Upload 8 files
Browse files- config.json +53 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +22 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2067 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/resnet-50",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ResNetForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"depths": [
|
| 7 |
+
3,
|
| 8 |
+
4,
|
| 9 |
+
6,
|
| 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": "COVID",
|
| 24 |
+
"1": "Lung opacity",
|
| 25 |
+
"2": "Normal",
|
| 26 |
+
"3": "Viral Pneumonia"
|
| 27 |
+
},
|
| 28 |
+
"label2id": {
|
| 29 |
+
"COVID": 0,
|
| 30 |
+
"Lung opacity": 1,
|
| 31 |
+
"Normal": 2,
|
| 32 |
+
"Viral Pneumonia": 3
|
| 33 |
+
},
|
| 34 |
+
"layer_type": "bottleneck",
|
| 35 |
+
"model_type": "resnet",
|
| 36 |
+
"num_channels": 3,
|
| 37 |
+
"out_features": [
|
| 38 |
+
"stage4"
|
| 39 |
+
],
|
| 40 |
+
"out_indices": [
|
| 41 |
+
4
|
| 42 |
+
],
|
| 43 |
+
"problem_type": "single_label_classification",
|
| 44 |
+
"stage_names": [
|
| 45 |
+
"stem",
|
| 46 |
+
"stage1",
|
| 47 |
+
"stage2",
|
| 48 |
+
"stage3",
|
| 49 |
+
"stage4"
|
| 50 |
+
],
|
| 51 |
+
"torch_dtype": "float32",
|
| 52 |
+
"transformers_version": "4.42.4"
|
| 53 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a519fbb0a3e371645e7b2de81a25851fd1a5431d5b8f8b8b8a3500cf7009c5a8
|
| 3 |
+
size 94319344
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59c08134eddb0047e92fc278b4d07329a2990e13ce2b942eff1fc081f4a2f59c
|
| 3 |
+
size 188265274
|
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 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37f947d6b54eeedc56f985f58c76b5403af723450de37052dd484bc469768e1b
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d93b3b4529555907ae84565fa978f9e641a53e6e4a74c902aca536002ae407dc
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2067 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9375,
|
| 3 |
+
"best_model_checkpoint": "./resnet50/checkpoint-2436",
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2520,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11904761904761904,
|
| 13 |
+
"grad_norm": 0.48410382866859436,
|
| 14 |
+
"learning_rate": 7.936507936507937e-07,
|
| 15 |
+
"loss": 1.3866,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.23809523809523808,
|
| 20 |
+
"grad_norm": 0.4765987694263458,
|
| 21 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 22 |
+
"loss": 1.3861,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.35714285714285715,
|
| 27 |
+
"grad_norm": 0.43654772639274597,
|
| 28 |
+
"learning_rate": 2.380952380952381e-06,
|
| 29 |
+
"loss": 1.3859,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.47619047619047616,
|
| 34 |
+
"grad_norm": 0.5733897686004639,
|
| 35 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 36 |
+
"loss": 1.3855,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5952380952380952,
|
| 41 |
+
"grad_norm": 0.457210898399353,
|
| 42 |
+
"learning_rate": 3.968253968253968e-06,
|
| 43 |
+
"loss": 1.385,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.7142857142857143,
|
| 48 |
+
"grad_norm": 0.5107349157333374,
|
| 49 |
+
"learning_rate": 4.761904761904762e-06,
|
| 50 |
+
"loss": 1.3843,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8333333333333334,
|
| 55 |
+
"grad_norm": 0.6200265288352966,
|
| 56 |
+
"learning_rate": 5.555555555555557e-06,
|
| 57 |
+
"loss": 1.3832,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9523809523809523,
|
| 62 |
+
"grad_norm": 0.6969241499900818,
|
| 63 |
+
"learning_rate": 6.349206349206349e-06,
|
| 64 |
+
"loss": 1.382,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.0,
|
| 69 |
+
"eval_accuracy": 0.58375,
|
| 70 |
+
"eval_loss": 1.3806676864624023,
|
| 71 |
+
"eval_runtime": 5.5764,
|
| 72 |
+
"eval_samples_per_second": 143.462,
|
| 73 |
+
"eval_steps_per_second": 17.933,
|
| 74 |
+
"step": 84
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.0714285714285714,
|
| 78 |
+
"grad_norm": 0.616682231426239,
|
| 79 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 80 |
+
"loss": 1.3805,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.1904761904761905,
|
| 85 |
+
"grad_norm": 0.774887204170227,
|
| 86 |
+
"learning_rate": 7.936507936507936e-06,
|
| 87 |
+
"loss": 1.3786,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.3095238095238095,
|
| 92 |
+
"grad_norm": 0.862555205821991,
|
| 93 |
+
"learning_rate": 8.730158730158731e-06,
|
| 94 |
+
"loss": 1.3754,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.4285714285714286,
|
| 99 |
+
"grad_norm": 0.7880147695541382,
|
| 100 |
+
"learning_rate": 9.523809523809525e-06,
|
| 101 |
+
"loss": 1.3714,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.5476190476190477,
|
| 106 |
+
"grad_norm": 0.9495492577552795,
|
| 107 |
+
"learning_rate": 1.031746031746032e-05,
|
| 108 |
+
"loss": 1.3663,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.6666666666666665,
|
| 113 |
+
"grad_norm": 1.0720769166946411,
|
| 114 |
+
"learning_rate": 1.1111111111111113e-05,
|
| 115 |
+
"loss": 1.3595,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.7857142857142856,
|
| 120 |
+
"grad_norm": 1.4009815454483032,
|
| 121 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 122 |
+
"loss": 1.3506,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.9047619047619047,
|
| 127 |
+
"grad_norm": 1.4374886751174927,
|
| 128 |
+
"learning_rate": 1.2698412698412699e-05,
|
| 129 |
+
"loss": 1.3401,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0,
|
| 134 |
+
"eval_accuracy": 0.44875,
|
| 135 |
+
"eval_loss": 1.31219482421875,
|
| 136 |
+
"eval_runtime": 5.4479,
|
| 137 |
+
"eval_samples_per_second": 146.845,
|
| 138 |
+
"eval_steps_per_second": 18.356,
|
| 139 |
+
"step": 168
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.0238095238095237,
|
| 143 |
+
"grad_norm": 1.4388487339019775,
|
| 144 |
+
"learning_rate": 1.3492063492063494e-05,
|
| 145 |
+
"loss": 1.3256,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.142857142857143,
|
| 150 |
+
"grad_norm": 2.0602617263793945,
|
| 151 |
+
"learning_rate": 1.4285714285714287e-05,
|
| 152 |
+
"loss": 1.3085,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.261904761904762,
|
| 157 |
+
"grad_norm": 2.5268349647521973,
|
| 158 |
+
"learning_rate": 1.507936507936508e-05,
|
| 159 |
+
"loss": 1.2787,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.380952380952381,
|
| 164 |
+
"grad_norm": 2.7516028881073,
|
| 165 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 166 |
+
"loss": 1.2603,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.5,
|
| 171 |
+
"grad_norm": 2.6878774166107178,
|
| 172 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 173 |
+
"loss": 1.2355,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.619047619047619,
|
| 178 |
+
"grad_norm": 1.6872291564941406,
|
| 179 |
+
"learning_rate": 1.7460317460317463e-05,
|
| 180 |
+
"loss": 1.201,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.738095238095238,
|
| 185 |
+
"grad_norm": 1.6558163166046143,
|
| 186 |
+
"learning_rate": 1.8253968253968254e-05,
|
| 187 |
+
"loss": 1.18,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.857142857142857,
|
| 192 |
+
"grad_norm": 2.9151134490966797,
|
| 193 |
+
"learning_rate": 1.904761904761905e-05,
|
| 194 |
+
"loss": 1.1485,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.9761904761904763,
|
| 199 |
+
"grad_norm": 1.5142790079116821,
|
| 200 |
+
"learning_rate": 1.9841269841269845e-05,
|
| 201 |
+
"loss": 1.1534,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 3.0,
|
| 206 |
+
"eval_accuracy": 0.5,
|
| 207 |
+
"eval_loss": 1.121809482574463,
|
| 208 |
+
"eval_runtime": 5.3391,
|
| 209 |
+
"eval_samples_per_second": 149.838,
|
| 210 |
+
"eval_steps_per_second": 18.73,
|
| 211 |
+
"step": 252
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.0952380952380953,
|
| 215 |
+
"grad_norm": 1.396998643875122,
|
| 216 |
+
"learning_rate": 1.9938271604938272e-05,
|
| 217 |
+
"loss": 1.1074,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.2142857142857144,
|
| 222 |
+
"grad_norm": 1.8247098922729492,
|
| 223 |
+
"learning_rate": 1.9850088183421517e-05,
|
| 224 |
+
"loss": 1.1146,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.3333333333333335,
|
| 229 |
+
"grad_norm": 3.359285831451416,
|
| 230 |
+
"learning_rate": 1.9761904761904763e-05,
|
| 231 |
+
"loss": 1.0917,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.4523809523809526,
