Hugging Face's logo

projectsidewalk
/
sidewalk-validator-ai-nocurbramp

Safetensors
dinov2
Model card Files
xet
Community
johnomeara commited on
Commit
83f1a09
·
verified ·
1 Parent(s): 9fa0f7e

Upload 8 files

Browse files
Files changed (8) hide show
  1. config.json +69 -0
  2. model.safetensors +3 -0
  3. optimizer.pt +3 -0
  4. preprocessor_config.json +27 -0
  5. rng_state.pth +3 -0
  6. scheduler.pt +3 -0
  7. trainer_state.json +3444 -0
  8. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/dinov2-large",
3
+ "apply_layernorm": true,
4
+ "architectures": [
5
+ "Dinov2ForImageClassification"
6
+ ],
7
+ "attention_probs_dropout_prob": 0.0,
8
+ "drop_path_rate": 0.0,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.0,
11
+ "hidden_size": 1024,
12
+ "id2label": {
13
+ "0": "correct",
14
+ "1": "incorrect"
15
+ },
16
+ "image_size": 518,
17
+ "initializer_range": 0.02,
18
+ "label2id": {
19
+ "correct": 0,
20
+ "incorrect": 1
21
+ },
22
+ "layer_norm_eps": 1e-06,
23
+ "layerscale_value": 1.0,
24
+ "mlp_ratio": 4,
25
+ "model_type": "dinov2",
26
+ "num_attention_heads": 16,
27
+ "num_channels": 3,
28
+ "num_hidden_layers": 24,
29
+ "out_features": [
30
+ "stage24"
31
+ ],
32
+ "out_indices": [
33
+ 24
34
+ ],
35
+ "patch_size": 14,
36
+ "problem_type": "single_label_classification",
37
+ "qkv_bias": true,
38
+ "reshape_hidden_states": true,
39
+ "stage_names": [
40
+ "stem",
41
+ "stage1",
42
+ "stage2",
43
+ "stage3",
44
+ "stage4",
45
+ "stage5",
46
+ "stage6",
47
+ "stage7",
48
+ "stage8",
49
+ "stage9",
50
+ "stage10",
51
+ "stage11",
52
+ "stage12",
53
+ "stage13",
54
+ "stage14",
55
+ "stage15",
56
+ "stage16",
57
+ "stage17",
58
+ "stage18",
59
+ "stage19",
60
+ "stage20",
61
+ "stage21",
62
+ "stage22",
63
+ "stage23",
64
+ "stage24"
65
+ ],
66
+ "torch_dtype": "float32",
67
+ "transformers_version": "4.47.1",
68
+ "use_swiglu_ffn": false
69
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02bff91182744b33217521cf6c773b0d524b08bdd6b67ca889d3ba1048fba441
3
+ size 1217542512
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1c10067ae4534fa78de4b5acfb43e085e1b1161fcf6d8e27c62f67c851fd050
3
+ size 2435341946
preprocessor_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 224,
4
+ "width": 224
5
+ },
6
+ "do_center_crop": true,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.485,
13
+ 0.456,
14
+ 0.406
15
+ ],
16
+ "image_processor_type": "BitImageProcessor",
17
+ "image_std": [
18
+ 0.229,
19
+ 0.224,
20
+ 0.225
21
+ ],
22
+ "resample": 3,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "shortest_edge": 256
26
+ }
27
+ }
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74044c1f0a8fb84fcc4673a98e2f3c18bb67ff0e1853b3dbfd2eae10cf7a8aea
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e523ea7272df54e091453fa5c2252ed849e2f5af01fc55f6d4453fdb4c1f28
3
+ size 1064
trainer_state.json ADDED
@@ -0,0 +1,3444 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8567568925713644,
3
+ "best_model_checkpoint": "NoCurbRamp/dinov2/checkpoint-4598",
4
+ "epoch": 22.0,
5
+ "eval_steps": 500,
6
+ "global_step": 4598,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04790419161676647,
13
+ "grad_norm": 102.78497314453125,
14
+ "learning_rate": 1.9230769230769234e-07,
15
+ "loss": 3.3285,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.09580838323353294,
20
+ "grad_norm": 108.06490325927734,
21
+ "learning_rate": 3.846153846153847e-07,
22
+ "loss": 3.3413,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.1437125748502994,
27
+ "grad_norm": 168.06533813476562,
28
+ "learning_rate": 5.76923076923077e-07,
29
+ "loss": 3.1682,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.19161676646706588,
34
+ "grad_norm": 107.77788543701172,
35
+ "learning_rate": 7.692307692307694e-07,
36
+ "loss": 2.9011,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.23952095808383234,
41
+ "grad_norm": 73.95760345458984,
42
+ "learning_rate": 9.615384615384617e-07,
43
+ "loss": 2.5857,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.2874251497005988,
48
+ "grad_norm": 105.05018615722656,
49
+ "learning_rate": 1.153846153846154e-06,
50
+ "loss": 2.5002,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.33532934131736525,
55
+ "grad_norm": 674.4759521484375,
56
+ "learning_rate": 1.3461538461538462e-06,
57
+ "loss": 2.4689,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.38323353293413176,
62
+ "grad_norm": 141.63890075683594,
63
+ "learning_rate": 1.5384615384615387e-06,
64
+ "loss": 2.3887,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.4311377245508982,
69
+ "grad_norm": 160.6332550048828,
70
+ "learning_rate": 1.7307692307692308e-06,
71
+ "loss": 2.2373,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.47904191616766467,
76
+ "grad_norm": 702.2935180664062,
77
+ "learning_rate": 1.9230769230769234e-06,
78
+ "loss": 2.1801,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.5269461077844312,
83
+ "grad_norm": 86.12798309326172,
84
+ "learning_rate": 2.1153846153846155e-06,
85
+ "loss": 2.0948,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.5748502994011976,
90
+ "grad_norm": 112.54278564453125,
91
+ "learning_rate": 2.307692307692308e-06,
92
+ "loss": 2.0035,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.6227544910179641,
97
+ "grad_norm": 187.13577270507812,
98
+ "learning_rate": 2.5e-06,
99
+ "loss": 2.13,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.6706586826347305,
104
+ "grad_norm": 74.73124694824219,
105
+ "learning_rate": 2.6923076923076923e-06,
106
+ "loss": 2.0507,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.718562874251497,
111
+ "grad_norm": 111.90137481689453,
112
+ "learning_rate": 2.8846153846153845e-06,
113
+ "loss": 1.9312,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.7664670658682635,
118
+ "grad_norm": 162.74754333496094,
119
+ "learning_rate": 3.0769230769230774e-06,
120
+ "loss": 2.3011,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.8143712574850299,
125
+ "grad_norm": 104.16117095947266,
126
+ "learning_rate": 3.2692307692307696e-06,
127
+ "loss": 1.8005,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.8622754491017964,
132
+ "grad_norm": 97.45256805419922,
133
+ "learning_rate": 3.4615384615384617e-06,
134
+ "loss": 1.8074,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.9101796407185628,
139
+ "grad_norm": 253.20359802246094,
140
+ "learning_rate": 3.653846153846154e-06,
141
+ "loss": 1.985,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.9580838323353293,
146
+ "grad_norm": 157.43246459960938,
147
+ "learning_rate": 3.846153846153847e-06,
148
+ "loss": 1.7699,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 1.0,
153
+ "eval_loss": 0.4299589693546295,
154
+ "eval_macro_f1": 0.7858166164735507,
155
+ "eval_runtime": 81.1893,
156
+ "eval_samples_per_second": 20.582,
157
+ "eval_steps_per_second": 2.574,
158
+ "step": 209
159
+ },
160
+ {
161
+ "epoch": 1.0047904191616766,
162
+ "grad_norm": 79.56900787353516,
163
+ "learning_rate": 4.0384615384615385e-06,
164
+ "loss": 2.0621,
165
+ "step": 210
166
+ },
167
+ {
168
+ "epoch": 1.0526946107784432,
169
