm": 1.6879889965057373,
|
| 1014 |
+
"learning_rate": 0.0002452528837622005,
|
| 1015 |
+
"loss": 0.101,
|
| 1016 |
+
"step": 7000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.9407490217998884,
|
| 1020 |
+
"grad_norm": 0.777055025100708,
|
| 1021 |
+
"learning_rate": 0.00024303460514640642,
|
| 1022 |
+
"loss": 0.1051,
|
| 1023 |
+
"step": 7050
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 3.9686975964225826,
|
| 1027 |
+
"grad_norm": 1.065346598625183,
|
| 1028 |
+
"learning_rate": 0.00024081632653061226,
|
| 1029 |
+
"loss": 0.1055,
|
| 1030 |
+
"step": 7100
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 3.9966461710452768,
|
| 1034 |
+
"grad_norm": 2.3775737285614014,
|
| 1035 |
+
"learning_rate": 0.0002385980479148181,
|
| 1036 |
+
"loss": 0.097,
|
| 1037 |
+
"step": 7150
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 4.0,
|
| 1041 |
+
"eval_f1": 0.9374824782730586,
|
| 1042 |
+
"eval_loss": 0.15188030898571014,
|
| 1043 |
+
"eval_runtime": 15.6619,
|
| 1044 |
+
"eval_samples_per_second": 233.688,
|
| 1045 |
+
"eval_steps_per_second": 7.343,
|
| 1046 |
+
"step": 7156
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 4.024594745667971,
|
| 1050 |
+
"grad_norm": 0.8017207980155945,
|
| 1051 |
+
"learning_rate": 0.00023637976929902397,
|
| 1052 |
+
"loss": 0.1102,
|
| 1053 |
+
"step": 7200
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 4.052543320290665,
|
| 1057 |
+
"grad_norm": 2.158339023590088,
|
| 1058 |
+
"learning_rate": 0.00023416149068322982,
|
| 1059 |
+
"loss": 0.1033,
|
| 1060 |
+
"step": 7250
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 4.08049189491336,
|
| 1064 |
+
"grad_norm": 0.8033800721168518,
|
| 1065 |
+
"learning_rate": 0.00023194321206743569,
|
| 1066 |
+
"loss": 0.1054,
|
| 1067 |
+
"step": 7300
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.1084404695360535,
|
| 1071 |
+
"grad_norm": 1.1149848699569702,
|
| 1072 |
+
"learning_rate": 0.00022972493345164153,
|
| 1073 |
+
"loss": 0.0884,
|
| 1074 |
+
"step": 7350
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.136389044158748,
|
| 1078 |
+
"grad_norm": 0.18090881407260895,
|
| 1079 |
+
"learning_rate": 0.0002275066548358474,
|
| 1080 |
+
"loss": 0.091,
|
| 1081 |
+
"step": 7400
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.164337618781442,
|
| 1085 |
+
"grad_norm": 0.6261000633239746,
|
| 1086 |
+
"learning_rate": 0.00022528837622005324,
|
| 1087 |
+
"loss": 0.0842,
|
| 1088 |
+
"step": 7450
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 4.192286193404136,
|
| 1092 |
+
"grad_norm": 0.44622424244880676,
|
| 1093 |
+
"learning_rate": 0.00022307009760425908,
|
| 1094 |
+
"loss": 0.0933,
|
| 1095 |
+
"step": 7500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 4.220234768026831,
|
| 1099 |
+
"grad_norm": 1.2650983333587646,
|
| 1100 |
+
"learning_rate": 0.00022085181898846495,
|
| 1101 |
+
"loss": 0.0982,
|
| 1102 |
+
"step": 7550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.248183342649525,
|
| 1106 |
+
"grad_norm": 2.512014150619507,
|
| 1107 |
+
"learning_rate": 0.0002186335403726708,
|
| 1108 |
+
"loss": 0.0959,
|
| 1109 |
+
"step": 7600
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 4.276131917272219,
|
| 1113 |
+
"grad_norm": 0.9572253227233887,
|
| 1114 |
+
"learning_rate": 0.00021641526175687667,
|
| 1115 |
+
"loss": 0.1197,
|
| 1116 |
+
"step": 7650
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 4.304080491894913,