|
| 236 |
+
"grad_norm": 2.1544318199157715,
|
| 237 |
+
"learning_rate": 1.9673721340388008e-05,
|
| 238 |
+
"loss": 1.0922,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.571428571428571,
|
| 243 |
+
"grad_norm": 1.6087664365768433,
|
| 244 |
+
"learning_rate": 1.9585537918871253e-05,
|
| 245 |
+
"loss": 1.076,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.6904761904761907,
|
| 250 |
+
"grad_norm": 2.3240392208099365,
|
| 251 |
+
"learning_rate": 1.94973544973545e-05,
|
| 252 |
+
"loss": 1.0453,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.8095238095238093,
|
| 257 |
+
"grad_norm": 2.1408257484436035,
|
| 258 |
+
"learning_rate": 1.9409171075837744e-05,
|
| 259 |
+
"loss": 1.0366,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.928571428571429,
|
| 264 |
+
"grad_norm": 4.0489583015441895,
|
| 265 |
+
"learning_rate": 1.932098765432099e-05,
|
| 266 |
+
"loss": 1.0343,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.0,
|
| 271 |
+
"eval_accuracy": 0.67625,
|
| 272 |
+
"eval_loss": 1.0234503746032715,
|
| 273 |
+
"eval_runtime": 5.3529,
|
| 274 |
+
"eval_samples_per_second": 149.453,
|
| 275 |
+
"eval_steps_per_second": 18.682,
|
| 276 |
+
"step": 336
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 4.0476190476190474,
|
| 280 |
+
"grad_norm": 3.3416860103607178,
|
| 281 |
+
"learning_rate": 1.9232804232804235e-05,
|
| 282 |
+
"loss": 1.0142,
|
| 283 |
+
"step": 340
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 4.166666666666667,
|
| 287 |
+
"grad_norm": 2.0407114028930664,
|
| 288 |
+
"learning_rate": 1.914462081128748e-05,
|
| 289 |
+
"loss": 0.9883,
|
| 290 |
+
"step": 350
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 4.285714285714286,
|
| 294 |
+
"grad_norm": 1.7006970643997192,
|
| 295 |
+
"learning_rate": 1.9056437389770726e-05,
|
| 296 |
+
"loss": 0.9741,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.404761904761905,
|
| 301 |
+
"grad_norm": 2.021171808242798,
|
| 302 |
+
"learning_rate": 1.8968253968253968e-05,
|
| 303 |
+
"loss": 0.9754,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.523809523809524,
|
| 308 |
+
"grad_norm": 4.418124675750732,
|
| 309 |
+
"learning_rate": 1.8880070546737216e-05,
|
| 310 |
+
"loss": 0.9428,
|
| 311 |
+
"step": 380
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.642857142857143,
|
| 315 |
+
"grad_norm": 2.6121156215667725,
|
| 316 |
+
"learning_rate": 1.8791887125220462e-05,
|
| 317 |
+
"loss": 0.9401,
|
| 318 |
+
"step": 390
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.761904761904762,
|
| 322 |
+
"grad_norm": 2.1322250366210938,
|
| 323 |
+
"learning_rate": 1.8703703703703707e-05,
|
| 324 |
+
"loss": 0.9409,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.880952380952381,
|
| 329 |
+
"grad_norm": 1.571338415145874,
|
| 330 |
+
"learning_rate": 1.861552028218695e-05,
|
| 331 |
+
"loss": 0.9062,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 5.0,
|
| 336 |
+
"grad_norm": Infinity,
|
| 337 |
+
"learning_rate": 1.853615520282187e-05,
|
| 338 |
+
"loss": 0.9117,
|
| 339 |
+
"step": 420
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.0,
|
| 343 |
+
"eval_accuracy": 0.7175,
|
| 344 |
+
"eval_loss": 0.9020006656646729,
|
| 345 |
+
"eval_runtime": 5.4379,
|
| 346 |
+
"eval_samples_per_second": 147.116,
|
| 347 |
+
"eval_steps_per_second": 18.389,
|
| 348 |
+
"step": 420
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.119047619047619,
|
| 352 |
+
"grad_norm": 4.1433868408203125,
|
| 353 |
+
"learning_rate": 1.8447971781305116e-05,
|
| 354 |
+
"loss": 0.8893,
|
| 355 |
+
"step": 430
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 5.238095238095238,
|
| 359 |
+
"grad_norm": 2.4193081855773926,
|
| 360 |
+
"learning_rate": 1.835978835978836e-05,
|
| 361 |
+
"loss": 0.8883,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.357142857142857,
|
| 366 |
+
"grad_norm": 3.0261456966400146,
|
| 367 |
+
"learning_rate": 1.8271604938271607e-05,
|
| 368 |
+
"loss": 0.8375,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.476190476190476,
|
| 373 |
+
"grad_norm": 4.649360656738281,
|
| 374 |
+
"learning_rate": 1.8183421516754852e-05,
|
| 375 |
+
"loss": 0.8617,
|
| 376 |
+
"step": 460
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.595238095238095,
|
| 380 |
+
"grad_norm": 2.217705011367798,
|
| 381 |
+
"learning_rate": 1.8095238095238097e-05,
|
| 382 |
+
"loss": 0.8377,
|
| 383 |
+
"step": 470
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.714285714285714,
|
| 387 |
+
"grad_norm": 3.888078212738037,
|
| 388 |
+
"learning_rate": 1.8007054673721343e-05,
|
| 389 |
+
"loss": 0.8027,
|
| 390 |
+
"step": 480
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.833333333333333,
|
| 394 |
+
"grad_norm": 3.240481376647949,
|
| 395 |
+
"learning_rate": 1.7918871252204585e-05,
|
| 396 |
+
"loss": 0.8258,
|
| 397 |
+
"step": 490
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.9523809523809526,
|
| 401 |
+
"grad_norm": 2.535261631011963,
|
| 402 |
+
"learning_rate": 1.783068783068783e-05,
|
| 403 |
+
"loss": 0.8169,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 6.0,
|
| 408 |
+
"eval_accuracy": 0.7925,
|
| 409 |
+
"eval_loss": 0.8129910230636597,
|
| 410 |
+
"eval_runtime": 5.2312,
|
| 411 |
+
"eval_samples_per_second": 152.927,
|
| 412 |
+
"eval_steps_per_second": 19.116,
|
| 413 |
+
"step": 504
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 6.071428571428571,
|
| 417 |
+
"grad_norm": 3.822288751602173,
|
| 418 |
+
"learning_rate": 1.774250440917108e-05,
|
| 419 |
+
"loss": 0.7932,
|
| 420 |
+
"step": 510
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 6.190476190476191,
|
| 424 |
+
"grad_norm": 2.448084592819214,
|
| 425 |
+
"learning_rate": 1.7654320987654324e-05,
|
| 426 |
+
"loss": 0.7807,
|
| 427 |
+
"step": 520
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 6.309523809523809,
|
| 431 |
+
"grad_norm": 4.276307106018066,
|
| 432 |
+
"learning_rate": 1.7566137566137566e-05,
|
| 433 |
+
"loss": 0.7736,
|
| 434 |
+
"step": 530
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.428571428571429,
|
| 438 |
+
"grad_norm": 3.5429012775421143,
|
| 439 |
+
"learning_rate": 1.747795414462081e-05,
|
| 440 |
+
"loss": 0.7453,
|
| 441 |
+
"step": 540
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.5476190476190474,
|
| 445 |
+
"grad_norm": 2.4782090187072754,
|
| 446 |
+
"learning_rate": 1.7389770723104057e-05,
|
| 447 |
+
"loss": 0.7327,
|
| 448 |
+
"step": 550
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.666666666666667,
|
| 452 |
+
"grad_norm": 5.342073440551758,
|
| 453 |
+
"learning_rate": 1.7301587301587302e-05,
|
| 454 |
+
"loss": 0.7247,
|
| 455 |
+
"step": 560
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.785714285714286,
|
| 459 |
+
"grad_norm": 10.06567668914795,
|
| 460 |
+
"learning_rate": 1.7213403880070548e-05,
|
| 461 |
+
"loss": 0.7088,
|
| 462 |
+
"step": 570
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.904761904761905,
|
| 466 |
+
"grad_norm": 3.217452049255371,
|
| 467 |
+
"learning_rate": 1.713403880070547e-05,
|
| 468 |
+
"loss": 0.7058,
|
| 469 |
+
"step": 580
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 7.0,
|
| 473 |
+
"eval_accuracy": 0.815,
|
| 474 |
+
"eval_loss": 0.7232482433319092,
|
| 475 |
+
"eval_runtime": 5.4248,
|
| 476 |
+
"eval_samples_per_second": 147.471,
|
| 477 |
+
"eval_steps_per_second": 18.434,
|
| 478 |
+
"step": 588
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.023809523809524,
|
| 482 |
+
"grad_norm": 4.111032009124756,
|
| 483 |
+
"learning_rate": 1.7045855379188714e-05,
|
| 484 |
+
"loss": 0.6987,
|
| 485 |
+
"step": 590
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.142857142857143,
|
| 489 |
+
"grad_norm": 2.667541027069092,
|
| 490 |
+
"learning_rate": 1.695767195767196e-05,
|
| 491 |
+
"loss": 0.684,
|
| 492 |
+
"step": 600
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.261904761904762,
|
| 496 |
+
"grad_norm": 8.819822311401367,
|
| 497 |
+
"learning_rate": 1.6869488536155205e-05,
|
| 498 |
+
"loss": 0.674,
|
| 499 |
+
"step": 610
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.380952380952381,
|
| 503 |
+
"grad_norm": 3.9701991081237793,
|
| 504 |
+
"learning_rate": 1.6781305114638447e-05,
|
| 505 |
+
"loss": 0.6199,
|
| 506 |
+
"step": 620
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.5,
|
| 510 |
+