+ "grad_norm": 80.78365325927734,
170
+ "learning_rate": 4.230769230769231e-06,
171
+ "loss": 1.9002,
172
+ "step": 220
173
+ },
174
+ {
175
+ "epoch": 1.1005988023952096,
176
+ "grad_norm": 131.77723693847656,
177
+ "learning_rate": 4.423076923076924e-06,
178
+ "loss": 1.8801,
179
+ "step": 230
180
+ },
181
+ {
182
+ "epoch": 1.148502994011976,
183
+ "grad_norm": 59.51331329345703,
184
+ "learning_rate": 4.615384615384616e-06,
185
+ "loss": 1.6218,
186
+ "step": 240
187
+ },
188
+ {
189
+ "epoch": 1.1964071856287426,
190
+ "grad_norm": 59.175621032714844,
191
+ "learning_rate": 4.807692307692308e-06,
192
+ "loss": 1.6202,
193
+ "step": 250
194
+ },
195
+ {
196
+ "epoch": 1.244311377245509,
197
+ "grad_norm": 175.007568359375,
198
+ "learning_rate": 5e-06,
199
+ "loss": 1.5911,
200
+ "step": 260
201
+ },
202
+ {
203
+ "epoch": 1.2922155688622754,
204
+ "grad_norm": 112.08436584472656,
205
+ "learning_rate": 5.192307692307693e-06,
206
+ "loss": 1.8472,
207
+ "step": 270
208
+ },
209
+ {
210
+ "epoch": 1.340119760479042,
211
+ "grad_norm": 49.89094924926758,
212
+ "learning_rate": 5.384615384615385e-06,
213
+ "loss": 1.7157,
214
+ "step": 280
215
+ },
216
+ {
217
+ "epoch": 1.3880239520958084,
218
+ "grad_norm": 105.93860626220703,
219
+ "learning_rate": 5.576923076923077e-06,
220
+ "loss": 1.7682,
221
+ "step": 290
222
+ },
223
+ {
224
+ "epoch": 1.4359281437125748,
225
+ "grad_norm": 73.6728515625,
226
+ "learning_rate": 5.769230769230769e-06,
227
+ "loss": 1.6694,
228
+ "step": 300
229
+ },
230
+ {
231
+ "epoch": 1.4838323353293412,
232
+ "grad_norm": 88.21549224853516,
233
+ "learning_rate": 5.961538461538462e-06,
234
+ "loss": 1.437,
235
+ "step": 310
236
+ },
237
+ {
238
+ "epoch": 1.5317365269461076,
239
+ "grad_norm": 95.41169738769531,
240
+ "learning_rate": 6.153846153846155e-06,
241
+ "loss": 1.6035,
242
+ "step": 320
243
+ },
244
+ {
245
+ "epoch": 1.5796407185628742,
246
+ "grad_norm": 55.19960021972656,
247
+ "learning_rate": 6.3461538461538466e-06,
248
+ "loss": 1.6455,
249
+ "step": 330
250
+ },
251
+ {
252
+ "epoch": 1.6275449101796409,
253
+ "grad_norm": 57.696773529052734,
254
+ "learning_rate": 6.538461538461539e-06,
255
+ "loss": 1.7315,
256
+ "step": 340
257
+ },
258
+ {
259
+ "epoch": 1.6754491017964073,
260
+ "grad_norm": 241.8914337158203,
261
+ "learning_rate": 6.730769230769232e-06,
262
+ "loss": 1.8734,
263
+ "step": 350
264
+ },
265
+ {
266
+ "epoch": 1.7233532934131737,
267
+ "grad_norm": 129.99835205078125,
268
+ "learning_rate": 6.923076923076923e-06,
269
+ "loss": 1.7628,
270
+ "step": 360
271
+ },
272
+ {
273
+ "epoch": 1.77125748502994,
274
+ "grad_norm": 125.50336456298828,
275
+ "learning_rate": 7.115384615384616e-06,
276
+ "loss": 1.5134,
277
+ "step": 370
278
+ },
279
+ {
280
+ "epoch": 1.8191616766467065,
281
+ "grad_norm": 66.20431518554688,
282
+ "learning_rate": 7.307692307692308e-06,
283
+ "loss": 1.7175,
284
+ "step": 380
285
+ },
286
+ {
287
+ "epoch": 1.867065868263473,
288
+ "grad_norm": 45.693199157714844,
289
+ "learning_rate": 7.500000000000001e-06,
290
+ "loss": 1.5907,
291
+ "step": 390
292
+ },
293
+ {
294
+ "epoch": 1.9149700598802395,
295
+ "grad_norm": 57.71452713012695,
296
+ "learning_rate": 7.692307692307694e-06,
297
+ "loss": 1.6598,
298
+ "step": 400
299
+ },
300
+ {
301
+ "epoch": 1.9628742514970061,
302
+ "grad_norm": 87.35894012451172,
303
+ "learning_rate": 7.884615384615384e-06,
304
+ "loss": 1.6557,
305
+ "step": 410
306
+ },
307
+ {
308
+ "epoch": 2.0,
309
+ "eval_loss": 0.41893911361694336,
310
+ "eval_macro_f1": 0.8138112665329834,
311
+ "eval_runtime": 41.195,
312
+ "eval_samples_per_second": 40.563,
313
+ "eval_steps_per_second": 5.073,
314
+ "step": 418
315
+ },
316
+ {
317
+ "epoch": 2.009580838323353,
318
+ "grad_norm": 71.44452667236328,
319
+ "learning_rate": 8.076923076923077e-06,
320
+ "loss": 1.7133,
321
+ "step": 420
322
+ },
323
+ {
324
+ "epoch": 2.05748502994012,
325
+ "grad_norm": 53.76054000854492,
326
+ "learning_rate": 8.26923076923077e-06,
327
+ "loss": 1.648,
328
+ "step": 430
329
+ },
330
+ {
331
+ "epoch": 2.1053892215568863,
332
+ "grad_norm": 1027.304931640625,
333
+ "learning_rate": 8.461538461538462e-06,
334
+ "loss": 1.7138,
335
+ "step": 440
336
+ },
337
+ {
338
+ "epoch": 2.1532934131736527,
339
+ "grad_norm": 226.78587341308594,
340
+ "learning_rate": 8.653846153846155e-06,
341
+ "loss": 1.589,
342
+ "step": 450
343
+ },
344
+ {
345
+ "epoch": 2.201197604790419,
346
+ "grad_norm": 83.36427307128906,
347
+ "learning_rate": 8.846153846153847e-06,
348
+ "loss": 1.4567,
349
+ "step": 460
350
+ },
351
+ {
352
+ "epoch": 2.2491017964071855,
353
+ "grad_norm": 56.77524185180664,
354
+ "learning_rate": 9.03846153846154e-06,
355
+ "loss": 1.6047,
356
+ "step": 470
357
+ },
358
+ {
359
+ "epoch": 2.297005988023952,
360
+ "grad_norm": 108.68399047851562,
361
+ "learning_rate": 9.230769230769232e-06,
362
+ "loss": 1.9924,
363
+ "step": 480
364
+ },
365
+ {
366
+ "epoch": 2.344910179640719,
367
+ "grad_norm": 82.6905288696289,
368
+ "learning_rate": 9.423076923076923e-06,
369
+ "loss": 2.0933,
370
+ "step": 490
371
+ },
372
+ {
373
+ "epoch": 2.392814371257485,
374
+ "grad_norm": 29.161964416503906,
375
+ "learning_rate": 9.615384615384616e-06,
376
+ "loss": 1.605,
377
+ "step": 500
378
+ },
379
+ {
380
+ "epoch": 2.4407185628742516,
381
+ "grad_norm": 37.83173370361328,
382
+ "learning_rate": 9.807692307692308e-06,
383
+ "loss": 1.7725,
384
+ "step": 510
385
+ },
386
+ {
387
+ "epoch": 2.488622754491018,
388
+ "grad_norm": 49.54322052001953,
389
+ "learning_rate": 1e-05,
390
+ "loss": 1.4123,
391
+ "step": 520
392
+ },
393
+ {
394
+ "epoch": 2.5365269461077844,
395
+ "grad_norm": 67.74500274658203,
396
+ "learning_rate": 9.97863247863248e-06,
397
+ "loss": 1.5403,
398
+ "step": 530
399
+ },
400
+ {
401
+ "epoch": 2.584431137724551,
402
+ "grad_norm": 55.931190490722656,
403
+ "learning_rate": 9.957264957264958e-06,
404
+ "loss": 1.4283,
405
+ "step": 540
406
+ },
407
+ {
408
+ "epoch": 2.632335329341317,
409
+ "grad_norm": 47.531734466552734,
410
+ "learning_rate": 9.935897435897437e-06,
411
+ "loss": 1.4491,
412
+ "step": 550
413
+ },
414
+ {
415
+ "epoch": 2.680239520958084,
416
+ "grad_norm": 30.019596099853516,
417
+ "learning_rate": 9.914529914529915e-06,
418
+ "loss": 1.4842,
419
+ "step": 560
420
+ },
421
+ {
422
+ "epoch": 2.7281437125748504,
423
+ "grad_norm": 67.44615173339844,
424
+ "learning_rate": 9.893162393162394e-06,
425
+ "loss": 1.6565,
426
+ "step": 570
427
+ },
428
+ {
429
+ "epoch": 2.776047904191617,
430
+ "grad_norm": 54.778350830078125,
431
+ "learning_rate": 9.871794871794872e-06,
432
+ "loss": 1.4563,
433
+ "step": 580
434
+ },
435
+ {
436
+ "epoch": 2.8239520958083832,
437
+ "grad_norm": 54.04096221923828,
438
+ "learning_rate": 9.850427350427351e-06,
439
+ "loss": 1.5794,
440
+ "step": 590
441
+ },
442
+ {
443
+ "epoch": 2.8718562874251496,
444
+ "grad_norm": 35.025691986083984,
445
+ "learning_rate": 9.82905982905983e-06,
446
+ "loss": 1.475,
447
+ "step": 600
448
+ },
449
+ {
450
+ "epoch": 2.919760479041916,
451
+ "grad_norm": 94.27930450439453,
452