|
| 1120 |
+
"grad_norm": 0.8612358570098877,
|
| 1121 |
+
"learning_rate": 0.0002141969831410825,
|
| 1122 |
+
"loss": 0.0898,
|
| 1123 |
+
"step": 7700
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 4.332029066517608,
|
| 1127 |
+
"grad_norm": 0.13563434779644012,
|
| 1128 |
+
"learning_rate": 0.00021197870452528838,
|
| 1129 |
+
"loss": 0.1268,
|
| 1130 |
+
"step": 7750
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 4.3599776411403015,
|
| 1134 |
+
"grad_norm": 0.9207750558853149,
|
| 1135 |
+
"learning_rate": 0.00020976042590949425,
|
| 1136 |
+
"loss": 0.1138,
|
| 1137 |
+
"step": 7800
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.387926215762996,
|
| 1141 |
+
"grad_norm": 0.15072673559188843,
|
| 1142 |
+
"learning_rate": 0.00020754214729370012,
|
| 1143 |
+
"loss": 0.0854,
|
| 1144 |
+
"step": 7850
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 4.41587479038569,
|
| 1148 |
+
"grad_norm": 0.5902436375617981,
|
| 1149 |
+
"learning_rate": 0.00020532386867790596,
|
| 1150 |
+
"loss": 0.0967,
|
| 1151 |
+
"step": 7900
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 4.4438233650083845,
|
| 1155 |
+
"grad_norm": 2.2286317348480225,
|
| 1156 |
+
"learning_rate": 0.0002031055900621118,
|
| 1157 |
+
"loss": 0.0939,
|
| 1158 |
+
"step": 7950
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 4.471771939631079,
|
| 1162 |
+
"grad_norm": 0.6109320521354675,
|
| 1163 |
+
"learning_rate": 0.00020088731144631767,
|
| 1164 |
+
"loss": 0.0887,
|
| 1165 |
+
"step": 8000
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 4.499720514253773,
|
| 1169 |
+
"grad_norm": 1.340773582458496,
|
| 1170 |
+
"learning_rate": 0.00019866903283052352,
|
| 1171 |
+
"loss": 0.0837,
|
| 1172 |
+
"step": 8050
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.5276690888764675,
|
| 1176 |
+
"grad_norm": 1.1809900999069214,
|
| 1177 |
+
"learning_rate": 0.00019645075421472938,
|
| 1178 |
+
"loss": 0.0765,
|
| 1179 |
+
"step": 8100
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 4.555617663499161,
|
| 1183 |
+
"grad_norm": 0.8299195170402527,
|
| 1184 |
+
"learning_rate": 0.00019423247559893523,
|
| 1185 |
+
"loss": 0.0972,
|
| 1186 |
+
"step": 8150
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 4.583566238121856,
|
| 1190 |
+
"grad_norm": 0.18145732581615448,
|
| 1191 |
+
"learning_rate": 0.0001920141969831411,
|
| 1192 |
+
"loss": 0.075,
|
| 1193 |
+
"step": 8200
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 4.61151481274455,
|
| 1197 |
+
"grad_norm": 0.2446570098400116,
|
| 1198 |
+
"learning_rate": 0.00018979591836734694,
|
| 1199 |
+
"loss": 0.0857,
|
| 1200 |
+
"step": 8250
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 4.639463387367244,
|
| 1204 |
+
"grad_norm": 1.534759759902954,
|
| 1205 |
+
"learning_rate": 0.00018757763975155278,
|
| 1206 |
+
"loss": 0.0793,
|
| 1207 |
+
"step": 8300
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 4.667411961989939,
|
| 1211 |
+
"grad_norm": 1.6062694787979126,
|
| 1212 |
+
"learning_rate": 0.00018535936113575865,
|
| 1213 |
+
"loss": 0.0931,
|
| 1214 |
+
"step": 8350
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 4.6953605366126325,
|
| 1218 |
+
"grad_norm": 2.1251513957977295,
|
| 1219 |
+
"learning_rate": 0.0001831410825199645,
|
| 1220 |
+
"loss": 0.1003,
|
| 1221 |
+
"step": 8400
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 4.723309111235327,
|
| 1225 |
+
"grad_norm": 0.9858797788619995,