"grad_norm": 8.835700035095215,
|
| 511 |
+
"learning_rate": 1.6693121693121696e-05,
|
| 512 |
+
"loss": 0.6241,
|
| 513 |
+
"step": 630
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.619047619047619,
|
| 517 |
+
"grad_norm": 3.7322373390197754,
|
| 518 |
+
"learning_rate": 1.660493827160494e-05,
|
| 519 |
+
"loss": 0.6087,
|
| 520 |
+
"step": 640
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.738095238095238,
|
| 524 |
+
"grad_norm": 4.442748069763184,
|
| 525 |
+
"learning_rate": 1.6516754850088187e-05,
|
| 526 |
+
"loss": 0.589,
|
| 527 |
+
"step": 650
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.857142857142857,
|
| 531 |
+
"grad_norm": 3.6890928745269775,
|
| 532 |
+
"learning_rate": 1.642857142857143e-05,
|
| 533 |
+
"loss": 0.5804,
|
| 534 |
+
"step": 660
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.976190476190476,
|
| 538 |
+
"grad_norm": 7.095566749572754,
|
| 539 |
+
"learning_rate": 1.6340388007054674e-05,
|
| 540 |
+
"loss": 0.5556,
|
| 541 |
+
"step": 670
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.0,
|
| 545 |
+
"eval_accuracy": 0.83375,
|
| 546 |
+
"eval_loss": 0.58147794008255,
|
| 547 |
+
"eval_runtime": 6.0471,
|
| 548 |
+
"eval_samples_per_second": 132.295,
|
| 549 |
+
"eval_steps_per_second": 16.537,
|
| 550 |
+
"step": 672
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 8.095238095238095,
|
| 554 |
+
"grad_norm": 4.2159318923950195,
|
| 555 |
+
"learning_rate": 1.625220458553792e-05,
|
| 556 |
+
"loss": 0.577,
|
| 557 |
+
"step": 680
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 8.214285714285714,
|
| 561 |
+
"grad_norm": 3.634282350540161,
|
| 562 |
+
"learning_rate": 1.6164021164021168e-05,
|
| 563 |
+
"loss": 0.5226,
|
| 564 |
+
"step": 690
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 8.333333333333334,
|
| 568 |
+
"grad_norm": 3.7204463481903076,
|
| 569 |
+
"learning_rate": 1.607583774250441e-05,
|
| 570 |
+
"loss": 0.5397,
|
| 571 |
+
"step": 700
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 8.452380952380953,
|
| 575 |
+
"grad_norm": 8.489058494567871,
|
| 576 |
+
"learning_rate": 1.5987654320987655e-05,
|
| 577 |
+
"loss": 0.515,
|
| 578 |
+
"step": 710
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 8.571428571428571,
|
| 582 |
+
"grad_norm": 8.01997184753418,
|
| 583 |
+
"learning_rate": 1.58994708994709e-05,
|
| 584 |
+
"loss": 0.5017,
|
| 585 |
+
"step": 720
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.69047619047619,
|
| 589 |
+
"grad_norm": 5.835849761962891,
|
| 590 |
+
"learning_rate": 1.5811287477954146e-05,
|
| 591 |
+
"loss": 0.5008,
|
| 592 |
+
"step": 730
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 8.80952380952381,
|
| 596 |
+
"grad_norm": 10.542766571044922,
|
| 597 |
+
"learning_rate": 1.572310405643739e-05,
|
| 598 |
+
"loss": 0.476,
|
| 599 |
+
"step": 740
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 8.928571428571429,
|
| 603 |
+
"grad_norm": 7.684937953948975,
|
| 604 |
+
"learning_rate": 1.5634920634920637e-05,
|
| 605 |
+
"loss": 0.4527,
|
| 606 |
+
"step": 750
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.0,
|
| 610 |
+
"eval_accuracy": 0.86625,
|
| 611 |
+
"eval_loss": 0.4813559651374817,
|
| 612 |
+
"eval_runtime": 6.5617,
|
| 613 |
+
"eval_samples_per_second": 121.92,
|
| 614 |
+
"eval_steps_per_second": 15.24,
|
| 615 |
+
"step": 756
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 9.047619047619047,
|
| 619 |
+
"grad_norm": 6.5577569007873535,
|
| 620 |
+
"learning_rate": 1.5546737213403882e-05,
|
| 621 |
+
"loss": 0.4875,
|
| 622 |
+
"step": 760
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 9.166666666666666,
|
| 626 |
+
"grad_norm": 6.667921543121338,
|
| 627 |
+
"learning_rate": 1.5458553791887128e-05,
|
| 628 |
+
"loss": 0.4581,
|
| 629 |
+
"step": 770
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 9.285714285714286,
|
| 633 |
+
"grad_norm": 5.33371114730835,
|
| 634 |
+
"learning_rate": 1.537037037037037e-05,
|
| 635 |
+
"loss": 0.4269,
|
| 636 |
+
"step": 780
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 9.404761904761905,
|
| 640 |
+
"grad_norm": 4.12121057510376,
|
| 641 |
+
"learning_rate": 1.5282186948853618e-05,
|
| 642 |
+
"loss": 0.4173,
|
| 643 |
+
"step": 790
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 9.523809523809524,
|
| 647 |
+
"grad_norm": 7.042411804199219,
|
| 648 |
+
"learning_rate": 1.5194003527336862e-05,
|
| 649 |
+
"loss": 0.4312,
|
| 650 |
+
"step": 800
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 9.642857142857142,
|
| 654 |
+
"grad_norm": 6.229002952575684,
|
| 655 |
+
"learning_rate": 1.5105820105820109e-05,
|
| 656 |
+
"loss": 0.421,
|
| 657 |
+
"step": 810
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 9.761904761904763,
|
| 661 |
+
"grad_norm": 9.01046085357666,
|
| 662 |
+
"learning_rate": 1.5017636684303351e-05,
|
| 663 |
+
"loss": 0.4182,
|
| 664 |
+
"step": 820
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 9.880952380952381,
|
| 668 |
+
"grad_norm": 9.333113670349121,
|
| 669 |
+
"learning_rate": 1.4929453262786598e-05,
|
| 670 |
+
"loss": 0.4065,
|
| 671 |
+
"step": 830
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 10.0,
|
| 675 |
+
"grad_norm": 5.7903876304626465,
|
| 676 |
+
"learning_rate": 1.4841269841269843e-05,
|
| 677 |
+
"loss": 0.3994,
|
| 678 |
+
"step": 840
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 10.0,
|
| 682 |
+
"eval_accuracy": 0.86875,
|
| 683 |
+
"eval_loss": 0.43535658717155457,
|
| 684 |
+
"eval_runtime": 6.424,
|
| 685 |
+
"eval_samples_per_second": 124.534,
|
| 686 |
+
"eval_steps_per_second": 15.567,
|
| 687 |
+
"step": 840
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 10.119047619047619,
|
| 691 |
+
"grad_norm": 4.672136306762695,
|
| 692 |
+
"learning_rate": 1.4753086419753087e-05,
|
| 693 |
+
"loss": 0.3692,
|
| 694 |
+
"step": 850
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 10.238095238095237,
|
| 698 |
+
"grad_norm": 9.825118064880371,
|
| 699 |
+
"learning_rate": 1.4664902998236332e-05,
|
| 700 |
+
"loss": 0.3946,
|
| 701 |
+
"step": 860
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 10.357142857142858,
|
| 705 |
+
"grad_norm": 10.594053268432617,
|
| 706 |
+
"learning_rate": 1.4576719576719578e-05,
|
| 707 |
+
"loss": 0.3836,
|
| 708 |
+
"step": 870
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 10.476190476190476,
|
| 712 |
+
"grad_norm": 6.456384658813477,
|
| 713 |
+
"learning_rate": 1.4488536155202823e-05,
|
| 714 |
+
"loss": 0.3922,
|
| 715 |
+
"step": 880
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 10.595238095238095,
|
| 719 |
+
"grad_norm": 7.883987903594971,
|
| 720 |
+
"learning_rate": 1.4400352733686067e-05,
|
| 721 |
+
"loss": 0.3767,
|
| 722 |
+
"step": 890
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 10.714285714285714,
|
| 726 |
+
"grad_norm": 3.752121686935425,
|
| 727 |
+
"learning_rate": 1.4312169312169312e-05,
|
| 728 |
+
"loss": 0.3731,
|
| 729 |
+
"step": 900
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 10.833333333333334,
|
| 733 |
+
"grad_norm": 5.878788948059082,
|
| 734 |
+
"learning_rate": 1.422398589065256e-05,
|
| 735 |
+
"loss": 0.3581,
|
| 736 |
+
"step": 910
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 10.952380952380953,
|
| 740 |
+
"grad_norm": 10.106928825378418,
|
| 741 |
+
"learning_rate": 1.4135802469135805e-05,
|
| 742 |
+
"loss": 0.3399,
|
| 743 |
+
"step": 920
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 11.0,
|
| 747 |
+
"eval_accuracy": 0.89,
|
| 748 |
+
"eval_loss": 0.3747052252292633,
|
| 749 |
+
"eval_runtime": 6.1766,
|
| 750 |
+
"eval_samples_per_second": 129.52,
|
| 751 |
+
"eval_steps_per_second": 16.19,
|
| 752 |
+
"step": 924
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 11.071428571428571,
|
| 756 |
+
"grad_norm": 13.221107482910156,
|
| 757 |
+
"learning_rate": 1.4047619047619048e-05,
|
| 758 |
+
"loss": 0.3512,
|
| 759 |
+
"step": 930
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 11.19047619047619,
|
| 763 |
+
"grad_norm": 5.013436317443848,
|
| 764 |
+
"learning_rate": 1.3959435626102294e-05,
|
| 765 |
+
"loss": 0.3217,
|
| 766 |
+
"step": 940
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 11.30952380952381,
|
| 770 |
+
"grad_norm": 4.782650470733643,
|
| 771 |
+
"learning_rate": 1.3871252204585539e-05,
|
| 772 |
+
"loss": 0.3603,
|
| 773 |
+
"step": 950