+ "learning_rate": 9.807692307692308e-06,
453
+ "loss": 1.7062,
454
+ "step": 610
455
+ },
456
+ {
457
+ "epoch": 2.9676646706586824,
458
+ "grad_norm": 83.62496185302734,
459
+ "learning_rate": 9.786324786324787e-06,
460
+ "loss": 1.3654,
461
+ "step": 620
462
+ },
463
+ {
464
+ "epoch": 3.0,
465
+ "eval_loss": 0.4371185898780823,
466
+ "eval_macro_f1": 0.7854326010045019,
467
+ "eval_runtime": 41.3851,
468
+ "eval_samples_per_second": 40.377,
469
+ "eval_steps_per_second": 5.05,
470
+ "step": 627
471
+ },
472
+ {
473
+ "epoch": 3.01437125748503,
474
+ "grad_norm": 270.2382507324219,
475
+ "learning_rate": 9.764957264957265e-06,
476
+ "loss": 1.4214,
477
+ "step": 630
478
+ },
479
+ {
480
+ "epoch": 3.0622754491017963,
481
+ "grad_norm": 39.00271224975586,
482
+ "learning_rate": 9.743589743589744e-06,
483
+ "loss": 1.4744,
484
+ "step": 640
485
+ },
486
+ {
487
+ "epoch": 3.1101796407185627,
488
+ "grad_norm": 32.14289474487305,
489
+ "learning_rate": 9.722222222222223e-06,
490
+ "loss": 1.6232,
491
+ "step": 650
492
+ },
493
+ {
494
+ "epoch": 3.1580838323353295,
495
+ "grad_norm": 39.31702423095703,
496
+ "learning_rate": 9.700854700854701e-06,
497
+ "loss": 1.4988,
498
+ "step": 660
499
+ },
500
+ {
501
+ "epoch": 3.205988023952096,
502
+ "grad_norm": 65.2395248413086,
503
+ "learning_rate": 9.67948717948718e-06,
504
+ "loss": 1.26,
505
+ "step": 670
506
+ },
507
+ {
508
+ "epoch": 3.2538922155688623,
509
+ "grad_norm": 57.717525482177734,
510
+ "learning_rate": 9.658119658119659e-06,
511
+ "loss": 1.331,
512
+ "step": 680
513
+ },
514
+ {
515
+ "epoch": 3.3017964071856287,
516
+ "grad_norm": 69.02164459228516,
517
+ "learning_rate": 9.636752136752137e-06,
518
+ "loss": 1.3799,
519
+ "step": 690
520
+ },
521
+ {
522
+ "epoch": 3.349700598802395,
523
+ "grad_norm": 60.07762145996094,
524
+ "learning_rate": 9.615384615384616e-06,
525
+ "loss": 1.6609,
526
+ "step": 700
527
+ },
528
+ {
529
+ "epoch": 3.3976047904191615,
530
+ "grad_norm": 44.97191619873047,
531
+ "learning_rate": 9.594017094017094e-06,
532
+ "loss": 1.4233,
533
+ "step": 710
534
+ },
535
+ {
536
+ "epoch": 3.4455089820359284,
537
+ "grad_norm": 65.7463150024414,
538
+ "learning_rate": 9.572649572649575e-06,
539
+ "loss": 1.3724,
540
+ "step": 720
541
+ },
542
+ {
543
+ "epoch": 3.4934131736526948,
544
+ "grad_norm": 57.15474319458008,
545
+ "learning_rate": 9.551282051282053e-06,
546
+ "loss": 1.6036,
547
+ "step": 730
548
+ },
549
+ {
550
+ "epoch": 3.541317365269461,
551
+ "grad_norm": 64.17628479003906,
552
+ "learning_rate": 9.52991452991453e-06,
553
+ "loss": 1.3152,
554
+ "step": 740
555
+ },
556
+ {
557
+ "epoch": 3.5892215568862276,
558
+ "grad_norm": 69.88690948486328,
559
+ "learning_rate": 9.508547008547009e-06,
560
+ "loss": 1.4094,
561
+ "step": 750
562
+ },
563
+ {
564
+ "epoch": 3.637125748502994,
565
+ "grad_norm": 98.0072021484375,
566
+ "learning_rate": 9.487179487179487e-06,
567
+ "loss": 1.441,
568
+ "step": 760
569
+ },
570
+ {
571
+ "epoch": 3.6850299401197604,
572
+ "grad_norm": 47.96156692504883,
573
+ "learning_rate": 9.465811965811966e-06,
574
+ "loss": 1.7361,
575
+ "step": 770
576
+ },
577
+ {
578
+ "epoch": 3.7329341317365268,
579
+ "grad_norm": 52.17365646362305,
580
+ "learning_rate": 9.444444444444445e-06,
581
+ "loss": 1.1775,
582
+ "step": 780
583
+ },
584
+ {
585
+ "epoch": 3.7808383233532936,
586
+ "grad_norm": 552.5225219726562,
587
+ "learning_rate": 9.423076923076923e-06,
588
+ "loss": 1.2409,
589
+ "step": 790
590
+ },
591
+ {
592
+ "epoch": 3.82874251497006,
593
+ "grad_norm": 34.344207763671875,
594
+ "learning_rate": 9.401709401709402e-06,
595
+ "loss": 1.4177,
596
+ "step": 800
597
+ },
598
+ {
599
+ "epoch": 3.8766467065868264,
600
+ "grad_norm": 112.69720458984375,
601
+ "learning_rate": 9.38034188034188e-06,
602
+ "loss": 1.4948,
603
+ "step": 810
604
+ },
605
+ {
606
+ "epoch": 3.924550898203593,
607
+ "grad_norm": 94.27716064453125,
608
+ "learning_rate": 9.358974358974359e-06,
609
+ "loss": 1.5927,
610
+ "step": 820
611
+ },
612
+ {
613
+ "epoch": 3.972455089820359,
614
+ "grad_norm": 48.06044387817383,
615
+ "learning_rate": 9.33760683760684e-06,
616
+ "loss": 1.4711,
617
+ "step": 830
618
+ },
619
+ {
620
+ "epoch": 4.0,
621
+ "eval_loss": 0.39647647738456726,
622
+ "eval_macro_f1": 0.8267623951048628,
623
+ "eval_runtime": 44.3505,
624
+ "eval_samples_per_second": 37.677,
625
+ "eval_steps_per_second": 4.712,
626
+ "step": 836
627
+ },
628
+ {
629
+ "epoch": 4.019161676646706,
630
+ "grad_norm": 130.36834716796875,
631
+ "learning_rate": 9.316239316239318e-06,
632
+ "loss": 1.8227,
633
+ "step": 840
634
+ },
635
+ {
636
+ "epoch": 4.067065868263473,
637
+ "grad_norm": 52.67673110961914,
638
+ "learning_rate": 9.294871794871796e-06,
639
+ "loss": 1.2007,
640
+ "step": 850
641
+ },
642
+ {
643
+ "epoch": 4.11497005988024,
644
+ "grad_norm": 38.394187927246094,
645
+ "learning_rate": 9.273504273504275e-06,
646
+ "loss": 1.2584,
647
+ "step": 860
648
+ },
649
+ {
650
+ "epoch": 4.162874251497006,
651
+ "grad_norm": 293.8218994140625,
652
+ "learning_rate": 9.252136752136754e-06,
653
+ "loss": 1.1846,
654
+ "step": 870
655
+ },
656
+ {
657
+ "epoch": 4.210778443113773,
658
+ "grad_norm": 85.88053894042969,
659
+ "learning_rate": 9.230769230769232e-06,
660
+ "loss": 1.3184,
661
+ "step": 880
662
+ },
663
+ {
664
+ "epoch": 4.258682634730539,
665
+ "grad_norm": 45.364723205566406,
666
+ "learning_rate": 9.20940170940171e-06,
667
+ "loss": 1.2298,
668
+ "step": 890
669
+ },
670
+ {
671
+ "epoch": 4.3065868263473055,
672
+ "grad_norm": 43.60586929321289,
673
+ "learning_rate": 9.188034188034188e-06,
674
+ "loss": 1.1389,
675
+ "step": 900
676
+ },
677
+ {
678
+ "epoch": 4.3544910179640715,
679
+ "grad_norm": 85.17373657226562,
680
+ "learning_rate": 9.166666666666666e-06,
681
+ "loss": 1.1531,
682
+ "step": 910
683
+ },
684
+ {
685
+ "epoch": 4.402395209580838,
686
+ "grad_norm": 38.943946838378906,
687
+ "learning_rate": 9.145299145299145e-06,
688
+ "loss": 1.2874,
689
+ "step": 920
690
+ },
691
+ {
692
+ "epoch": 4.450299401197605,
693
+ "grad_norm": 29.44234275817871,
694
+ "learning_rate": 9.123931623931624e-06,
695
+ "loss": 1.3225,
696
+ "step": 930
697
+ },
698
+ {
699
+ "epoch": 4.498203592814371,
700
+ "grad_norm": 119.2593002319336,
701
+ "learning_rate": 9.102564102564104e-06,
702
+ "loss": 1.5807,
703
+ "step": 940
704
+ },
705
+ {
706
+ "epoch": 4.546107784431138,
707
+ "grad_norm": 30.237037658691406,
708
+ "learning_rate": 9.081196581196583e-06,
709
+ "loss": 0.9587,
710
+ "step": 950
711
+ },
712
+ {
713
+ "epoch": 4.594011976047904,
714
+ "grad_norm": 100.8051986694336,
715
+ "learning_rate": 9.059829059829061e-06,
716
+ "loss": 1.1353,
717
+ "step": 960
718
+ },
719
+ {
720
+ "epoch": 4.641916167664671,
721
+ "grad_norm": 64.73004150390625,
722
+ "learning_rate": 9.03846153846154e-06,
723
+ "loss": 1.271,
724
+ "step": 970
725
+ },
726
+ {
727
+ "epoch": 4.689820359281438,
728
+ "grad_norm": 55.64018249511719,
729
+ "learning_rate": 9.017094017094018e-06,
730
+ "loss": 1.2463,
731
+ "step": 980
732
+ },
733
+ {
734
+ "epoch": 4.7377245508982035,
735