|
| 1226 |
+
"learning_rate": 0.00018092280390417036,
|
| 1227 |
+
"loss": 0.082,
|
| 1228 |
+
"step": 8450
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 4.751257685858021,
|
| 1232 |
+
"grad_norm": 1.1654304265975952,
|
| 1233 |
+
"learning_rate": 0.0001787045252883762,
|
| 1234 |
+
"loss": 0.0603,
|
| 1235 |
+
"step": 8500
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 4.7792062604807155,
|
| 1239 |
+
"grad_norm": 1.731586217880249,
|
| 1240 |
+
"learning_rate": 0.00017648624667258208,
|
| 1241 |
+
"loss": 0.0593,
|
| 1242 |
+
"step": 8550
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 4.807154835103409,
|
| 1246 |
+
"grad_norm": 0.524991512298584,
|
| 1247 |
+
"learning_rate": 0.00017426796805678795,
|
| 1248 |
+
"loss": 0.0928,
|
| 1249 |
+
"step": 8600
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 4.835103409726104,
|
| 1253 |
+
"grad_norm": 0.76012122631073,
|
| 1254 |
+
"learning_rate": 0.00017204968944099382,
|
| 1255 |
+
"loss": 0.0901,
|
| 1256 |
+
"step": 8650
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.8630519843487985,
|
| 1260 |
+
"grad_norm": 1.038218379020691,
|
| 1261 |
+
"learning_rate": 0.00016983141082519966,
|
| 1262 |
+
"loss": 0.0712,
|
| 1263 |
+
"step": 8700
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 4.891000558971492,
|
| 1267 |
+
"grad_norm": 0.7694293260574341,
|
| 1268 |
+
"learning_rate": 0.0001676131322094055,
|
| 1269 |
+
"loss": 0.0801,
|
| 1270 |
+
"step": 8750
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 4.918949133594187,
|
| 1274 |
+
"grad_norm": 1.4149705171585083,
|
| 1275 |
+
"learning_rate": 0.00016539485359361137,
|
| 1276 |
+
"loss": 0.1054,
|
| 1277 |
+
"step": 8800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 4.946897708216881,
|
| 1281 |
+
"grad_norm": 0.35135123133659363,
|
| 1282 |
+
"learning_rate": 0.00016317657497781721,
|
| 1283 |
+
"loss": 0.0917,
|
| 1284 |
+
"step": 8850
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 4.974846282839575,
|
| 1288 |
+
"grad_norm": 1.8134843111038208,
|
| 1289 |
+
"learning_rate": 0.00016095829636202308,
|
| 1290 |
+
"loss": 0.0946,
|
| 1291 |
+
"step": 8900
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 5.0,
|
| 1295 |
+
"eval_f1": 0.9494430162810625,
|
| 1296 |
+
"eval_loss": 0.12213503569364548,
|
| 1297 |
+
"eval_runtime": 16.0206,
|
| 1298 |
+
"eval_samples_per_second": 228.456,
|
| 1299 |
+
"eval_steps_per_second": 7.178,
|
| 1300 |
+
"step": 8945
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 5.00279485746227,
|
| 1304 |
+
"grad_norm": 0.8835757374763489,
|
| 1305 |
+
"learning_rate": 0.00015874001774622893,
|
| 1306 |
+
"loss": 0.0792,
|
| 1307 |
+
"step": 8950
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 5.0307434320849636,
|
| 1311 |
+
"grad_norm": 0.5846070647239685,
|
| 1312 |
+
"learning_rate": 0.0001565217391304348,
|
| 1313 |
+
"loss": 0.0834,
|
| 1314 |
+
"step": 9000
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 5.058692006707658,
|
| 1318 |
+
"grad_norm": 0.4132053256034851,
|
| 1319 |
+
"learning_rate": 0.00015430346051464064,
|
| 1320 |
+
"loss": 0.0622,
|
| 1321 |
+
"step": 9050
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 5.086640581330352,
|
| 1325 |
+
"grad_norm": 3.0904829502105713,
|
| 1326 |
+
"learning_rate": 0.00015208518189884648,
|
| 1327 |
+
"loss": 0.0626,
|
| 1328 |
+
"step": 9100
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 5.1145891559530465,