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 11.428571428571429,
|
| 777 |
+
"grad_norm": 7.278000354766846,
|
| 778 |
+
"learning_rate": 1.3783068783068784e-05,
|
| 779 |
+
"loss": 0.3193,
|
| 780 |
+
"step": 960
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 11.547619047619047,
|
| 784 |
+
"grad_norm": 13.20319652557373,
|
| 785 |
+
"learning_rate": 1.3694885361552028e-05,
|
| 786 |
+
"loss": 0.3429,
|
| 787 |
+
"step": 970
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 11.666666666666666,
|
| 791 |
+
"grad_norm": 10.208937644958496,
|
| 792 |
+
"learning_rate": 1.3606701940035273e-05,
|
| 793 |
+
"loss": 0.3391,
|
| 794 |
+
"step": 980
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 11.785714285714286,
|
| 798 |
+
"grad_norm": 8.55504322052002,
|
| 799 |
+
"learning_rate": 1.351851851851852e-05,
|
| 800 |
+
"loss": 0.3286,
|
| 801 |
+
"step": 990
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 11.904761904761905,
|
| 805 |
+
"grad_norm": 9.174301147460938,
|
| 806 |
+
"learning_rate": 1.3430335097001766e-05,
|
| 807 |
+
"loss": 0.3157,
|
| 808 |
+
"step": 1000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 12.0,
|
| 812 |
+
"eval_accuracy": 0.895,
|
| 813 |
+
"eval_loss": 0.33855053782463074,
|
| 814 |
+
"eval_runtime": 5.5685,
|
| 815 |
+
"eval_samples_per_second": 143.664,
|
| 816 |
+
"eval_steps_per_second": 17.958,
|
| 817 |
+
"step": 1008
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 12.023809523809524,
|
| 821 |
+
"grad_norm": 9.196775436401367,
|
| 822 |
+
"learning_rate": 1.334215167548501e-05,
|
| 823 |
+
"loss": 0.32,
|
| 824 |
+
"step": 1010
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 12.142857142857142,
|
| 828 |
+
"grad_norm": 7.491143703460693,
|
| 829 |
+
"learning_rate": 1.3253968253968255e-05,
|
| 830 |
+
"loss": 0.3211,
|
| 831 |
+
"step": 1020
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 12.261904761904763,
|
| 835 |
+
"grad_norm": 10.411942481994629,
|
| 836 |
+
"learning_rate": 1.31657848324515e-05,
|
| 837 |
+
"loss": 0.31,
|
| 838 |
+
"step": 1030
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 12.380952380952381,
|
| 842 |
+
"grad_norm": 7.7979021072387695,
|
| 843 |
+
"learning_rate": 1.3077601410934746e-05,
|
| 844 |
+
"loss": 0.301,
|
| 845 |
+
"step": 1040
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 12.5,
|
| 849 |
+
"grad_norm": 16.540781021118164,
|
| 850 |
+
"learning_rate": 1.298941798941799e-05,
|
| 851 |
+
"loss": 0.2879,
|
| 852 |
+
"step": 1050
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 12.619047619047619,
|
| 856 |
+
"grad_norm": 8.34303092956543,
|
| 857 |
+
"learning_rate": 1.2901234567901235e-05,
|
| 858 |
+
"loss": 0.2837,
|
| 859 |
+
"step": 1060
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 12.738095238095237,
|
| 863 |
+
"grad_norm": 5.1471405029296875,
|
| 864 |
+
"learning_rate": 1.2813051146384482e-05,
|
| 865 |
+
"loss": 0.2943,
|
| 866 |
+
"step": 1070
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 12.857142857142858,
|
| 870 |
+
"grad_norm": 12.514649391174316,
|
| 871 |
+
"learning_rate": 1.2724867724867727e-05,
|
| 872 |
+
"loss": 0.3033,
|
| 873 |
+
"step": 1080
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 12.976190476190476,
|
| 877 |
+
"grad_norm": 14.337882041931152,
|
| 878 |
+
"learning_rate": 1.263668430335097e-05,
|
| 879 |
+
"loss": 0.3094,
|
| 880 |
+
"step": 1090
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 13.0,
|
| 884 |
+
"eval_accuracy": 0.90375,
|
| 885 |
+
"eval_loss": 0.3166072368621826,
|
| 886 |
+
"eval_runtime": 5.6029,
|
| 887 |
+
"eval_samples_per_second": 142.783,
|
| 888 |
+
"eval_steps_per_second": 17.848,
|
| 889 |
+
"step": 1092
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 13.095238095238095,
|
| 893 |
+
"grad_norm": 7.664865493774414,
|
| 894 |
+
"learning_rate": 1.2548500881834216e-05,
|
| 895 |
+
"loss": 0.3027,
|
| 896 |
+
"step": 1100
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 13.214285714285714,
|
| 900 |
+
"grad_norm": 7.5357489585876465,
|
| 901 |
+
"learning_rate": 1.2460317460317461e-05,
|
| 902 |
+
"loss": 0.303,
|
| 903 |
+
"step": 1110
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 13.333333333333334,
|
| 907 |
+
"grad_norm": 10.726656913757324,
|
| 908 |
+
"learning_rate": 1.2372134038800707e-05,
|
| 909 |
+
"loss": 0.2897,
|
| 910 |
+
"step": 1120
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 13.452380952380953,
|
| 914 |
+
"grad_norm": 12.0068359375,
|
| 915 |
+
"learning_rate": 1.228395061728395e-05,
|
| 916 |
+
"loss": 0.2949,
|
| 917 |
+
"step": 1130
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 13.571428571428571,
|
| 921 |
+
"grad_norm": 9.883105278015137,
|
| 922 |
+
"learning_rate": 1.2195767195767196e-05,
|
| 923 |
+
"loss": 0.2573,
|
| 924 |
+
"step": 1140
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 13.69047619047619,
|
| 928 |
+
"grad_norm": 9.254776000976562,
|
| 929 |
+
"learning_rate": 1.2107583774250443e-05,
|
| 930 |
+
"loss": 0.2838,
|
| 931 |
+
"step": 1150
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 13.80952380952381,
|
| 935 |
+
"grad_norm": 6.0406880378723145,
|
| 936 |
+
"learning_rate": 1.2019400352733688e-05,
|
| 937 |
+
"loss": 0.2558,
|
| 938 |
+
"step": 1160
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 13.928571428571429,
|
| 942 |
+
"grad_norm": 6.664582252502441,
|
| 943 |
+
"learning_rate": 1.1931216931216932e-05,
|
| 944 |
+
"loss": 0.2839,
|
| 945 |
+
"step": 1170
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 14.0,
|
| 949 |
+
"eval_accuracy": 0.90125,
|
| 950 |
+
"eval_loss": 0.31680428981781006,
|
| 951 |
+
"eval_runtime": 5.4208,
|
| 952 |
+
"eval_samples_per_second": 147.579,
|
| 953 |
+
"eval_steps_per_second": 18.447,
|
| 954 |
+
"step": 1176
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 14.047619047619047,
|
| 958 |
+
"grad_norm": 11.964251518249512,
|
| 959 |
+
"learning_rate": 1.1843033509700177e-05,
|
| 960 |
+
"loss": 0.2688,
|
| 961 |
+
"step": 1180
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 14.166666666666666,
|
| 965 |
+
"grad_norm": 6.548354148864746,
|
| 966 |
+
"learning_rate": 1.1754850088183423e-05,
|
| 967 |
+
"loss": 0.2545,
|
| 968 |
+
"step": 1190
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 14.285714285714286,
|
| 972 |
+
"grad_norm": 14.916826248168945,
|
| 973 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 974 |
+
"loss": 0.2618,
|
| 975 |
+
"step": 1200
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 14.404761904761905,
|
| 979 |
+
"grad_norm": 15.267316818237305,
|
| 980 |
+
"learning_rate": 1.1578483245149912e-05,
|
| 981 |
+
"loss": 0.2442,
|
| 982 |
+
"step": 1210
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 14.523809523809524,
|
| 986 |
+
"grad_norm": 8.463184356689453,
|
| 987 |
+
"learning_rate": 1.1490299823633157e-05,
|
| 988 |
+
"loss": 0.2618,
|
| 989 |
+
"step": 1220
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 14.642857142857142,
|
| 993 |
+
"grad_norm": 13.68342113494873,
|
| 994 |
+
"learning_rate": 1.1402116402116404e-05,
|
| 995 |
+
"loss": 0.2447,
|
| 996 |
+
"step": 1230
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 14.761904761904763,
|
| 1000 |
+
"grad_norm": 4.8457112312316895,
|
| 1001 |
+
"learning_rate": 1.131393298059965e-05,
|
| 1002 |
+
"loss": 0.2361,
|
| 1003 |
+
"step": 1240
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 14.880952380952381,
|
| 1007 |
+
"grad_norm": 6.4722185134887695,
|
| 1008 |
+
"learning_rate": 1.1225749559082893e-05,
|
| 1009 |
+
"loss": 0.2519,
|
| 1010 |
+
"step": 1250
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 15.0,
|
| 1014 |
+
"grad_norm": 15.844517707824707,
|
| 1015 |
+
"learning_rate": 1.1137566137566138e-05,
|
| 1016 |
+
"loss": 0.2658,
|
| 1017 |
+
"step": 1260
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 15.0,
|
| 1021 |
+
"eval_accuracy": 0.91,
|
| 1022 |
+
"eval_loss": 0.2803078591823578,
|
| 1023 |
+
"eval_runtime": 5.3029,
|
| 1024 |
+
"eval_samples_per_second": 150.862,
|
| 1025 |
+
"eval_steps_per_second": 18.858,
|
| 1026 |
+
"step": 1260
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 15.119047619047619,
|
| 1030 |
+
"grad_norm": 20.211238861083984,
|
| 1031 |
+
"learning_rate": 1.1049382716049384e-05,
|
| 1032 |
+
"loss": 0.2386,
|
| 1033 |
+
"step": 1270
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 15.238095238095237,
|
| 1037 |
+