+ "grad_norm": 32.8985710144043,
736
+ "learning_rate": 8.995726495726497e-06,
737
+ "loss": 1.2084,
738
+ "step": 990
739
+ },
740
+ {
741
+ "epoch": 4.78562874251497,
742
+ "grad_norm": 71.72270202636719,
743
+ "learning_rate": 8.974358974358976e-06,
744
+ "loss": 0.9659,
745
+ "step": 1000
746
+ },
747
+ {
748
+ "epoch": 4.833532934131736,
749
+ "grad_norm": 60.666831970214844,
750
+ "learning_rate": 8.952991452991454e-06,
751
+ "loss": 1.037,
752
+ "step": 1010
753
+ },
754
+ {
755
+ "epoch": 4.881437125748503,
756
+ "grad_norm": 65.69544219970703,
757
+ "learning_rate": 8.931623931623933e-06,
758
+ "loss": 1.4125,
759
+ "step": 1020
760
+ },
761
+ {
762
+ "epoch": 4.929341317365269,
763
+ "grad_norm": 78.90387725830078,
764
+ "learning_rate": 8.910256410256411e-06,
765
+ "loss": 1.1816,
766
+ "step": 1030
767
+ },
768
+ {
769
+ "epoch": 4.977245508982036,
770
+ "grad_norm": 51.65810775756836,
771
+ "learning_rate": 8.888888888888888e-06,
772
+ "loss": 1.1849,
773
+ "step": 1040
774
+ },
775
+ {
776
+ "epoch": 5.0,
777
+ "eval_loss": 0.4067809581756592,
778
+ "eval_macro_f1": 0.838654650788542,
779
+ "eval_runtime": 53.3705,
780
+ "eval_samples_per_second": 31.309,
781
+ "eval_steps_per_second": 3.916,
782
+ "step": 1045
783
+ },
784
+ {
785
+ "epoch": 5.023952095808383,
786
+ "grad_norm": 61.138328552246094,
787
+ "learning_rate": 8.867521367521369e-06,
788
+ "loss": 1.1483,
789
+ "step": 1050
790
+ },
791
+ {
792
+ "epoch": 5.07185628742515,
793
+ "grad_norm": 61.026058197021484,
794
+ "learning_rate": 8.846153846153847e-06,
795
+ "loss": 0.7873,
796
+ "step": 1060
797
+ },
798
+ {
799
+ "epoch": 5.119760479041916,
800
+ "grad_norm": 100.50831604003906,
801
+ "learning_rate": 8.824786324786326e-06,
802
+ "loss": 0.9951,
803
+ "step": 1070
804
+ },
805
+ {
806
+ "epoch": 5.167664670658683,
807
+ "grad_norm": 70.29264068603516,
808
+ "learning_rate": 8.803418803418804e-06,
809
+ "loss": 1.1178,
810
+ "step": 1080
811
+ },
812
+ {
813
+ "epoch": 5.2155688622754495,
814
+ "grad_norm": 62.474613189697266,
815
+ "learning_rate": 8.782051282051283e-06,
816
+ "loss": 1.0041,
817
+ "step": 1090
818
+ },
819
+ {
820
+ "epoch": 5.263473053892215,
821
+ "grad_norm": 32.0721321105957,
822
+ "learning_rate": 8.760683760683762e-06,
823
+ "loss": 1.0825,
824
+ "step": 1100
825
+ },
826
+ {
827
+ "epoch": 5.311377245508982,
828
+ "grad_norm": 78.72416687011719,
829
+ "learning_rate": 8.73931623931624e-06,
830
+ "loss": 1.2565,
831
+ "step": 1110
832
+ },
833
+ {
834
+ "epoch": 5.359281437125748,
835
+ "grad_norm": 119.30426788330078,
836
+ "learning_rate": 8.717948717948719e-06,
837
+ "loss": 1.3718,
838
+ "step": 1120
839
+ },
840
+ {
841
+ "epoch": 5.407185628742515,
842
+ "grad_norm": 45.799991607666016,
843
+ "learning_rate": 8.696581196581197e-06,
844
+ "loss": 1.2321,
845
+ "step": 1130
846
+ },
847
+ {
848
+ "epoch": 5.455089820359281,
849
+ "grad_norm": 25.331722259521484,
850
+ "learning_rate": 8.675213675213676e-06,
851
+ "loss": 0.9521,
852
+ "step": 1140
853
+ },
854
+ {
855
+ "epoch": 5.502994011976048,
856
+ "grad_norm": 96.21393585205078,
857
+ "learning_rate": 8.653846153846155e-06,
858
+ "loss": 0.9443,
859
+ "step": 1150
860
+ },
861
+ {
862
+ "epoch": 5.550898203592815,
863
+ "grad_norm": 45.29869079589844,
864
+ "learning_rate": 8.632478632478633e-06,
865
+ "loss": 0.8862,
866
+ "step": 1160
867
+ },
868
+ {
869
+ "epoch": 5.598802395209581,
870
+ "grad_norm": 56.898780822753906,
871
+ "learning_rate": 8.611111111111112e-06,
872
+ "loss": 1.0529,
873
+ "step": 1170
874
+ },
875
+ {
876
+ "epoch": 5.6467065868263475,
877
+ "grad_norm": 63.27139663696289,
878
+ "learning_rate": 8.58974358974359e-06,
879
+ "loss": 1.0086,
880
+ "step": 1180
881
+ },
882
+ {
883
+ "epoch": 5.6946107784431135,
884
+ "grad_norm": 64.9346923828125,
885
+ "learning_rate": 8.568376068376069e-06,
886
+ "loss": 1.1777,
887
+ "step": 1190
888
+ },
889
+ {
890
+ "epoch": 5.74251497005988,
891
+ "grad_norm": 34.74249267578125,
892
+ "learning_rate": 8.547008547008548e-06,
893
+ "loss": 1.1921,
894
+ "step": 1200
895
+ },
896
+ {
897
+ "epoch": 5.790419161676647,
898
+ "grad_norm": 85.9405517578125,
899
+ "learning_rate": 8.525641025641026e-06,
900
+ "loss": 1.0274,
901
+ "step": 1210
902
+ },
903
+ {
904
+ "epoch": 5.838323353293413,
905
+ "grad_norm": 51.7358512878418,
906
+ "learning_rate": 8.504273504273505e-06,
907
+ "loss": 1.088,
908
+ "step": 1220
909
+ },
910
+ {
911
+ "epoch": 5.88622754491018,
912
+ "grad_norm": 43.08283615112305,
913
+ "learning_rate": 8.482905982905983e-06,
914
+ "loss": 0.8124,
915
+ "step": 1230
916
+ },
917
+ {
918
+ "epoch": 5.934131736526946,
919
+ "grad_norm": 90.1866455078125,
920
+ "learning_rate": 8.461538461538462e-06,
921
+ "loss": 1.0978,
922
+ "step": 1240
923
+ },
924
+ {
925
+ "epoch": 5.982035928143713,
926
+ "grad_norm": 31.511568069458008,
927
+ "learning_rate": 8.44017094017094e-06,
928
+ "loss": 1.2058,
929
+ "step": 1250
930
+ },
931
+ {
932
+ "epoch": 6.0,
933
+ "eval_loss": 0.40154314041137695,
934
+ "eval_macro_f1": 0.828210643143795,
935
+ "eval_runtime": 65.0826,
936
+ "eval_samples_per_second": 25.675,
937
+ "eval_steps_per_second": 3.211,
938
+ "step": 1254
939
+ },
940
+ {
941
+ "epoch": 6.02874251497006,
942
+ "grad_norm": 24.663558959960938,
943
+ "learning_rate": 8.41880341880342e-06,
944
+ "loss": 0.7289,
945
+ "step": 1260
946
+ },
947
+ {
948
+ "epoch": 6.076646706586827,
949
+ "grad_norm": 32.12767028808594,
950
+ "learning_rate": 8.397435897435898e-06,
951
+ "loss": 0.9549,
952
+ "step": 1270
953
+ },
954
+ {
955
+ "epoch": 6.1245508982035926,
956
+ "grad_norm": 45.44089126586914,
957
+ "learning_rate": 8.376068376068377e-06,
958
+ "loss": 0.8573,
959
+ "step": 1280
960
+ },
961
+ {
962
+ "epoch": 6.172455089820359,
963
+ "grad_norm": 95.96544647216797,
964
+ "learning_rate": 8.354700854700855e-06,
965
+ "loss": 0.7697,
966
+ "step": 1290
967
+ },
968
+ {
969
+ "epoch": 6.220359281437125,
970
+ "grad_norm": 88.79412841796875,
971
+ "learning_rate": 8.333333333333334e-06,
972
+ "loss": 0.9044,
973
+ "step": 1300
974
+ },
975
+ {
976
+ "epoch": 6.268263473053892,
977
+ "grad_norm": 57.126304626464844,
978
+ "learning_rate": 8.311965811965812e-06,
979
+ "loss": 0.7112,
980
+ "step": 1310
981
+ },
982
+ {
983
+ "epoch": 6.316167664670659,
984
+ "grad_norm": 76.69857788085938,
985
+ "learning_rate": 8.290598290598293e-06,
986
+ "loss": 0.7156,
987
+ "step": 1320
988
+ },
989
+ {
990
+ "epoch": 6.364071856287425,
991
+ "grad_norm": 54.02803039550781,
992
+ "learning_rate": 8.26923076923077e-06,
993
+ "loss": 0.8275,
994
+ "step": 1330
995
+ },
996
+ {
997
+ "epoch": 6.411976047904192,
998
+ "grad_norm": 60.05768966674805,
999
+ "learning_rate": 8.247863247863248e-06,
1000
+ "loss": 0.8879,
1001
+ "step": 1340
1002
+ },
1003
+ {
1004
+ "epoch": 6.459880239520958,
1005
+ "grad_norm": 80.0739517211914,
1006
+ "learning_rate": 8.226495726495727e-06,
1007
+ "loss": 0.9576,
1008
+ "step": 1350
1009
+ },
1010
+ {
1011
+ "epoch": 6.507784431137725,
1012
+ "grad_norm": 44.03201675415039,
1013
+ "learning_rate": 8.205128205128205e-06,
1014
+ "loss": 0.996,