|
| 1332 |
+
"grad_norm": 2.243579387664795,
|
| 1333 |
+
"learning_rate": 0.00014986690328305235,
|
| 1334 |
+
"loss": 0.0795,
|
| 1335 |
+
"step": 9150
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 5.14253773057574,
|
| 1339 |
+
"grad_norm": 1.4317448139190674,
|
| 1340 |
+
"learning_rate": 0.0001476486246672582,
|
| 1341 |
+
"loss": 0.0642,
|
| 1342 |
+
"step": 9200
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 5.170486305198435,
|
| 1346 |
+
"grad_norm": 0.3954031765460968,
|
| 1347 |
+
"learning_rate": 0.00014543034605146406,
|
| 1348 |
+
"loss": 0.0748,
|
| 1349 |
+
"step": 9250
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 5.1984348798211295,
|
| 1353 |
+
"grad_norm": 0.7129592895507812,
|
| 1354 |
+
"learning_rate": 0.0001432120674356699,
|
| 1355 |
+
"loss": 0.0631,
|
| 1356 |
+
"step": 9300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 5.226383454443823,
|
| 1360 |
+
"grad_norm": 0.7119801044464111,
|
| 1361 |
+
"learning_rate": 0.0001409937888198758,
|
| 1362 |
+
"loss": 0.0672,
|
| 1363 |
+
"step": 9350
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 5.254332029066518,
|
| 1367 |
+
"grad_norm": 1.8398321866989136,
|
| 1368 |
+
"learning_rate": 0.00013877551020408165,
|
| 1369 |
+
"loss": 0.0627,
|
| 1370 |
+
"step": 9400
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 5.282280603689212,
|
| 1374 |
+
"grad_norm": 1.1668697595596313,
|
| 1375 |
+
"learning_rate": 0.0001365572315882875,
|
| 1376 |
+
"loss": 0.0661,
|
| 1377 |
+
"step": 9450
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 5.310229178311906,
|
| 1381 |
+
"grad_norm": 0.6019765734672546,
|
| 1382 |
+
"learning_rate": 0.00013433895297249336,
|
| 1383 |
+
"loss": 0.0615,
|
| 1384 |
+
"step": 9500
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 5.3381777529346,
|
| 1388 |
+
"grad_norm": 2.365870237350464,
|
| 1389 |
+
"learning_rate": 0.0001321206743566992,
|
| 1390 |
+
"loss": 0.0582,
|
| 1391 |
+
"step": 9550
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 5.366126327557295,
|
| 1395 |
+
"grad_norm": 0.10710655897855759,
|
| 1396 |
+
"learning_rate": 0.00012990239574090507,
|
| 1397 |
+
"loss": 0.052,
|
| 1398 |
+
"step": 9600
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 5.394074902179989,
|
| 1402 |
+
"grad_norm": 0.6959599852561951,
|
| 1403 |
+
"learning_rate": 0.0001276841171251109,
|
| 1404 |
+
"loss": 0.0558,
|
| 1405 |
+
"step": 9650
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 5.422023476802683,
|
| 1409 |
+
"grad_norm": 0.11615774035453796,
|
| 1410 |
+
"learning_rate": 0.00012546583850931678,
|
| 1411 |
+
"loss": 0.0625,
|
| 1412 |
+
"step": 9700
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 5.4499720514253776,
|
| 1416 |
+
"grad_norm": 0.057329438626766205,
|
| 1417 |
+
"learning_rate": 0.00012324755989352263,
|
| 1418 |
+
"loss": 0.0455,
|
| 1419 |
+
"step": 9750
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 5.477920626048071,
|
| 1423 |
+
"grad_norm": 3.285327196121216,
|
| 1424 |
+
"learning_rate": 0.00012102928127772848,
|
| 1425 |
+
"loss": 0.0674,
|
| 1426 |
+
"step": 9800
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 5.505869200670766,
|
| 1430 |
+
"grad_norm": 2.17206072807312,
|
| 1431 |
+
"learning_rate": 0.00011881100266193434,
|
| 1432 |
+
"loss": 0.0678,
|
| 1433 |
+
"step": 9850
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 5.5338177752934605,
|
| 1437 |
+
"grad_norm": 0.2885390520095825,