"grad_norm": 6.853188991546631,
|
| 1038 |
+
"learning_rate": 1.0961199294532629e-05,
|
| 1039 |
+
"loss": 0.2326,
|
| 1040 |
+
"step": 1280
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 15.357142857142858,
|
| 1044 |
+
"grad_norm": 8.602685928344727,
|
| 1045 |
+
"learning_rate": 1.0873015873015873e-05,
|
| 1046 |
+
"loss": 0.2553,
|
| 1047 |
+
"step": 1290
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 15.476190476190476,
|
| 1051 |
+
"grad_norm": 4.291443824768066,
|
| 1052 |
+
"learning_rate": 1.0784832451499118e-05,
|
| 1053 |
+
"loss": 0.2434,
|
| 1054 |
+
"step": 1300
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 15.595238095238095,
|
| 1058 |
+
"grad_norm": 6.306275367736816,
|
| 1059 |
+
"learning_rate": 1.0696649029982365e-05,
|
| 1060 |
+
"loss": 0.2277,
|
| 1061 |
+
"step": 1310
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 15.714285714285714,
|
| 1065 |
+
"grad_norm": 5.804137706756592,
|
| 1066 |
+
"learning_rate": 1.060846560846561e-05,
|
| 1067 |
+
"loss": 0.2248,
|
| 1068 |
+
"step": 1320
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 15.833333333333334,
|
| 1072 |
+
"grad_norm": 13.754738807678223,
|
| 1073 |
+
"learning_rate": 1.0520282186948854e-05,
|
| 1074 |
+
"loss": 0.2313,
|
| 1075 |
+
"step": 1330
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 15.952380952380953,
|
| 1079 |
+
"grad_norm": 8.410860061645508,
|
| 1080 |
+
"learning_rate": 1.04320987654321e-05,
|
| 1081 |
+
"loss": 0.2331,
|
| 1082 |
+
"step": 1340
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 16.0,
|
| 1086 |
+
"eval_accuracy": 0.9125,
|
| 1087 |
+
"eval_loss": 0.26975059509277344,
|
| 1088 |
+
"eval_runtime": 5.4078,
|
| 1089 |
+
"eval_samples_per_second": 147.934,
|
| 1090 |
+
"eval_steps_per_second": 18.492,
|
| 1091 |
+
"step": 1344
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 16.071428571428573,
|
| 1095 |
+
"grad_norm": 9.222555160522461,
|
| 1096 |
+
"learning_rate": 1.0352733686067021e-05,
|
| 1097 |
+
"loss": 0.2394,
|
| 1098 |
+
"step": 1350
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 16.19047619047619,
|
| 1102 |
+
"grad_norm": 9.493171691894531,
|
| 1103 |
+
"learning_rate": 1.0264550264550266e-05,
|
| 1104 |
+
"loss": 0.2213,
|
| 1105 |
+
"step": 1360
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 16.30952380952381,
|
| 1109 |
+
"grad_norm": 4.863091945648193,
|
| 1110 |
+
"learning_rate": 1.0176366843033512e-05,
|
| 1111 |
+
"loss": 0.2447,
|
| 1112 |
+
"step": 1370
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 16.428571428571427,
|
| 1116 |
+
"grad_norm": 9.915203094482422,
|
| 1117 |
+
"learning_rate": 1.0088183421516755e-05,
|
| 1118 |
+
"loss": 0.2369,
|
| 1119 |
+
"step": 1380
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 16.547619047619047,
|
| 1123 |
+
"grad_norm": 7.485960960388184,
|
| 1124 |
+
"learning_rate": 1e-05,
|
| 1125 |
+
"loss": 0.213,
|
| 1126 |
+
"step": 1390
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 16.666666666666668,
|
| 1130 |
+
"grad_norm": 5.043095111846924,
|
| 1131 |
+
"learning_rate": 9.911816578483246e-06,
|
| 1132 |
+
"loss": 0.2035,
|
| 1133 |
+
"step": 1400
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 16.785714285714285,
|
| 1137 |
+
"grad_norm": 12.636078834533691,
|
| 1138 |
+
"learning_rate": 9.823633156966492e-06,
|
| 1139 |
+
"loss": 0.2312,
|
| 1140 |
+
"step": 1410
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 16.904761904761905,
|
| 1144 |
+
"grad_norm": 13.119421005249023,
|
| 1145 |
+
"learning_rate": 9.735449735449735e-06,
|
| 1146 |
+
"loss": 0.2271,
|
| 1147 |
+
"step": 1420
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 17.0,
|
| 1151 |
+
"eval_accuracy": 0.90625,
|
| 1152 |
+
"eval_loss": 0.2662568986415863,
|
| 1153 |
+
"eval_runtime": 5.5739,
|
| 1154 |
+
"eval_samples_per_second": 143.526,
|
| 1155 |
+
"eval_steps_per_second": 17.941,
|
| 1156 |
+
"step": 1428
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 17.023809523809526,
|
| 1160 |
+
"grad_norm": 19.145736694335938,
|
| 1161 |
+
"learning_rate": 9.64726631393298e-06,
|
| 1162 |
+
"loss": 0.2089,
|
| 1163 |
+
"step": 1430
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 17.142857142857142,
|
| 1167 |
+
"grad_norm": 13.143487930297852,
|
| 1168 |
+
"learning_rate": 9.559082892416226e-06,
|
| 1169 |
+
"loss": 0.2036,
|
| 1170 |
+
"step": 1440
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 17.261904761904763,
|
| 1174 |
+
"grad_norm": 13.245809555053711,
|
| 1175 |
+
"learning_rate": 9.470899470899471e-06,
|
| 1176 |
+
"loss": 0.1902,
|
| 1177 |
+
"step": 1450
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 17.38095238095238,
|
| 1181 |
+
"grad_norm": 15.815811157226562,
|
| 1182 |
+
"learning_rate": 9.382716049382717e-06,
|
| 1183 |
+
"loss": 0.2234,
|
| 1184 |
+
"step": 1460
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 17.5,
|
| 1188 |
+
"grad_norm": 8.950321197509766,
|
| 1189 |
+
"learning_rate": 9.294532627865962e-06,
|
| 1190 |
+
"loss": 0.2258,
|
| 1191 |
+
"step": 1470
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 17.61904761904762,
|
| 1195 |
+
"grad_norm": 6.937575817108154,
|
| 1196 |
+
"learning_rate": 9.206349206349207e-06,
|
| 1197 |
+
"loss": 0.212,
|
| 1198 |
+
"step": 1480
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 17.738095238095237,
|
| 1202 |
+
"grad_norm": 9.044116020202637,
|
| 1203 |
+
"learning_rate": 9.118165784832453e-06,
|
| 1204 |
+
"loss": 0.2084,
|
| 1205 |
+
"step": 1490
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 17.857142857142858,
|
| 1209 |
+
"grad_norm": 9.224007606506348,
|
| 1210 |
+
"learning_rate": 9.029982363315696e-06,
|
| 1211 |
+
"loss": 0.2036,
|
| 1212 |
+
"step": 1500
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 17.976190476190474,
|
| 1216 |
+
"grad_norm": 5.575143814086914,
|
| 1217 |
+
"learning_rate": 8.941798941798942e-06,
|
| 1218 |
+
"loss": 0.1895,
|
| 1219 |
+
"step": 1510
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 18.0,
|
| 1223 |
+
"eval_accuracy": 0.92375,
|
| 1224 |
+
"eval_loss": 0.2639971375465393,
|
| 1225 |
+
"eval_runtime": 5.9356,
|
| 1226 |
+
"eval_samples_per_second": 134.779,
|
| 1227 |
+
"eval_steps_per_second": 16.847,
|
| 1228 |
+
"step": 1512
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 18.095238095238095,
|
| 1232 |
+
"grad_norm": 10.496304512023926,
|
| 1233 |
+
"learning_rate": 8.853615520282187e-06,
|
| 1234 |
+
"loss": 0.1908,
|
| 1235 |
+
"step": 1520
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 18.214285714285715,
|
| 1239 |
+
"grad_norm": 8.395342826843262,
|
| 1240 |
+
"learning_rate": 8.765432098765432e-06,
|
| 1241 |
+
"loss": 0.202,
|
| 1242 |
+
"step": 1530
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 18.333333333333332,
|
| 1246 |
+
"grad_norm": 6.879480838775635,
|
| 1247 |
+
"learning_rate": 8.677248677248678e-06,
|
| 1248 |
+
"loss": 0.1825,
|
| 1249 |
+
"step": 1540
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 18.452380952380953,
|
| 1253 |
+
"grad_norm": 6.519415855407715,
|
| 1254 |
+
"learning_rate": 8.589065255731923e-06,
|
| 1255 |
+
"loss": 0.2025,
|
| 1256 |
+
"step": 1550
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 18.571428571428573,
|
| 1260 |
+
"grad_norm": 14.108360290527344,
|
| 1261 |
+
"learning_rate": 8.500881834215169e-06,
|
| 1262 |
+
"loss": 0.2156,
|
| 1263 |
+
"step": 1560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 18.69047619047619,
|
| 1267 |
+
"grad_norm": 8.98913288116455,
|
| 1268 |
+
"learning_rate": 8.412698412698414e-06,
|
| 1269 |
+
"loss": 0.1895,
|
| 1270 |
+
"step": 1570
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 18.80952380952381,
|
| 1274 |
+
"grad_norm": 7.803181171417236,
|
| 1275 |
+
"learning_rate": 8.324514991181658e-06,
|
| 1276 |
+
"loss": 0.1829,
|
| 1277 |
+
"step": 1580
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 18.928571428571427,
|
| 1281 |
+
"grad_norm": 6.913029193878174,
|
| 1282 |
+
"learning_rate": 8.236331569664903e-06,
|
| 1283 |
+
"loss": 0.1914,
|
| 1284 |
+
"step": 1590
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 19.0,
|
| 1288 |
+
"eval_accuracy": 0.9225,
|
| 1289 |
+
"eval_loss": 0.23814032971858978,
|
| 1290 |
+
"eval_runtime": 6.1225,
|
| 1291 |
+
"eval_samples_per_second": 130.665,
|
| 1292 |
+
"eval_steps_per_second": 16.333,
|
| 1293 |
+
"step": 1596
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 19.047619047619047,