1015
+ "step": 1360
1016
+ },
1017
+ {
1018
+ "epoch": 6.5556886227544915,
1019
+ "grad_norm": 54.739044189453125,
1020
+ "learning_rate": 8.183760683760684e-06,
1021
+ "loss": 0.8699,
1022
+ "step": 1370
1023
+ },
1024
+ {
1025
+ "epoch": 6.6035928143712574,
1026
+ "grad_norm": 44.991119384765625,
1027
+ "learning_rate": 8.162393162393163e-06,
1028
+ "loss": 0.894,
1029
+ "step": 1380
1030
+ },
1031
+ {
1032
+ "epoch": 6.651497005988024,
1033
+ "grad_norm": 58.03591537475586,
1034
+ "learning_rate": 8.141025641025641e-06,
1035
+ "loss": 0.9078,
1036
+ "step": 1390
1037
+ },
1038
+ {
1039
+ "epoch": 6.69940119760479,
1040
+ "grad_norm": 51.95082473754883,
1041
+ "learning_rate": 8.11965811965812e-06,
1042
+ "loss": 0.8032,
1043
+ "step": 1400
1044
+ },
1045
+ {
1046
+ "epoch": 6.747305389221557,
1047
+ "grad_norm": 47.06685256958008,
1048
+ "learning_rate": 8.098290598290598e-06,
1049
+ "loss": 0.8379,
1050
+ "step": 1410
1051
+ },
1052
+ {
1053
+ "epoch": 6.795209580838323,
1054
+ "grad_norm": 32.85672378540039,
1055
+ "learning_rate": 8.076923076923077e-06,
1056
+ "loss": 0.6956,
1057
+ "step": 1420
1058
+ },
1059
+ {
1060
+ "epoch": 6.84311377245509,
1061
+ "grad_norm": 72.61941528320312,
1062
+ "learning_rate": 8.055555555555557e-06,
1063
+ "loss": 0.8305,
1064
+ "step": 1430
1065
+ },
1066
+ {
1067
+ "epoch": 6.891017964071857,
1068
+ "grad_norm": 50.65372085571289,
1069
+ "learning_rate": 8.034188034188036e-06,
1070
+ "loss": 1.294,
1071
+ "step": 1440
1072
+ },
1073
+ {
1074
+ "epoch": 6.938922155688623,
1075
+ "grad_norm": 30.090587615966797,
1076
+ "learning_rate": 8.012820512820515e-06,
1077
+ "loss": 0.7265,
1078
+ "step": 1450
1079
+ },
1080
+ {
1081
+ "epoch": 6.9868263473053895,
1082
+ "grad_norm": 61.50442886352539,
1083
+ "learning_rate": 7.991452991452993e-06,
1084
+ "loss": 0.9642,
1085
+ "step": 1460
1086
+ },
1087
+ {
1088
+ "epoch": 7.0,
1089
+ "eval_loss": 0.4302707612514496,
1090
+ "eval_macro_f1": 0.8384768009768009,
1091
+ "eval_runtime": 44.1392,
1092
+ "eval_samples_per_second": 37.857,
1093
+ "eval_steps_per_second": 4.735,
1094
+ "step": 1463
1095
+ },
1096
+ {
1097
+ "epoch": 7.0335329341317365,
1098
+ "grad_norm": 51.26984786987305,
1099
+ "learning_rate": 7.970085470085472e-06,
1100
+ "loss": 0.5057,
1101
+ "step": 1470
1102
+ },
1103
+ {
1104
+ "epoch": 7.081437125748503,
1105
+ "grad_norm": 40.46345138549805,
1106
+ "learning_rate": 7.948717948717949e-06,
1107
+ "loss": 0.6524,
1108
+ "step": 1480
1109
+ },
1110
+ {
1111
+ "epoch": 7.129341317365269,
1112
+ "grad_norm": 86.76705169677734,
1113
+ "learning_rate": 7.927350427350427e-06,
1114
+ "loss": 0.7067,
1115
+ "step": 1490
1116
+ },
1117
+ {
1118
+ "epoch": 7.177245508982036,
1119
+ "grad_norm": 47.121002197265625,
1120
+ "learning_rate": 7.905982905982906e-06,
1121
+ "loss": 0.6437,
1122
+ "step": 1500
1123
+ },
1124
+ {
1125
+ "epoch": 7.225149700598802,
1126
+ "grad_norm": 63.37303924560547,
1127
+ "learning_rate": 7.884615384615384e-06,
1128
+ "loss": 0.7471,
1129
+ "step": 1510
1130
+ },
1131
+ {
1132
+ "epoch": 7.273053892215569,
1133
+ "grad_norm": 47.41765594482422,
1134
+ "learning_rate": 7.863247863247863e-06,
1135
+ "loss": 0.8865,
1136
+ "step": 1520
1137
+ },
1138
+ {
1139
+ "epoch": 7.320958083832335,
1140
+ "grad_norm": 49.318092346191406,
1141
+ "learning_rate": 7.841880341880342e-06,
1142
+ "loss": 0.804,
1143
+ "step": 1530
1144
+ },
1145
+ {
1146
+ "epoch": 7.368862275449102,
1147
+ "grad_norm": 45.841331481933594,
1148
+ "learning_rate": 7.820512820512822e-06,
1149
+ "loss": 0.8108,
1150
+ "step": 1540
1151
+ },
1152
+ {
1153
+ "epoch": 7.416766467065869,
1154
+ "grad_norm": 50.48727035522461,
1155
+ "learning_rate": 7.7991452991453e-06,
1156
+ "loss": 0.8056,
1157
+ "step": 1550
1158
+ },
1159
+ {
1160
+ "epoch": 7.464670658682635,
1161
+ "grad_norm": 32.53761291503906,
1162
+ "learning_rate": 7.77777777777778e-06,
1163
+ "loss": 0.652,
1164
+ "step": 1560
1165
+ },
1166
+ {
1167
+ "epoch": 7.512574850299401,
1168
+ "grad_norm": 62.68672561645508,
1169
+ "learning_rate": 7.756410256410258e-06,
1170
+ "loss": 1.0458,
1171
+ "step": 1570
1172
+ },
1173
+ {
1174
+ "epoch": 7.560479041916167,
1175
+ "grad_norm": 45.77931594848633,
1176
+ "learning_rate": 7.735042735042736e-06,
1177
+ "loss": 0.712,
1178
+ "step": 1580
1179
+ },
1180
+ {
1181
+ "epoch": 7.608383233532934,
1182
+ "grad_norm": 46.453208923339844,
1183
+ "learning_rate": 7.713675213675215e-06,
1184
+ "loss": 0.8081,
1185
+ "step": 1590
1186
+ },
1187
+ {
1188
+ "epoch": 7.656287425149701,
1189
+ "grad_norm": 97.16567993164062,
1190
+ "learning_rate": 7.692307692307694e-06,
1191
+ "loss": 0.7669,
1192
+ "step": 1600
1193
+ },
1194
+ {
1195
+ "epoch": 7.704191616766467,
1196
+ "grad_norm": 40.78181076049805,
1197
+ "learning_rate": 7.670940170940172e-06,
1198
+ "loss": 0.7844,
1199
+ "step": 1610
1200
+ },
1201
+ {
1202
+ "epoch": 7.752095808383234,
1203
+ "grad_norm": 72.332763671875,
1204
+ "learning_rate": 7.649572649572649e-06,
1205
+ "loss": 0.7251,
1206
+ "step": 1620
1207
+ },
1208
+ {
1209
+ "epoch": 7.8,
1210
+ "grad_norm": 61.692543029785156,
1211
+ "learning_rate": 7.6282051282051286e-06,
1212
+ "loss": 0.8466,
1213
+ "step": 1630
1214
+ },
1215
+ {
1216
+ "epoch": 7.847904191616767,
1217
+ "grad_norm": 76.77676391601562,
1218
+ "learning_rate": 7.606837606837607e-06,
1219
+ "loss": 0.8042,
1220
+ "step": 1640
1221
+ },
1222
+ {
1223
+ "epoch": 7.895808383233533,
1224
+ "grad_norm": 29.141651153564453,
1225
+ "learning_rate": 7.585470085470086e-06,
1226
+ "loss": 0.7552,
1227
+ "step": 1650
1228
+ },
1229
+ {
1230
+ "epoch": 7.9437125748502995,
1231
+ "grad_norm": 58.72688293457031,
1232
+ "learning_rate": 7.564102564102564e-06,
1233
+ "loss": 0.5293,
1234
+ "step": 1660
1235
+ },
1236
+ {
1237
+ "epoch": 7.991616766467066,
1238
+ "grad_norm": 52.79533386230469,
1239
+ "learning_rate": 7.542735042735043e-06,
1240
+ "loss": 0.7987,
1241
+ "step": 1670
1242
+ },
1243
+ {
1244
+ "epoch": 8.0,
1245
+ "eval_loss": 0.4819260835647583,
1246
+ "eval_macro_f1": 0.8445069931839906,
1247
+ "eval_runtime": 49.8575,
1248
+ "eval_samples_per_second": 33.516,
1249
+ "eval_steps_per_second": 4.192,
1250
+ "step": 1672
1251
+ },
1252
+ {
1253
+ "epoch": 8.038323353293412,
1254
+ "grad_norm": 15.822467803955078,
1255
+ "learning_rate": 7.521367521367522e-06,
1256
+ "loss": 0.4624,
1257
+ "step": 1680
1258
+ },
1259
+ {
1260
+ "epoch": 8.08622754491018,
1261
+ "grad_norm": 17.05672264099121,
1262
+ "learning_rate": 7.500000000000001e-06,
1263
+ "loss": 0.518,
1264
+ "step": 1690
1265
+ },
1266
+ {
1267
+ "epoch": 8.134131736526946,
1268
+ "grad_norm": 404.2013244628906,
1269
+ "learning_rate": 7.47863247863248e-06,
1270
+ "loss": 0.6137,
1271
+ "step": 1700
1272
+ },
1273
+ {
1274
+ "epoch": 8.182035928143712,
1275
+ "grad_norm": 72.656982421875,
1276
+ "learning_rate": 7.457264957264958e-06,
1277
+ "loss": 0.7263,
1278
+ "step": 1710
1279
+ },
1280
+ {
1281
+ "epoch": 8.22994011976048,
1282
+ "grad_norm": 42.37464141845703,
1283
+ "learning_rate": 7.435897435897437e-06,
1284
+ "loss": 0.5209,
1285
+ "step": 1720