|
| 1438 |
+
"learning_rate": 0.0001165927240461402,
|
| 1439 |
+
"loss": 0.0512,
|
| 1440 |
+
"step": 9900
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 5.561766349916154,
|
| 1444 |
+
"grad_norm": 2.1970298290252686,
|
| 1445 |
+
"learning_rate": 0.00011437444543034606,
|
| 1446 |
+
"loss": 0.0598,
|
| 1447 |
+
"step": 9950
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 5.589714924538849,
|
| 1451 |
+
"grad_norm": 0.1274396926164627,
|
| 1452 |
+
"learning_rate": 0.00011215616681455192,
|
| 1453 |
+
"loss": 0.0717,
|
| 1454 |
+
"step": 10000
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 5.617663499161543,
|
| 1458 |
+
"grad_norm": 0.7299688458442688,
|
| 1459 |
+
"learning_rate": 0.00010993788819875776,
|
| 1460 |
+
"loss": 0.0797,
|
| 1461 |
+
"step": 10050
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 5.645612073784237,
|
| 1465 |
+
"grad_norm": 0.5048921704292297,
|
| 1466 |
+
"learning_rate": 0.00010771960958296362,
|
| 1467 |
+
"loss": 0.0592,
|
| 1468 |
+
"step": 10100
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 5.673560648406931,
|
| 1472 |
+
"grad_norm": 1.5433499813079834,
|
| 1473 |
+
"learning_rate": 0.00010550133096716947,
|
| 1474 |
+
"loss": 0.0702,
|
| 1475 |
+
"step": 10150
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 5.701509223029626,
|
| 1479 |
+
"grad_norm": 0.23040899634361267,
|
| 1480 |
+
"learning_rate": 0.00010328305235137533,
|
| 1481 |
+
"loss": 0.0654,
|
| 1482 |
+
"step": 10200
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 5.729457797652319,
|
| 1486 |
+
"grad_norm": 1.693513035774231,
|
| 1487 |
+
"learning_rate": 0.00010106477373558119,
|
| 1488 |
+
"loss": 0.0687,
|
| 1489 |
+
"step": 10250
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 5.757406372275014,
|
| 1493 |
+
"grad_norm": 4.076319694519043,
|
| 1494 |
+
"learning_rate": 9.884649511978706e-05,
|
| 1495 |
+
"loss": 0.0639,
|
| 1496 |
+
"step": 10300
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 5.785354946897709,
|
| 1500 |
+
"grad_norm": 0.48212429881095886,
|
| 1501 |
+
"learning_rate": 9.662821650399291e-05,
|
| 1502 |
+
"loss": 0.0604,
|
| 1503 |
+
"step": 10350
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 5.813303521520402,
|
| 1507 |
+
"grad_norm": 0.04306049272418022,
|
| 1508 |
+
"learning_rate": 9.440993788819877e-05,
|
| 1509 |
+
"loss": 0.0495,
|
| 1510 |
+
"step": 10400
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 5.841252096143097,
|
| 1514 |
+
"grad_norm": 0.23358555138111115,
|
| 1515 |
+
"learning_rate": 9.223602484472051e-05,
|
| 1516 |
+
"loss": 0.0615,
|
| 1517 |
+
"step": 10450
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 5.869200670765791,
|
| 1521 |
+
"grad_norm": 1.8047716617584229,
|
| 1522 |
+
"learning_rate": 9.001774622892636e-05,
|
| 1523 |
+
"loss": 0.0531,
|
| 1524 |
+
"step": 10500
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 5.897149245388485,
|
| 1528 |
+
"grad_norm": 0.1505836844444275,
|
| 1529 |
+
"learning_rate": 8.779946761313221e-05,
|
| 1530 |
+
"loss": 0.0529,
|
| 1531 |
+
"step": 10550
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 5.92509782001118,
|
| 1535 |
+
"grad_norm": 0.42754456400871277,
|
| 1536 |
+
"learning_rate": 8.558118899733806e-05,
|
| 1537 |
+
"loss": 0.0567,
|
| 1538 |
+
"step": 10600
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 5.953046394633874,
|
| 1542 |
+
"grad_norm": 1.9926128387451172,
|
| 1543 |
+