|
| 1297 |
+
"grad_norm": 8.117836952209473,
|
| 1298 |
+
"learning_rate": 8.148148148148148e-06,
|
| 1299 |
+
"loss": 0.1953,
|
| 1300 |
+
"step": 1600
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 19.166666666666668,
|
| 1304 |
+
"grad_norm": 7.4035139083862305,
|
| 1305 |
+
"learning_rate": 8.059964726631394e-06,
|
| 1306 |
+
"loss": 0.1755,
|
| 1307 |
+
"step": 1610
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 19.285714285714285,
|
| 1311 |
+
"grad_norm": 9.961386680603027,
|
| 1312 |
+
"learning_rate": 7.971781305114639e-06,
|
| 1313 |
+
"loss": 0.1826,
|
| 1314 |
+
"step": 1620
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 19.404761904761905,
|
| 1318 |
+
"grad_norm": 3.902977705001831,
|
| 1319 |
+
"learning_rate": 7.883597883597884e-06,
|
| 1320 |
+
"loss": 0.2038,
|
| 1321 |
+
"step": 1630
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 19.523809523809526,
|
| 1325 |
+
"grad_norm": 11.7597074508667,
|
| 1326 |
+
"learning_rate": 7.79541446208113e-06,
|
| 1327 |
+
"loss": 0.1755,
|
| 1328 |
+
"step": 1640
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 19.642857142857142,
|
| 1332 |
+
"grad_norm": 5.295619964599609,
|
| 1333 |
+
"learning_rate": 7.707231040564375e-06,
|
| 1334 |
+
"loss": 0.1821,
|
| 1335 |
+
"step": 1650
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 19.761904761904763,
|
| 1339 |
+
"grad_norm": 8.691162109375,
|
| 1340 |
+
"learning_rate": 7.61904761904762e-06,
|
| 1341 |
+
"loss": 0.1931,
|
| 1342 |
+
"step": 1660
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 19.88095238095238,
|
| 1346 |
+
"grad_norm": 14.22559928894043,
|
| 1347 |
+
"learning_rate": 7.530864197530865e-06,
|
| 1348 |
+
"loss": 0.1842,
|
| 1349 |
+
"step": 1670
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 20.0,
|
| 1353 |
+
"grad_norm": 19.08734130859375,
|
| 1354 |
+
"learning_rate": 7.4426807760141095e-06,
|
| 1355 |
+
"loss": 0.1741,
|
| 1356 |
+
"step": 1680
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 20.0,
|
| 1360 |
+
"eval_accuracy": 0.92125,
|
| 1361 |
+
"eval_loss": 0.2286679595708847,
|
| 1362 |
+
"eval_runtime": 6.3863,
|
| 1363 |
+
"eval_samples_per_second": 125.269,
|
| 1364 |
+
"eval_steps_per_second": 15.659,
|
| 1365 |
+
"step": 1680
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 20.11904761904762,
|
| 1369 |
+
"grad_norm": 11.843003273010254,
|
| 1370 |
+
"learning_rate": 7.354497354497355e-06,
|
| 1371 |
+
"loss": 0.1704,
|
| 1372 |
+
"step": 1690
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 20.238095238095237,
|
| 1376 |
+
"grad_norm": 4.4936299324035645,
|
| 1377 |
+
"learning_rate": 7.2663139329806e-06,
|
| 1378 |
+
"loss": 0.1833,
|
| 1379 |
+
"step": 1700
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 20.357142857142858,
|
| 1383 |
+
"grad_norm": 3.9449779987335205,
|
| 1384 |
+
"learning_rate": 7.178130511463846e-06,
|
| 1385 |
+
"loss": 0.1779,
|
| 1386 |
+
"step": 1710
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 20.476190476190474,
|
| 1390 |
+
"grad_norm": 9.936880111694336,
|
| 1391 |
+
"learning_rate": 7.08994708994709e-06,
|
| 1392 |
+
"loss": 0.1789,
|
| 1393 |
+
"step": 1720
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 20.595238095238095,
|
| 1397 |
+
"grad_norm": 5.483986854553223,
|
| 1398 |
+
"learning_rate": 7.0017636684303355e-06,
|
| 1399 |
+
"loss": 0.1776,
|
| 1400 |
+
"step": 1730
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 20.714285714285715,
|
| 1404 |
+
"grad_norm": 8.04780101776123,
|
| 1405 |
+
"learning_rate": 6.913580246913581e-06,
|
| 1406 |
+
"loss": 0.1737,
|
| 1407 |
+
"step": 1740
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 20.833333333333332,
|
| 1411 |
+
"grad_norm": 6.709885120391846,
|
| 1412 |
+
"learning_rate": 6.825396825396826e-06,
|
| 1413 |
+
"loss": 0.1855,
|
| 1414 |
+
"step": 1750
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 20.952380952380953,
|
| 1418 |
+
"grad_norm": 7.882541656494141,
|
| 1419 |
+
"learning_rate": 6.737213403880071e-06,
|
| 1420 |
+
"loss": 0.1682,
|
| 1421 |
+
"step": 1760
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 21.0,
|
| 1425 |
+
"eval_accuracy": 0.92625,
|
| 1426 |
+
"eval_loss": 0.21940070390701294,
|
| 1427 |
+
"eval_runtime": 5.4374,
|
| 1428 |
+
"eval_samples_per_second": 147.128,
|
| 1429 |
+
"eval_steps_per_second": 18.391,
|
| 1430 |
+
"step": 1764
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 21.071428571428573,
|
| 1434 |
+
"grad_norm": 5.953088760375977,
|
| 1435 |
+
"learning_rate": 6.649029982363316e-06,
|
| 1436 |
+
"loss": 0.1741,
|
| 1437 |
+
"step": 1770
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 21.19047619047619,
|
| 1441 |
+
"grad_norm": 7.106684684753418,
|
| 1442 |
+
"learning_rate": 6.560846560846561e-06,
|
| 1443 |
+
"loss": 0.1745,
|
| 1444 |
+
"step": 1780
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 21.30952380952381,
|
| 1448 |
+
"grad_norm": 13.026501655578613,
|
| 1449 |
+
"learning_rate": 6.472663139329807e-06,
|
| 1450 |
+
"loss": 0.1844,
|
| 1451 |
+
"step": 1790
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 21.428571428571427,
|
| 1455 |
+
"grad_norm": 3.5783817768096924,
|
| 1456 |
+
"learning_rate": 6.384479717813051e-06,
|
| 1457 |
+
"loss": 0.1636,
|
| 1458 |
+
"step": 1800
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 21.547619047619047,
|
| 1462 |
+
"grad_norm": 10.339872360229492,
|
| 1463 |
+
"learning_rate": 6.296296296296297e-06,
|
| 1464 |
+
"loss": 0.1766,
|
| 1465 |
+
"step": 1810
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 21.666666666666668,
|
| 1469 |
+
"grad_norm": 13.413084030151367,
|
| 1470 |
+
"learning_rate": 6.208112874779542e-06,
|
| 1471 |
+
"loss": 0.1643,
|
| 1472 |
+
"step": 1820
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 21.785714285714285,
|
| 1476 |
+
"grad_norm": 6.75968074798584,
|
| 1477 |
+
"learning_rate": 6.119929453262787e-06,
|
| 1478 |
+
"loss": 0.1589,
|
| 1479 |
+
"step": 1830
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 21.904761904761905,
|
| 1483 |
+
"grad_norm": 3.4033656120300293,
|
| 1484 |
+
"learning_rate": 6.031746031746032e-06,
|
| 1485 |
+
"loss": 0.1569,
|
| 1486 |
+
"step": 1840
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 22.0,
|
| 1490 |
+
"eval_accuracy": 0.93375,
|
| 1491 |
+
"eval_loss": 0.21772493422031403,
|
| 1492 |
+
"eval_runtime": 5.4159,
|
| 1493 |
+
"eval_samples_per_second": 147.714,
|
| 1494 |
+
"eval_steps_per_second": 18.464,
|
| 1495 |
+
"step": 1848
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 22.023809523809526,
|
| 1499 |
+
"grad_norm": 28.122211456298828,
|
| 1500 |
+
"learning_rate": 5.943562610229277e-06,
|
| 1501 |
+
"loss": 0.1648,
|
| 1502 |
+
"step": 1850
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 22.142857142857142,
|
| 1506 |
+
"grad_norm": 9.448488235473633,
|
| 1507 |
+
"learning_rate": 5.855379188712523e-06,
|
| 1508 |
+
"loss": 0.1397,
|
| 1509 |
+
"step": 1860
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 22.261904761904763,
|
| 1513 |
+
"grad_norm": 7.661059379577637,
|
| 1514 |
+
"learning_rate": 5.767195767195768e-06,
|
| 1515 |
+
"loss": 0.1756,
|
| 1516 |
+
"step": 1870
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 22.38095238095238,
|
| 1520 |
+
"grad_norm": 10.659194946289062,
|
| 1521 |
+
"learning_rate": 5.6790123456790125e-06,
|
| 1522 |
+
"loss": 0.1588,
|
| 1523 |
+
"step": 1880
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 22.5,
|
| 1527 |
+
"grad_norm": 3.608710527420044,
|
| 1528 |
+
"learning_rate": 5.590828924162258e-06,
|
| 1529 |
+
"loss": 0.1656,
|
| 1530 |
+
"step": 1890
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 22.61904761904762,
|
| 1534 |
+
"grad_norm": 10.583845138549805,
|
| 1535 |
+
"learning_rate": 5.502645502645503e-06,
|
| 1536 |
+
"loss": 0.1618,
|
| 1537 |
+
"step": 1900
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 22.738095238095237,
|
| 1541 |
+
"grad_norm": 8.316349029541016,
|
| 1542 |
+
"learning_rate": 5.4144620811287486e-06,
|
| 1543 |
+
"loss": 0.1541,
|
| 1544 |
+
"step": 1910
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 22.857142857142858,
|
| 1548 |
+
"grad_norm": 9.836575508117676,
|
| 1549 |
+
"learning_rate": 5.326278659611993e-06,
|
| 1550 |
+
"loss": 0.1588,
|
| 1551 |
+
"step": 1920
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 22.976190476190474,
|
| 1555 |