1286
+ },
1287
+ {
1288
+ "epoch": 8.277844311377246,
1289
+ "grad_norm": 41.12651062011719,
1290
+ "learning_rate": 7.4145299145299155e-06,
1291
+ "loss": 0.6419,
1292
+ "step": 1730
1293
+ },
1294
+ {
1295
+ "epoch": 8.325748502994012,
1296
+ "grad_norm": 33.06435012817383,
1297
+ "learning_rate": 7.393162393162394e-06,
1298
+ "loss": 0.6121,
1299
+ "step": 1740
1300
+ },
1301
+ {
1302
+ "epoch": 8.373652694610778,
1303
+ "grad_norm": 54.51243209838867,
1304
+ "learning_rate": 7.371794871794873e-06,
1305
+ "loss": 0.554,
1306
+ "step": 1750
1307
+ },
1308
+ {
1309
+ "epoch": 8.421556886227545,
1310
+ "grad_norm": 23.176687240600586,
1311
+ "learning_rate": 7.350427350427351e-06,
1312
+ "loss": 0.8463,
1313
+ "step": 1760
1314
+ },
1315
+ {
1316
+ "epoch": 8.469461077844311,
1317
+ "grad_norm": 27.295547485351562,
1318
+ "learning_rate": 7.329059829059829e-06,
1319
+ "loss": 0.5512,
1320
+ "step": 1770
1321
+ },
1322
+ {
1323
+ "epoch": 8.517365269461077,
1324
+ "grad_norm": 19.80516242980957,
1325
+ "learning_rate": 7.307692307692308e-06,
1326
+ "loss": 0.7343,
1327
+ "step": 1780
1328
+ },
1329
+ {
1330
+ "epoch": 8.565269461077845,
1331
+ "grad_norm": 100.30529022216797,
1332
+ "learning_rate": 7.286324786324786e-06,
1333
+ "loss": 0.5751,
1334
+ "step": 1790
1335
+ },
1336
+ {
1337
+ "epoch": 8.613173652694611,
1338
+ "grad_norm": 40.754974365234375,
1339
+ "learning_rate": 7.264957264957266e-06,
1340
+ "loss": 0.8093,
1341
+ "step": 1800
1342
+ },
1343
+ {
1344
+ "epoch": 8.661077844311377,
1345
+ "grad_norm": 76.589111328125,
1346
+ "learning_rate": 7.243589743589744e-06,
1347
+ "loss": 0.7388,
1348
+ "step": 1810
1349
+ },
1350
+ {
1351
+ "epoch": 8.708982035928143,
1352
+ "grad_norm": 54.932838439941406,
1353
+ "learning_rate": 7.222222222222223e-06,
1354
+ "loss": 0.6432,
1355
+ "step": 1820
1356
+ },
1357
+ {
1358
+ "epoch": 8.75688622754491,
1359
+ "grad_norm": 54.24689483642578,
1360
+ "learning_rate": 7.2008547008547015e-06,
1361
+ "loss": 0.8261,
1362
+ "step": 1830
1363
+ },
1364
+ {
1365
+ "epoch": 8.804790419161677,
1366
+ "grad_norm": 26.55350112915039,
1367
+ "learning_rate": 7.17948717948718e-06,
1368
+ "loss": 0.701,
1369
+ "step": 1840
1370
+ },
1371
+ {
1372
+ "epoch": 8.852694610778443,
1373
+ "grad_norm": 28.47566032409668,
1374
+ "learning_rate": 7.158119658119659e-06,
1375
+ "loss": 0.8115,
1376
+ "step": 1850
1377
+ },
1378
+ {
1379
+ "epoch": 8.90059880239521,
1380
+ "grad_norm": 60.1458740234375,
1381
+ "learning_rate": 7.136752136752137e-06,
1382
+ "loss": 0.7178,
1383
+ "step": 1860
1384
+ },
1385
+ {
1386
+ "epoch": 8.948502994011976,
1387
+ "grad_norm": 47.09638214111328,
1388
+ "learning_rate": 7.115384615384616e-06,
1389
+ "loss": 0.8113,
1390
+ "step": 1870
1391
+ },
1392
+ {
1393
+ "epoch": 8.996407185628742,
1394
+ "grad_norm": 33.51266098022461,
1395
+ "learning_rate": 7.0940170940170945e-06,
1396
+ "loss": 0.7928,
1397
+ "step": 1880
1398
+ },
1399
+ {
1400
+ "epoch": 9.0,
1401
+ "eval_loss": 0.41668668389320374,
1402
+ "eval_macro_f1": 0.8441824722148172,
1403
+ "eval_runtime": 44.5976,
1404
+ "eval_samples_per_second": 37.468,
1405
+ "eval_steps_per_second": 4.686,
1406
+ "step": 1881
1407
+ },
1408
+ {
1409
+ "epoch": 9.04311377245509,
1410
+ "grad_norm": 81.31795501708984,
1411
+ "learning_rate": 7.072649572649574e-06,
1412
+ "loss": 0.6389,
1413
+ "step": 1890
1414
+ },
1415
+ {
1416
+ "epoch": 9.091017964071856,
1417
+ "grad_norm": 13.542675971984863,
1418
+ "learning_rate": 7.051282051282053e-06,
1419
+ "loss": 0.5636,
1420
+ "step": 1900
1421
+ },
1422
+ {
1423
+ "epoch": 9.138922155688622,
1424
+ "grad_norm": 82.46122741699219,
1425
+ "learning_rate": 7.02991452991453e-06,
1426
+ "loss": 0.6297,
1427
+ "step": 1910
1428
+ },
1429
+ {
1430
+ "epoch": 9.18682634730539,
1431
+ "grad_norm": 47.003292083740234,
1432
+ "learning_rate": 7.008547008547009e-06,
1433
+ "loss": 0.5565,
1434
+ "step": 1920
1435
+ },
1436
+ {
1437
+ "epoch": 9.234730538922156,
1438
+ "grad_norm": 26.577322006225586,
1439
+ "learning_rate": 6.9871794871794876e-06,
1440
+ "loss": 0.9967,
1441
+ "step": 1930
1442
+ },
1443
+ {
1444
+ "epoch": 9.282634730538922,
1445
+ "grad_norm": 47.41004180908203,
1446
+ "learning_rate": 6.965811965811966e-06,
1447
+ "loss": 0.5435,
1448
+ "step": 1940
1449
+ },
1450
+ {
1451
+ "epoch": 9.33053892215569,
1452
+ "grad_norm": 64.1046371459961,
1453
+ "learning_rate": 6.944444444444445e-06,
1454
+ "loss": 0.6405,
1455
+ "step": 1950
1456
+ },
1457
+ {
1458
+ "epoch": 9.378443113772455,
1459
+ "grad_norm": 48.981136322021484,
1460
+ "learning_rate": 6.923076923076923e-06,
1461
+ "loss": 0.5692,
1462
+ "step": 1960
1463
+ },
1464
+ {
1465
+ "epoch": 9.426347305389221,
1466
+ "grad_norm": 113.02758026123047,
1467
+ "learning_rate": 6.901709401709402e-06,
1468
+ "loss": 0.5958,
1469
+ "step": 1970
1470
+ },
1471
+ {
1472
+ "epoch": 9.474251497005987,
1473
+ "grad_norm": 87.0408935546875,
1474
+ "learning_rate": 6.880341880341881e-06,
1475
+ "loss": 0.7034,
1476
+ "step": 1980
1477
+ },
1478
+ {
1479
+ "epoch": 9.522155688622755,
1480
+ "grad_norm": 59.14445495605469,
1481
+ "learning_rate": 6.858974358974359e-06,
1482
+ "loss": 0.7209,
1483
+ "step": 1990
1484
+ },
1485
+ {
1486
+ "epoch": 9.570059880239521,
1487
+ "grad_norm": 53.770408630371094,
1488
+ "learning_rate": 6.837606837606839e-06,
1489
+ "loss": 0.8953,
1490
+ "step": 2000
1491
+ },
1492
+ {
1493
+ "epoch": 9.617964071856287,
1494
+ "grad_norm": 46.554481506347656,
1495
+ "learning_rate": 6.816239316239317e-06,
1496
+ "loss": 0.7084,
1497
+ "step": 2010
1498
+ },
1499
+ {
1500
+ "epoch": 9.665868263473055,
1501
+ "grad_norm": 41.41537857055664,
1502
+ "learning_rate": 6.794871794871796e-06,
1503
+ "loss": 0.5916,
1504
+ "step": 2020
1505
+ },
1506
+ {
1507
+ "epoch": 9.71377245508982,
1508
+ "grad_norm": 46.17745590209961,
1509
+ "learning_rate": 6.7735042735042745e-06,
1510
+ "loss": 0.5613,
1511
+ "step": 2030
1512
+ },
1513
+ {
1514
+ "epoch": 9.761676646706587,
1515
+ "grad_norm": 61.96057891845703,
1516
+ "learning_rate": 6.752136752136753e-06,
1517
+ "loss": 0.7686,
1518
+ "step": 2040
1519
+ },
1520
+ {
1521
+ "epoch": 9.809580838323352,
1522
+ "grad_norm": 17.55687713623047,
1523
+ "learning_rate": 6.730769230769232e-06,
1524
+ "loss": 0.7887,
1525
+ "step": 2050
1526
+ },
1527
+ {
1528
+ "epoch": 9.85748502994012,
1529
+ "grad_norm": 45.83679962158203,
1530
+ "learning_rate": 6.7094017094017094e-06,
1531
+ "loss": 0.4342,
1532
+ "step": 2060
1533
+ },
1534
+ {
1535
+ "epoch": 9.905389221556886,
1536
+ "grad_norm": 65.41094970703125,
1537
+ "learning_rate": 6.688034188034188e-06,
1538
+ "loss": 0.7364,
1539
+ "step": 2070
1540
+ },
1541
+ {
1542
+ "epoch": 9.953293413173652,
1543
+ "grad_norm": 63.39105987548828,
1544
+ "learning_rate": 6.666666666666667e-06,
1545
+ "loss": 0.5624,
1546
+ "step": 2080
1547
+ },
1548
+ {
1549
+ "epoch": 10.0,
1550
+ "grad_norm": 35.72822189331055,
1551
+ "learning_rate": 6.645299145299145e-06,
1552
+ "loss": 0.5399,
1553
+ "step": 2090
1554
+ },
1555
+ {
1556
+ "epoch": 10.0,
1557
+ "eval_loss": 0.5035926103591919,
1558