"learning_rate": 8.336291038154392e-05,
|
| 1544 |
+
"loss": 0.0473,
|
| 1545 |
+
"step": 10650
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 5.980994969256568,
|
| 1549 |
+
"grad_norm": 0.3347943425178528,
|
| 1550 |
+
"learning_rate": 8.114463176574978e-05,
|
| 1551 |
+
"loss": 0.0688,
|
| 1552 |
+
"step": 10700
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 6.0,
|
| 1556 |
+
"eval_f1": 0.9666100735710244,
|
| 1557 |
+
"eval_loss": 0.08690152317285538,
|
| 1558 |
+
"eval_runtime": 15.4749,
|
| 1559 |
+
"eval_samples_per_second": 236.512,
|
| 1560 |
+
"eval_steps_per_second": 7.431,
|
| 1561 |
+
"step": 10734
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 6.008943543879262,
|
| 1565 |
+
"grad_norm": 0.8442544341087341,
|
| 1566 |
+
"learning_rate": 7.892635314995563e-05,
|
| 1567 |
+
"loss": 0.0456,
|
| 1568 |
+
"step": 10750
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 6.036892118501957,
|
| 1572 |
+
"grad_norm": 1.1793649196624756,
|
| 1573 |
+
"learning_rate": 7.67080745341615e-05,
|
| 1574 |
+
"loss": 0.0434,
|
| 1575 |
+
"step": 10800
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 6.06484069312465,
|
| 1579 |
+
"grad_norm": 0.9821094870567322,
|
| 1580 |
+
"learning_rate": 7.448979591836736e-05,
|
| 1581 |
+
"loss": 0.0455,
|
| 1582 |
+
"step": 10850
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 6.092789267747345,
|
| 1586 |
+
"grad_norm": 0.5856317281723022,
|
| 1587 |
+
"learning_rate": 7.227151730257321e-05,
|
| 1588 |
+
"loss": 0.0365,
|
| 1589 |
+
"step": 10900
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 6.120737842370039,
|
| 1593 |
+
"grad_norm": 0.5461094379425049,
|
| 1594 |
+
"learning_rate": 7.005323868677906e-05,
|
| 1595 |
+
"loss": 0.0342,
|
| 1596 |
+
"step": 10950
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 6.148686416992733,
|
| 1600 |
+
"grad_norm": 2.0931904315948486,
|
| 1601 |
+
"learning_rate": 6.783496007098491e-05,
|
| 1602 |
+
"loss": 0.0423,
|
| 1603 |
+
"step": 11000
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 6.176634991615428,
|
| 1607 |
+
"grad_norm": 0.046953145414590836,
|
| 1608 |
+
"learning_rate": 6.561668145519077e-05,
|
| 1609 |
+
"loss": 0.0523,
|
| 1610 |
+
"step": 11050
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 6.204583566238122,
|
| 1614 |
+
"grad_norm": 1.0641871690750122,
|
| 1615 |
+
"learning_rate": 6.339840283939662e-05,
|
| 1616 |
+
"loss": 0.0421,
|
| 1617 |
+
"step": 11100
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 6.232532140860816,
|
| 1621 |
+
"grad_norm": 0.4122409224510193,
|
| 1622 |
+
"learning_rate": 6.118012422360248e-05,
|
| 1623 |
+
"loss": 0.0441,
|
| 1624 |
+
"step": 11150
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 6.26048071548351,
|
| 1628 |
+
"grad_norm": 0.2935519218444824,
|
| 1629 |
+
"learning_rate": 5.8961845607808344e-05,
|
| 1630 |
+
"loss": 0.0354,
|
| 1631 |
+
"step": 11200
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 6.288429290106205,
|
| 1635 |
+
"grad_norm": 0.8135849833488464,
|
| 1636 |
+
"learning_rate": 5.6743566992014193e-05,
|
| 1637 |
+
"loss": 0.0333,
|
| 1638 |
+
"step": 11250
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 6.316377864728899,
|
| 1642 |
+
"grad_norm": 1.1806564331054688,
|
| 1643 |
+
"learning_rate": 5.4525288376220056e-05,
|
| 1644 |
+
"loss": 0.0614,
|
| 1645 |
+
"step": 11300
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 6.344326439351593,
|
| 1649 |