+
"grad_norm": 7.263607501983643,
|
| 1556 |
+
"learning_rate": 5.2380952380952384e-06,
|
| 1557 |
+
"loss": 0.144,
|
| 1558 |
+
"step": 1930
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 23.0,
|
| 1562 |
+
"eval_accuracy": 0.93375,
|
| 1563 |
+
"eval_loss": 0.21350684762001038,
|
| 1564 |
+
"eval_runtime": 5.4573,
|
| 1565 |
+
"eval_samples_per_second": 146.592,
|
| 1566 |
+
"eval_steps_per_second": 18.324,
|
| 1567 |
+
"step": 1932
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 23.095238095238095,
|
| 1571 |
+
"grad_norm": 9.257414817810059,
|
| 1572 |
+
"learning_rate": 5.149911816578484e-06,
|
| 1573 |
+
"loss": 0.1624,
|
| 1574 |
+
"step": 1940
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 23.214285714285715,
|
| 1578 |
+
"grad_norm": 23.810653686523438,
|
| 1579 |
+
"learning_rate": 5.061728395061729e-06,
|
| 1580 |
+
"loss": 0.1528,
|
| 1581 |
+
"step": 1950
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 23.333333333333332,
|
| 1585 |
+
"grad_norm": 8.397602081298828,
|
| 1586 |
+
"learning_rate": 4.973544973544974e-06,
|
| 1587 |
+
"loss": 0.1506,
|
| 1588 |
+
"step": 1960
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 23.452380952380953,
|
| 1592 |
+
"grad_norm": 3.9385476112365723,
|
| 1593 |
+
"learning_rate": 4.885361552028219e-06,
|
| 1594 |
+
"loss": 0.1427,
|
| 1595 |
+
"step": 1970
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 23.571428571428573,
|
| 1599 |
+
"grad_norm": 9.915690422058105,
|
| 1600 |
+
"learning_rate": 4.7971781305114636e-06,
|
| 1601 |
+
"loss": 0.1548,
|
| 1602 |
+
"step": 1980
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 23.69047619047619,
|
| 1606 |
+
"grad_norm": 10.970438003540039,
|
| 1607 |
+
"learning_rate": 4.708994708994709e-06,
|
| 1608 |
+
"loss": 0.165,
|
| 1609 |
+
"step": 1990
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 23.80952380952381,
|
| 1613 |
+
"grad_norm": 12.917925834655762,
|
| 1614 |
+
"learning_rate": 4.620811287477954e-06,
|
| 1615 |
+
"loss": 0.1658,
|
| 1616 |
+
"step": 2000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 23.928571428571427,
|
| 1620 |
+
"grad_norm": 16.563697814941406,
|
| 1621 |
+
"learning_rate": 4.5326278659612e-06,
|
| 1622 |
+
"loss": 0.1581,
|
| 1623 |
+
"step": 2010
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 24.0,
|
| 1627 |
+
"eval_accuracy": 0.9325,
|
| 1628 |
+
"eval_loss": 0.20941905677318573,
|
| 1629 |
+
"eval_runtime": 6.2026,
|
| 1630 |
+
"eval_samples_per_second": 128.978,
|
| 1631 |
+
"eval_steps_per_second": 16.122,
|
| 1632 |
+
"step": 2016
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 24.047619047619047,
|
| 1636 |
+
"grad_norm": 5.663702487945557,
|
| 1637 |
+
"learning_rate": 4.444444444444444e-06,
|
| 1638 |
+
"loss": 0.1742,
|
| 1639 |
+
"step": 2020
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 24.166666666666668,
|
| 1643 |
+
"grad_norm": 21.897550582885742,
|
| 1644 |
+
"learning_rate": 4.3562610229276895e-06,
|
| 1645 |
+
"loss": 0.1663,
|
| 1646 |
+
"step": 2030
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 24.285714285714285,
|
| 1650 |
+
"grad_norm": 7.413302421569824,
|
| 1651 |
+
"learning_rate": 4.268077601410935e-06,
|
| 1652 |
+
"loss": 0.1478,
|
| 1653 |
+
"step": 2040
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 24.404761904761905,
|
| 1657 |
+
"grad_norm": 9.5319242477417,
|
| 1658 |
+
"learning_rate": 4.17989417989418e-06,
|
| 1659 |
+
"loss": 0.1616,
|
| 1660 |
+
"step": 2050
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 24.523809523809526,
|
| 1664 |
+
"grad_norm": 7.299759864807129,
|
| 1665 |
+
"learning_rate": 4.091710758377425e-06,
|
| 1666 |
+
"loss": 0.1601,
|
| 1667 |
+
"step": 2060
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 24.642857142857142,
|
| 1671 |
+
"grad_norm": 7.582053184509277,
|
| 1672 |
+
"learning_rate": 4.00352733686067e-06,
|
| 1673 |
+
"loss": 0.1481,
|
| 1674 |
+
"step": 2070
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 24.761904761904763,
|
| 1678 |
+
"grad_norm": 6.607138633728027,
|
| 1679 |
+
"learning_rate": 3.9153439153439155e-06,
|
| 1680 |
+
"loss": 0.1447,
|
| 1681 |
+
"step": 2080
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 24.88095238095238,
|
| 1685 |
+
"grad_norm": 5.343249320983887,
|
| 1686 |
+
"learning_rate": 3.827160493827161e-06,
|
| 1687 |
+
"loss": 0.1366,
|
| 1688 |
+
"step": 2090
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 25.0,
|
| 1692 |
+
"grad_norm": 11.388031005859375,
|
| 1693 |
+
"learning_rate": 3.7389770723104058e-06,
|
| 1694 |
+
"loss": 0.1426,
|
| 1695 |
+
"step": 2100
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 25.0,
|
| 1699 |
+
"eval_accuracy": 0.935,
|
| 1700 |
+
"eval_loss": 0.2057761698961258,
|
| 1701 |
+
"eval_runtime": 6.4445,
|
| 1702 |
+
"eval_samples_per_second": 124.136,
|
| 1703 |
+
"eval_steps_per_second": 15.517,
|
| 1704 |
+
"step": 2100
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 25.11904761904762,
|
| 1708 |
+
"grad_norm": 13.238511085510254,
|
| 1709 |
+
"learning_rate": 3.6507936507936507e-06,
|
| 1710 |
+
"loss": 0.1451,
|
| 1711 |
+
"step": 2110
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 25.238095238095237,
|
| 1715 |
+
"grad_norm": 18.473859786987305,
|
| 1716 |
+
"learning_rate": 3.562610229276896e-06,
|
| 1717 |
+
"loss": 0.146,
|
| 1718 |
+
"step": 2120
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 25.357142857142858,
|
| 1722 |
+
"grad_norm": 8.533565521240234,
|
| 1723 |
+
"learning_rate": 3.474426807760141e-06,
|
| 1724 |
+
"loss": 0.15,
|
| 1725 |
+
"step": 2130
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 25.476190476190474,
|
| 1729 |
+
"grad_norm": 8.290474891662598,
|
| 1730 |
+
"learning_rate": 3.3862433862433864e-06,
|
| 1731 |
+
"loss": 0.1367,
|
| 1732 |
+
"step": 2140
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 25.595238095238095,
|
| 1736 |
+
"grad_norm": 3.6419670581817627,
|
| 1737 |
+
"learning_rate": 3.2980599647266313e-06,
|
| 1738 |
+
"loss": 0.1392,
|
| 1739 |
+
"step": 2150
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 25.714285714285715,
|
| 1743 |
+
"grad_norm": 16.031980514526367,
|
| 1744 |
+
"learning_rate": 3.2098765432098767e-06,
|
| 1745 |
+
"loss": 0.1496,
|
| 1746 |
+
"step": 2160
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 25.833333333333332,
|
| 1750 |
+
"grad_norm": 16.360395431518555,
|
| 1751 |
+
"learning_rate": 3.1216931216931216e-06,
|
| 1752 |
+
"loss": 0.1705,
|
| 1753 |
+
"step": 2170
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 25.952380952380953,
|
| 1757 |
+
"grad_norm": 7.6555280685424805,
|
| 1758 |
+
"learning_rate": 3.033509700176367e-06,
|
| 1759 |
+
"loss": 0.1409,
|
| 1760 |
+
"step": 2180
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 26.0,
|
| 1764 |
+
"eval_accuracy": 0.93375,
|
| 1765 |
+
"eval_loss": 0.2027110904455185,
|
| 1766 |
+
"eval_runtime": 6.4093,
|
| 1767 |
+
"eval_samples_per_second": 124.818,
|
| 1768 |
+
"eval_steps_per_second": 15.602,
|
| 1769 |
+
"step": 2184
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 26.071428571428573,
|
| 1773 |
+
"grad_norm": 13.298584938049316,
|
| 1774 |
+
"learning_rate": 2.945326278659612e-06,
|
| 1775 |
+
"loss": 0.1339,
|
| 1776 |
+
"step": 2190
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 26.19047619047619,
|
| 1780 |
+
"grad_norm": 5.4013800621032715,
|
| 1781 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 1782 |
+
"loss": 0.1388,
|
| 1783 |
+
"step": 2200
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 26.30952380952381,
|
| 1787 |
+
"grad_norm": 6.846310138702393,
|
| 1788 |
+
"learning_rate": 2.768959435626102e-06,
|
| 1789 |
+
"loss": 0.1379,
|
| 1790 |
+
"step": 2210
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 26.428571428571427,
|
| 1794 |
+
"grad_norm": 8.436290740966797,
|
| 1795 |
+
"learning_rate": 2.6807760141093476e-06,
|
| 1796 |
+
"loss": 0.1404,
|
| 1797 |
+
"step": 2220
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 26.547619047619047,
|
| 1801 |
+
"grad_norm": 8.173290252685547,
|
| 1802 |
+
"learning_rate": 2.5925925925925925e-06,
|
| 1803 |
+
"loss": 0.1526,
|
| 1804 |
+
"step": 2230
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 26.666666666666668,
|
| 1808 |
+
"grad_norm": 7.008592128753662,
|
| 1809 |
+
"learning_rate": 2.504409171075838e-06,
|
| 1810 |
+
"loss": 0.1656,
|
| 1811 |
+
"step": 2240