+ "eval_macro_f1": 0.8252431536013625,
1559
+ "eval_runtime": 43.3409,
1560
+ "eval_samples_per_second": 38.555,
1561
+ "eval_steps_per_second": 4.822,
1562
+ "step": 2090
1563
+ },
1564
+ {
1565
+ "epoch": 10.047904191616766,
1566
+ "grad_norm": 43.547481536865234,
1567
+ "learning_rate": 6.623931623931624e-06,
1568
+ "loss": 0.6161,
1569
+ "step": 2100
1570
+ },
1571
+ {
1572
+ "epoch": 10.095808383233534,
1573
+ "grad_norm": 9.250253677368164,
1574
+ "learning_rate": 6.602564102564103e-06,
1575
+ "loss": 0.6646,
1576
+ "step": 2110
1577
+ },
1578
+ {
1579
+ "epoch": 10.1437125748503,
1580
+ "grad_norm": 37.02254867553711,
1581
+ "learning_rate": 6.581196581196582e-06,
1582
+ "loss": 0.4968,
1583
+ "step": 2120
1584
+ },
1585
+ {
1586
+ "epoch": 10.191616766467066,
1587
+ "grad_norm": 54.5052490234375,
1588
+ "learning_rate": 6.5598290598290605e-06,
1589
+ "loss": 0.3927,
1590
+ "step": 2130
1591
+ },
1592
+ {
1593
+ "epoch": 10.239520958083832,
1594
+ "grad_norm": 64.98939514160156,
1595
+ "learning_rate": 6.538461538461539e-06,
1596
+ "loss": 0.5233,
1597
+ "step": 2140
1598
+ },
1599
+ {
1600
+ "epoch": 10.2874251497006,
1601
+ "grad_norm": 41.95616912841797,
1602
+ "learning_rate": 6.517094017094018e-06,
1603
+ "loss": 0.7305,
1604
+ "step": 2150
1605
+ },
1606
+ {
1607
+ "epoch": 10.335329341317365,
1608
+ "grad_norm": 26.417930603027344,
1609
+ "learning_rate": 6.495726495726496e-06,
1610
+ "loss": 0.5581,
1611
+ "step": 2160
1612
+ },
1613
+ {
1614
+ "epoch": 10.383233532934131,
1615
+ "grad_norm": 60.45222473144531,
1616
+ "learning_rate": 6.474358974358975e-06,
1617
+ "loss": 0.5982,
1618
+ "step": 2170
1619
+ },
1620
+ {
1621
+ "epoch": 10.431137724550899,
1622
+ "grad_norm": 47.69218826293945,
1623
+ "learning_rate": 6.4529914529914535e-06,
1624
+ "loss": 0.3558,
1625
+ "step": 2180
1626
+ },
1627
+ {
1628
+ "epoch": 10.479041916167665,
1629
+ "grad_norm": 33.18900680541992,
1630
+ "learning_rate": 6.431623931623933e-06,
1631
+ "loss": 0.6521,
1632
+ "step": 2190
1633
+ },
1634
+ {
1635
+ "epoch": 10.52694610778443,
1636
+ "grad_norm": 39.501426696777344,
1637
+ "learning_rate": 6.410256410256412e-06,
1638
+ "loss": 0.6646,
1639
+ "step": 2200
1640
+ },
1641
+ {
1642
+ "epoch": 10.574850299401197,
1643
+ "grad_norm": 61.5817985534668,
1644
+ "learning_rate": 6.3888888888888885e-06,
1645
+ "loss": 0.6448,
1646
+ "step": 2210
1647
+ },
1648
+ {
1649
+ "epoch": 10.622754491017965,
1650
+ "grad_norm": 37.70832061767578,
1651
+ "learning_rate": 6.367521367521368e-06,
1652
+ "loss": 0.4272,
1653
+ "step": 2220
1654
+ },
1655
+ {
1656
+ "epoch": 10.67065868263473,
1657
+ "grad_norm": 56.18067169189453,
1658
+ "learning_rate": 6.3461538461538466e-06,
1659
+ "loss": 0.5429,
1660
+ "step": 2230
1661
+ },
1662
+ {
1663
+ "epoch": 10.718562874251496,
1664
+ "grad_norm": 38.271278381347656,
1665
+ "learning_rate": 6.324786324786325e-06,
1666
+ "loss": 0.503,
1667
+ "step": 2240
1668
+ },
1669
+ {
1670
+ "epoch": 10.766467065868264,
1671
+ "grad_norm": 53.65212631225586,
1672
+ "learning_rate": 6.303418803418804e-06,
1673
+ "loss": 0.5207,
1674
+ "step": 2250
1675
+ },
1676
+ {
1677
+ "epoch": 10.81437125748503,
1678
+ "grad_norm": 15.48988151550293,
1679
+ "learning_rate": 6.282051282051282e-06,
1680
+ "loss": 0.6707,
1681
+ "step": 2260
1682
+ },
1683
+ {
1684
+ "epoch": 10.862275449101796,
1685
+ "grad_norm": 44.31208038330078,
1686
+ "learning_rate": 6.260683760683761e-06,
1687
+ "loss": 0.3833,
1688
+ "step": 2270
1689
+ },
1690
+ {
1691
+ "epoch": 10.910179640718562,
1692
+ "grad_norm": 48.97806167602539,
1693
+ "learning_rate": 6.23931623931624e-06,
1694
+ "loss": 0.5623,
1695
+ "step": 2280
1696
+ },
1697
+ {
1698
+ "epoch": 10.95808383233533,
1699
+ "grad_norm": 73.36396789550781,
1700
+ "learning_rate": 6.217948717948718e-06,
1701
+ "loss": 0.6053,
1702
+ "step": 2290
1703
+ },
1704
+ {
1705
+ "epoch": 11.0,
1706
+ "eval_loss": 0.4976274073123932,
1707
+ "eval_macro_f1": 0.8447303083464661,
1708
+ "eval_runtime": 44.1148,
1709
+ "eval_samples_per_second": 37.878,
1710
+ "eval_steps_per_second": 4.738,
1711
+ "step": 2299
1712
+ },
1713
+ {
1714
+ "epoch": 11.004790419161676,
1715
+ "grad_norm": 52.19700622558594,
1716
+ "learning_rate": 6.196581196581198e-06,
1717
+ "loss": 0.4362,
1718
+ "step": 2300
1719
+ },
1720
+ {
1721
+ "epoch": 11.052694610778444,
1722
+ "grad_norm": 12.376163482666016,
1723
+ "learning_rate": 6.175213675213676e-06,
1724
+ "loss": 0.2856,
1725
+ "step": 2310
1726
+ },
1727
+ {
1728
+ "epoch": 11.10059880239521,
1729
+ "grad_norm": 55.741554260253906,
1730
+ "learning_rate": 6.153846153846155e-06,
1731
+ "loss": 0.4785,
1732
+ "step": 2320
1733
+ },
1734
+ {
1735
+ "epoch": 11.148502994011976,
1736
+ "grad_norm": 23.254854202270508,
1737
+ "learning_rate": 6.1324786324786335e-06,
1738
+ "loss": 0.5646,
1739
+ "step": 2330
1740
+ },
1741
+ {
1742
+ "epoch": 11.196407185628743,
1743
+ "grad_norm": 54.249332427978516,
1744
+ "learning_rate": 6.111111111111112e-06,
1745
+ "loss": 0.4413,
1746
+ "step": 2340
1747
+ },
1748
+ {
1749
+ "epoch": 11.24431137724551,
1750
+ "grad_norm": 91.06517791748047,
1751
+ "learning_rate": 6.08974358974359e-06,
1752
+ "loss": 0.4921,
1753
+ "step": 2350
1754
+ },
1755
+ {
1756
+ "epoch": 11.292215568862275,
1757
+ "grad_norm": 12.940735816955566,
1758
+ "learning_rate": 6.0683760683760684e-06,
1759
+ "loss": 0.5419,
1760
+ "step": 2360
1761
+ },
1762
+ {
1763
+ "epoch": 11.340119760479041,
1764
+ "grad_norm": 88.24459075927734,
1765
+ "learning_rate": 6.047008547008547e-06,
1766
+ "loss": 0.582,
1767
+ "step": 2370
1768
+ },
1769
+ {
1770
+ "epoch": 11.388023952095809,
1771
+ "grad_norm": 28.848173141479492,
1772
+ "learning_rate": 6.025641025641026e-06,
1773
+ "loss": 0.5232,
1774
+ "step": 2380
1775
+ },
1776
+ {
1777
+ "epoch": 11.435928143712575,
1778
+ "grad_norm": 18.56818389892578,
1779
+ "learning_rate": 6.004273504273504e-06,
1780
+ "loss": 0.4061,
1781
+ "step": 2390
1782
+ },
1783
+ {
1784
+ "epoch": 11.48383233532934,
1785
+ "grad_norm": 88.79745483398438,
1786
+ "learning_rate": 5.982905982905983e-06,
1787
+ "loss": 0.4603,
1788
+ "step": 2400
1789
+ },
1790
+ {
1791
+ "epoch": 11.531736526946109,
1792
+ "grad_norm": 59.039039611816406,
1793
+ "learning_rate": 5.961538461538462e-06,
1794
+ "loss": 0.4677,
1795
+ "step": 2410
1796
+ },
1797
+ {
1798
+ "epoch": 11.579640718562874,
1799
+ "grad_norm": 60.00382614135742,
1800
+ "learning_rate": 5.940170940170941e-06,
1801
+ "loss": 0.3458,
1802
+ "step": 2420
1803
+ },
1804
+ {
1805
+ "epoch": 11.62754491017964,
1806
+ "grad_norm": 37.48514938354492,
1807
+ "learning_rate": 5.9188034188034195e-06,
1808
+ "loss": 0.6475,
1809
+ "step": 2430
1810
+ },
1811
+ {
1812
+ "epoch": 11.675449101796406,
1813
+ "grad_norm": 53.149925231933594,
1814
+ "learning_rate": 5.897435897435898e-06,
1815
+ "loss": 0.4033,
1816
+ "step": 2440
1817
+ },
1818
+ {
1819
+ "epoch": 11.723353293413174,
1820
+ "grad_norm": 57.217735290527344,
1821
+ "learning_rate": 5.876068376068377e-06,
1822
+ "loss": 0.5827,
1823
+ "step": 2450
1824
+ },
1825
+ {
1826
+ "epoch": 11.77125748502994,
1827