+
"grad_norm": 0.5362429022789001,
|
| 1650 |
+
"learning_rate": 5.230700976042591e-05,
|
| 1651 |
+
"loss": 0.0453,
|
| 1652 |
+
"step": 11350
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 6.372275013974288,
|
| 1656 |
+
"grad_norm": 0.10853467881679535,
|
| 1657 |
+
"learning_rate": 5.008873114463176e-05,
|
| 1658 |
+
"loss": 0.0371,
|
| 1659 |
+
"step": 11400
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 6.400223588596981,
|
| 1663 |
+
"grad_norm": 2.1878767013549805,
|
| 1664 |
+
"learning_rate": 4.787045252883762e-05,
|
| 1665 |
+
"loss": 0.0418,
|
| 1666 |
+
"step": 11450
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 6.428172163219676,
|
| 1670 |
+
"grad_norm": 1.1209208965301514,
|
| 1671 |
+
"learning_rate": 4.565217391304348e-05,
|
| 1672 |
+
"loss": 0.0257,
|
| 1673 |
+
"step": 11500
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 6.45612073784237,
|
| 1677 |
+
"grad_norm": 1.4330588579177856,
|
| 1678 |
+
"learning_rate": 4.343389529724934e-05,
|
| 1679 |
+
"loss": 0.0595,
|
| 1680 |
+
"step": 11550
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 6.484069312465064,
|
| 1684 |
+
"grad_norm": 1.9309511184692383,
|
| 1685 |
+
"learning_rate": 4.1215616681455187e-05,
|
| 1686 |
+
"loss": 0.0357,
|
| 1687 |
+
"step": 11600
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 6.512017887087758,
|
| 1691 |
+
"grad_norm": 0.004909028764814138,
|
| 1692 |
+
"learning_rate": 3.899733806566105e-05,
|
| 1693 |
+
"loss": 0.0332,
|
| 1694 |
+
"step": 11650
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 6.539966461710453,
|
| 1698 |
+
"grad_norm": 0.5627462267875671,
|
| 1699 |
+
"learning_rate": 3.6779059449866906e-05,
|
| 1700 |
+
"loss": 0.0416,
|
| 1701 |
+
"step": 11700
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 6.567915036333147,
|
| 1705 |
+
"grad_norm": 4.456967830657959,
|
| 1706 |
+
"learning_rate": 3.456078083407276e-05,
|
| 1707 |
+
"loss": 0.045,
|
| 1708 |
+
"step": 11750
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 6.595863610955841,
|
| 1712 |
+
"grad_norm": 2.8317513465881348,
|
| 1713 |
+
"learning_rate": 3.234250221827861e-05,
|
| 1714 |
+
"loss": 0.0305,
|
| 1715 |
+
"step": 11800
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 6.623812185578536,
|
| 1719 |
+
"grad_norm": 2.5559463500976562,
|
| 1720 |
+
"learning_rate": 3.012422360248447e-05,
|
| 1721 |
+
"loss": 0.0297,
|
| 1722 |
+
"step": 11850
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 6.651760760201229,
|
| 1726 |
+
"grad_norm": 0.5523830056190491,
|
| 1727 |
+
"learning_rate": 2.790594498669033e-05,
|
| 1728 |
+
"loss": 0.0325,
|
| 1729 |
+
"step": 11900
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 6.679709334823924,
|
| 1733 |
+
"grad_norm": 0.04988823086023331,
|
| 1734 |
+
"learning_rate": 2.5687666370896183e-05,
|
| 1735 |
+
"loss": 0.0313,
|
| 1736 |
+
"step": 11950
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 6.707657909446619,
|
| 1740 |
+
"grad_norm": 1.5902944803237915,
|
| 1741 |
+
"learning_rate": 2.3469387755102043e-05,
|
| 1742 |
+
"loss": 0.0369,
|
| 1743 |
+
"step": 12000
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 6.735606484069312,
|
| 1747 |
+
"grad_norm": 1.3617613315582275,
|
| 1748 |
+
"learning_rate": 2.1251109139307896e-05,
|
| 1749 |
+
"loss": 0.0338,
|
| 1750 |
+
"step": 12050
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 6.763555058692007,
|
| 1754 |
+
"grad_norm": 0.13116395473480225,