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 26.785714285714285,
|
| 1815 |
+
"grad_norm": 14.528263092041016,
|
| 1816 |
+
"learning_rate": 2.416225749559083e-06,
|
| 1817 |
+
"loss": 0.1403,
|
| 1818 |
+
"step": 2250
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 26.904761904761905,
|
| 1822 |
+
"grad_norm": 19.173906326293945,
|
| 1823 |
+
"learning_rate": 2.328042328042328e-06,
|
| 1824 |
+
"loss": 0.1445,
|
| 1825 |
+
"step": 2260
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 27.0,
|
| 1829 |
+
"eval_accuracy": 0.9275,
|
| 1830 |
+
"eval_loss": 0.20720510184764862,
|
| 1831 |
+
"eval_runtime": 5.4406,
|
| 1832 |
+
"eval_samples_per_second": 147.044,
|
| 1833 |
+
"eval_steps_per_second": 18.38,
|
| 1834 |
+
"step": 2268
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 27.023809523809526,
|
| 1838 |
+
"grad_norm": 7.463446140289307,
|
| 1839 |
+
"learning_rate": 2.239858906525573e-06,
|
| 1840 |
+
"loss": 0.1377,
|
| 1841 |
+
"step": 2270
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 27.142857142857142,
|
| 1845 |
+
"grad_norm": 11.728429794311523,
|
| 1846 |
+
"learning_rate": 2.1516754850088184e-06,
|
| 1847 |
+
"loss": 0.1315,
|
| 1848 |
+
"step": 2280
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 27.261904761904763,
|
| 1852 |
+
"grad_norm": 4.965158939361572,
|
| 1853 |
+
"learning_rate": 2.0634920634920634e-06,
|
| 1854 |
+
"loss": 0.1466,
|
| 1855 |
+
"step": 2290
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 27.38095238095238,
|
| 1859 |
+
"grad_norm": 5.408288955688477,
|
| 1860 |
+
"learning_rate": 1.9753086419753087e-06,
|
| 1861 |
+
"loss": 0.1418,
|
| 1862 |
+
"step": 2300
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 27.5,
|
| 1866 |
+
"grad_norm": 5.822543621063232,
|
| 1867 |
+
"learning_rate": 1.887125220458554e-06,
|
| 1868 |
+
"loss": 0.1189,
|
| 1869 |
+
"step": 2310
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 27.61904761904762,
|
| 1873 |
+
"grad_norm": 16.357952117919922,
|
| 1874 |
+
"learning_rate": 1.798941798941799e-06,
|
| 1875 |
+
"loss": 0.1595,
|
| 1876 |
+
"step": 2320
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 27.738095238095237,
|
| 1880 |
+
"grad_norm": 7.142815589904785,
|
| 1881 |
+
"learning_rate": 1.7107583774250442e-06,
|
| 1882 |
+
"loss": 0.1362,
|
| 1883 |
+
"step": 2330
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 27.857142857142858,
|
| 1887 |
+
"grad_norm": 3.4048264026641846,
|
| 1888 |
+
"learning_rate": 1.6225749559082893e-06,
|
| 1889 |
+
"loss": 0.1226,
|
| 1890 |
+
"step": 2340
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 27.976190476190474,
|
| 1894 |
+
"grad_norm": 3.4568896293640137,
|
| 1895 |
+
"learning_rate": 1.5343915343915345e-06,
|
| 1896 |
+
"loss": 0.1472,
|
| 1897 |
+
"step": 2350
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 28.0,
|
| 1901 |
+
"eval_accuracy": 0.93375,
|
| 1902 |
+
"eval_loss": 0.2012936919927597,
|
| 1903 |
+
"eval_runtime": 5.3691,
|
| 1904 |
+
"eval_samples_per_second": 149.0,
|
| 1905 |
+
"eval_steps_per_second": 18.625,
|
| 1906 |
+
"step": 2352
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 28.095238095238095,
|
| 1910 |
+
"grad_norm": 25.373470306396484,
|
| 1911 |
+
"learning_rate": 1.4462081128747796e-06,
|
| 1912 |
+
"loss": 0.1368,
|
| 1913 |
+
"step": 2360
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 28.214285714285715,
|
| 1917 |
+
"grad_norm": 14.415323257446289,
|
| 1918 |
+
"learning_rate": 1.3580246913580248e-06,
|
| 1919 |
+
"loss": 0.145,
|
| 1920 |
+
"step": 2370
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 28.333333333333332,
|
| 1924 |
+
"grad_norm": 9.555523872375488,
|
| 1925 |
+
"learning_rate": 1.26984126984127e-06,
|
| 1926 |
+
"loss": 0.1556,
|
| 1927 |
+
"step": 2380
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 28.452380952380953,
|
| 1931 |
+
"grad_norm": 13.204785346984863,
|
| 1932 |
+
"learning_rate": 1.181657848324515e-06,
|
| 1933 |
+
"loss": 0.1495,
|
| 1934 |
+
"step": 2390
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 28.571428571428573,
|
| 1938 |
+
"grad_norm": 13.89194393157959,
|
| 1939 |
+
"learning_rate": 1.0934744268077602e-06,
|
| 1940 |
+
"loss": 0.1221,
|
| 1941 |
+
"step": 2400
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 28.69047619047619,
|
| 1945 |
+
"grad_norm": 21.069202423095703,
|
| 1946 |
+
"learning_rate": 1.0052910052910054e-06,
|
| 1947 |
+
"loss": 0.1376,
|
| 1948 |
+
"step": 2410
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 28.80952380952381,
|
| 1952 |
+
"grad_norm": 13.420437812805176,
|
| 1953 |
+
"learning_rate": 9.171075837742504e-07,
|
| 1954 |
+
"loss": 0.135,
|
| 1955 |
+
"step": 2420
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 28.928571428571427,
|
| 1959 |
+
"grad_norm": 6.938393592834473,
|
| 1960 |
+
"learning_rate": 8.289241622574956e-07,
|
| 1961 |
+
"loss": 0.1329,
|
| 1962 |
+
"step": 2430
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 29.0,
|
| 1966 |
+
"eval_accuracy": 0.9375,
|
| 1967 |
+
"eval_loss": 0.19804717600345612,
|
| 1968 |
+
"eval_runtime": 5.2492,
|
| 1969 |
+
"eval_samples_per_second": 152.406,
|
| 1970 |
+
"eval_steps_per_second": 19.051,
|
| 1971 |
+
"step": 2436
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 29.047619047619047,
|
| 1975 |
+
"grad_norm": 6.732052803039551,
|
| 1976 |
+
"learning_rate": 7.407407407407407e-07,
|
| 1977 |
+
"loss": 0.1373,
|
| 1978 |
+
"step": 2440
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 29.166666666666668,
|
| 1982 |
+
"grad_norm": 11.9608154296875,
|
| 1983 |
+
"learning_rate": 6.525573192239859e-07,
|
| 1984 |
+
"loss": 0.1288,
|
| 1985 |
+
"step": 2450
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 29.285714285714285,
|
| 1989 |
+
"grad_norm": 9.937846183776855,
|
| 1990 |
+
"learning_rate": 5.64373897707231e-07,
|
| 1991 |
+
"loss": 0.1267,
|
| 1992 |
+
"step": 2460
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 29.404761904761905,
|
| 1996 |
+
"grad_norm": 6.138334274291992,
|
| 1997 |
+
"learning_rate": 4.7619047619047623e-07,
|
| 1998 |
+
"loss": 0.1358,
|
| 1999 |
+
"step": 2470
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 29.523809523809526,
|
| 2003 |
+
"grad_norm": 5.230373859405518,
|
| 2004 |
+
"learning_rate": 3.880070546737214e-07,
|
| 2005 |
+
"loss": 0.1473,
|
| 2006 |
+
"step": 2480
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 29.642857142857142,
|
| 2010 |
+
"grad_norm": 8.851503372192383,
|
| 2011 |
+
"learning_rate": 2.9982363315696647e-07,
|
| 2012 |
+
"loss": 0.1507,
|
| 2013 |
+
"step": 2490
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 29.761904761904763,
|
| 2017 |
+
"grad_norm": 5.106894493103027,
|
| 2018 |
+
"learning_rate": 2.1164021164021165e-07,
|
| 2019 |
+
"loss": 0.1262,
|
| 2020 |
+
"step": 2500
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 29.88095238095238,
|
| 2024 |
+
"grad_norm": 9.42648696899414,
|
| 2025 |
+
"learning_rate": 1.234567901234568e-07,
|
| 2026 |
+
"loss": 0.1397,
|
| 2027 |
+
"step": 2510
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 30.0,
|
| 2031 |
+
"grad_norm": 8.953765869140625,
|
| 2032 |
+
"learning_rate": 3.527336860670194e-08,
|
| 2033 |
+
"loss": 0.132,
|
| 2034 |
+
"step": 2520
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 30.0,
|
| 2038 |
+
"eval_accuracy": 0.93625,
|
| 2039 |
+
"eval_loss": 0.1968877911567688,
|
| 2040 |
+
"eval_runtime": 5.803,
|
| 2041 |
+
"eval_samples_per_second": 137.859,
|
| 2042 |
+
"eval_steps_per_second": 17.232,
|
| 2043 |
+
"step": 2520
|
| 2044 |
+
}
|
| 2045 |
+
],
|
| 2046 |
+
"logging_steps": 10,
|
| 2047 |
+
"max_steps": 2520,
|
| 2048 |
+
"num_input_tokens_seen": 0,
|
| 2049 |
+
"num_train_epochs": 30,
|
| 2050 |
+
"save_steps": 500,
|
| 2051 |
+
"stateful_callbacks": {
|
| 2052 |
+
"TrainerControl": {
|
| 2053 |
+
"args": {
|
| 2054 |
+
"should_epoch_stop": false,
|
| 2055 |
+
"should_evaluate": false,
|
| 2056 |
+
"should_log": false,
|
| 2057 |
+
"should_save": true,
|
| 2058 |
+
"should_training_stop": true
|
| 2059 |
+
},
|
| 2060 |
+
"attributes": {}
|
| 2061 |
+
}
|
| 2062 |
+
},
|
| 2063 |
+
"total_flos": 1.019477021294592e+19,
|
| 2064 |
+
"train_batch_size": 192,
|
| 2065 |
+
"trial_name": null,
|
| 2066 |
+
"trial_params": null
|
| 2067 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d043779ba98c244cf21fe965ab08cc6f63a07c2d49949c74e46624c71feca47f
|
| 3 |
+
size 5112
|