|
| 1755 |
+
"learning_rate": 1.9032830523513755e-05,
|
| 1756 |
+
"loss": 0.0254,
|
| 1757 |
+
"step": 12100
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 6.791503633314701,
|
| 1761 |
+
"grad_norm": 0.7528799176216125,
|
| 1762 |
+
"learning_rate": 1.6814551907719608e-05,
|
| 1763 |
+
"loss": 0.0322,
|
| 1764 |
+
"step": 12150
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 6.819452207937395,
|
| 1768 |
+
"grad_norm": 1.894806146621704,
|
| 1769 |
+
"learning_rate": 1.4596273291925467e-05,
|
| 1770 |
+
"loss": 0.0254,
|
| 1771 |
+
"step": 12200
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 6.847400782560089,
|
| 1775 |
+
"grad_norm": 0.5565615892410278,
|
| 1776 |
+
"learning_rate": 1.2377994676131324e-05,
|
| 1777 |
+
"loss": 0.0284,
|
| 1778 |
+
"step": 12250
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 6.875349357182784,
|
| 1782 |
+
"grad_norm": 0.45591992139816284,
|
| 1783 |
+
"learning_rate": 1.015971606033718e-05,
|
| 1784 |
+
"loss": 0.0356,
|
| 1785 |
+
"step": 12300
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 6.9032979318054775,
|
| 1789 |
+
"grad_norm": 1.1384029388427734,
|
| 1790 |
+
"learning_rate": 7.941437444543036e-06,
|
| 1791 |
+
"loss": 0.0308,
|
| 1792 |
+
"step": 12350
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 6.931246506428172,
|
| 1796 |
+
"grad_norm": 0.24669720232486725,
|
| 1797 |
+
"learning_rate": 5.7231588287488905e-06,
|
| 1798 |
+
"loss": 0.0385,
|
| 1799 |
+
"step": 12400
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 6.959195081050867,
|
| 1803 |
+
"grad_norm": 2.1776230335235596,
|
| 1804 |
+
"learning_rate": 3.5048802129547475e-06,
|
| 1805 |
+
"loss": 0.0271,
|
| 1806 |
+
"step": 12450
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 6.98714365567356,
|
| 1810 |
+
"grad_norm": 0.8091051578521729,
|
| 1811 |
+
"learning_rate": 1.2866015971606034e-06,
|
| 1812 |
+
"loss": 0.0303,
|
| 1813 |
+
"step": 12500
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 7.0,
|
| 1817 |
+
"eval_f1": 0.9760225669957687,
|
| 1818 |
+
"eval_loss": 0.06743501126766205,
|
| 1819 |
+
"eval_runtime": 16.4262,
|
| 1820 |
+
"eval_samples_per_second": 222.815,
|
| 1821 |
+
"eval_steps_per_second": 7.001,
|
| 1822 |
+
"step": 12523
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.0,
|
| 1826 |
+
"step": 12523,
|
| 1827 |
+
"total_flos": 2.1045216318570566e+19,
|
| 1828 |
+
"train_loss": 0.12371139868869004,
|
| 1829 |
+
"train_runtime": 2341.9199,
|
| 1830 |
+
"train_samples_per_second": 171.111,
|
| 1831 |
+
"train_steps_per_second": 5.347
|
| 1832 |
+
}
|
| 1833 |
+
],
|
| 1834 |
+
"logging_steps": 50,
|
| 1835 |
+
"max_steps": 12523,
|
| 1836 |
+
"num_input_tokens_seen": 0,
|
| 1837 |
+
"num_train_epochs": 7,
|
| 1838 |
+
"save_steps": 500,
|
| 1839 |
+
"stateful_callbacks": {
|
| 1840 |
+
"TrainerControl": {
|
| 1841 |
+
"args": {
|
| 1842 |
+
"should_epoch_stop": false,
|
| 1843 |
+
"should_evaluate": false,
|
| 1844 |
+
"should_log": false,
|
| 1845 |
+
"should_save": true,
|
| 1846 |
+
"should_training_stop": true
|
| 1847 |
+
},
|
| 1848 |
+
"attributes": {}
|
| 1849 |
+
}
|
| 1850 |
+
},
|
| 1851 |
+
"total_flos": 2.1045216318570566e+19,
|
| 1852 |
+
"train_batch_size": 32,
|
| 1853 |
+
"trial_name": null,
|
| 1854 |
+
"trial_params": null
|
| 1855 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:018b1cc79f445b7fda77d6931acd6c935bbc6899a188e5565257443a139ef323
|
| 3 |
+
size 5304
|