+ "grad_norm": 32.07596206665039,
1828
+ "learning_rate": 5.854700854700855e-06,
1829
+ "loss": 0.4473,
1830
+ "step": 2460
1831
+ },
1832
+ {
1833
+ "epoch": 11.819161676646706,
1834
+ "grad_norm": 39.394474029541016,
1835
+ "learning_rate": 5.833333333333334e-06,
1836
+ "loss": 0.6156,
1837
+ "step": 2470
1838
+ },
1839
+ {
1840
+ "epoch": 11.867065868263474,
1841
+ "grad_norm": 42.22713088989258,
1842
+ "learning_rate": 5.8119658119658126e-06,
1843
+ "loss": 0.5249,
1844
+ "step": 2480
1845
+ },
1846
+ {
1847
+ "epoch": 11.91497005988024,
1848
+ "grad_norm": 44.544944763183594,
1849
+ "learning_rate": 5.790598290598292e-06,
1850
+ "loss": 0.5426,
1851
+ "step": 2490
1852
+ },
1853
+ {
1854
+ "epoch": 11.962874251497006,
1855
+ "grad_norm": 37.793052673339844,
1856
+ "learning_rate": 5.769230769230769e-06,
1857
+ "loss": 0.7017,
1858
+ "step": 2500
1859
+ },
1860
+ {
1861
+ "epoch": 12.0,
1862
+ "eval_loss": 0.5270896553993225,
1863
+ "eval_macro_f1": 0.8419574495547526,
1864
+ "eval_runtime": 52.4362,
1865
+ "eval_samples_per_second": 31.867,
1866
+ "eval_steps_per_second": 3.986,
1867
+ "step": 2508
1868
+ },
1869
+ {
1870
+ "epoch": 12.009580838323354,
1871
+ "grad_norm": 33.670814514160156,
1872
+ "learning_rate": 5.7478632478632475e-06,
1873
+ "loss": 0.348,
1874
+ "step": 2510
1875
+ },
1876
+ {
1877
+ "epoch": 12.05748502994012,
1878
+ "grad_norm": 16.13005256652832,
1879
+ "learning_rate": 5.726495726495727e-06,
1880
+ "loss": 0.3518,
1881
+ "step": 2520
1882
+ },
1883
+ {
1884
+ "epoch": 12.105389221556885,
1885
+ "grad_norm": 117.29934692382812,
1886
+ "learning_rate": 5.705128205128206e-06,
1887
+ "loss": 0.4044,
1888
+ "step": 2530
1889
+ },
1890
+ {
1891
+ "epoch": 12.153293413173653,
1892
+ "grad_norm": 66.2587661743164,
1893
+ "learning_rate": 5.683760683760684e-06,
1894
+ "loss": 0.3835,
1895
+ "step": 2540
1896
+ },
1897
+ {
1898
+ "epoch": 12.20119760479042,
1899
+ "grad_norm": 39.18204879760742,
1900
+ "learning_rate": 5.662393162393163e-06,
1901
+ "loss": 0.5428,
1902
+ "step": 2550
1903
+ },
1904
+ {
1905
+ "epoch": 12.249101796407185,
1906
+ "grad_norm": 29.139354705810547,
1907
+ "learning_rate": 5.641025641025641e-06,
1908
+ "loss": 0.2677,
1909
+ "step": 2560
1910
+ },
1911
+ {
1912
+ "epoch": 12.297005988023953,
1913
+ "grad_norm": 49.968666076660156,
1914
+ "learning_rate": 5.61965811965812e-06,
1915
+ "loss": 0.4607,
1916
+ "step": 2570
1917
+ },
1918
+ {
1919
+ "epoch": 12.344910179640719,
1920
+ "grad_norm": 30.069171905517578,
1921
+ "learning_rate": 5.598290598290599e-06,
1922
+ "loss": 0.4371,
1923
+ "step": 2580
1924
+ },
1925
+ {
1926
+ "epoch": 12.392814371257485,
1927
+ "grad_norm": 21.969675064086914,
1928
+ "learning_rate": 5.576923076923077e-06,
1929
+ "loss": 0.473,
1930
+ "step": 2590
1931
+ },
1932
+ {
1933
+ "epoch": 12.44071856287425,
1934
+ "grad_norm": 76.71914672851562,
1935
+ "learning_rate": 5.555555555555557e-06,
1936
+ "loss": 0.3471,
1937
+ "step": 2600
1938
+ },
1939
+ {
1940
+ "epoch": 12.488622754491018,
1941
+ "grad_norm": 50.17422866821289,
1942
+ "learning_rate": 5.534188034188035e-06,
1943
+ "loss": 0.7815,
1944
+ "step": 2610
1945
+ },
1946
+ {
1947
+ "epoch": 12.536526946107784,
1948
+ "grad_norm": 44.04591751098633,
1949
+ "learning_rate": 5.512820512820514e-06,
1950
+ "loss": 0.5847,
1951
+ "step": 2620
1952
+ },
1953
+ {
1954
+ "epoch": 12.58443113772455,
1955
+ "grad_norm": 17.07801055908203,
1956
+ "learning_rate": 5.4914529914529925e-06,
1957
+ "loss": 0.5836,
1958
+ "step": 2630
1959
+ },
1960
+ {
1961
+ "epoch": 12.632335329341318,
1962
+ "grad_norm": 37.61981201171875,
1963
+ "learning_rate": 5.470085470085471e-06,
1964
+ "loss": 0.6164,
1965
+ "step": 2640
1966
+ },
1967
+ {
1968
+ "epoch": 12.680239520958084,
1969
+ "grad_norm": 45.67870330810547,
1970
+ "learning_rate": 5.448717948717949e-06,
1971
+ "loss": 0.4233,
1972
+ "step": 2650
1973
+ },
1974
+ {
1975
+ "epoch": 12.72814371257485,
1976
+ "grad_norm": 6.8784918785095215,
1977
+ "learning_rate": 5.4273504273504275e-06,
1978
+ "loss": 0.4341,
1979
+ "step": 2660
1980
+ },
1981
+ {
1982
+ "epoch": 12.776047904191616,
1983
+ "grad_norm": 106.60427856445312,
1984
+ "learning_rate": 5.405982905982906e-06,
1985
+ "loss": 0.4442,
1986
+ "step": 2670
1987
+ },
1988
+ {
1989
+ "epoch": 12.823952095808384,
1990
+ "grad_norm": 10.980769157409668,
1991
+ "learning_rate": 5.384615384615385e-06,
1992
+ "loss": 0.3069,
1993
+ "step": 2680
1994
+ },
1995
+ {
1996
+ "epoch": 12.87185628742515,
1997
+ "grad_norm": 108.9623794555664,
1998
+ "learning_rate": 5.363247863247863e-06,
1999
+ "loss": 0.461,
2000
+ "step": 2690
2001
+ },
2002
+ {
2003
+ "epoch": 12.919760479041916,
2004
+ "grad_norm": 49.222408294677734,
2005
+ "learning_rate": 5.341880341880342e-06,
2006
+ "loss": 0.5779,
2007
+ "step": 2700
2008
+ },
2009
+ {
2010
+ "epoch": 12.967664670658683,
2011
+ "grad_norm": 77.78028869628906,
2012
+ "learning_rate": 5.320512820512821e-06,
2013
+ "loss": 0.5748,
2014
+ "step": 2710
2015
+ },
2016
+ {
2017
+ "epoch": 13.0,
2018
+ "eval_loss": 0.6038709878921509,
2019
+ "eval_macro_f1": 0.82953734826579,
2020
+ "eval_runtime": 45.778,
2021
+ "eval_samples_per_second": 36.502,
2022
+ "eval_steps_per_second": 4.566,
2023
+ "step": 2717
2024
+ },
2025
+ {
2026
+ "epoch": 13.01437125748503,
2027
+ "grad_norm": 64.35317993164062,
2028
+ "learning_rate": 5.2991452991453e-06,
2029
+ "loss": 0.4684,
2030
+ "step": 2720
2031
+ },
2032
+ {
2033
+ "epoch": 13.062275449101797,
2034
+ "grad_norm": 27.278182983398438,
2035
+ "learning_rate": 5.2777777777777785e-06,
2036
+ "loss": 0.2145,
2037
+ "step": 2730
2038
+ },
2039
+ {
2040
+ "epoch": 13.110179640718563,
2041
+ "grad_norm": 12.438913345336914,
2042
+ "learning_rate": 5.256410256410257e-06,
2043
+ "loss": 0.3516,
2044
+ "step": 2740
2045
+ },
2046
+ {
2047
+ "epoch": 13.158083832335329,
2048
+ "grad_norm": 111.6015396118164,
2049
+ "learning_rate": 5.235042735042736e-06,
2050
+ "loss": 0.4645,
2051
+ "step": 2750
2052
+ },
2053
+ {
2054
+ "epoch": 13.205988023952095,
2055
+ "grad_norm": 29.416948318481445,
2056
+ "learning_rate": 5.213675213675214e-06,
2057
+ "loss": 0.4283,
2058
+ "step": 2760
2059
+ },
2060
+ {
2061
+ "epoch": 13.253892215568863,
2062
+ "grad_norm": 42.40279006958008,
2063
+ "learning_rate": 5.192307692307693e-06,
2064
+ "loss": 0.3558,
2065
+ "step": 2770
2066
+ },
2067
+ {
2068
+ "epoch": 13.301796407185629,
2069
+ "grad_norm": 90.29447937011719,
2070
+ "learning_rate": 5.1709401709401716e-06,
2071
+ "loss": 0.4245,
2072
+ "step": 2780
2073
+ },
2074
+ {
2075
+ "epoch": 13.349700598802395,
2076
+ "grad_norm": 51.8712272644043,
2077
+ "learning_rate": 5.149572649572649e-06,
2078
+ "loss": 0.3301,
2079
+ "step": 2790
2080
+ },
2081
+ {
2082
+ "epoch": 13.397604790419162,
2083
+ "grad_norm": 32.75868606567383,
2084
+ "learning_rate": 5.128205128205128e-06,
2085
+ "loss": 0.401,
2086
+ "step": 2800
2087
+ },
2088
+ {
2089
+ "epoch": 13.445508982035928,
2090
+ "grad_norm": 14.076770782470703,
2091
+ "learning_rate": 5.1068376068376065e-06,
2092
+ "loss": 0.2116,
2093
+ "step": 2810
2094
+ },
2095
+ {
2096
+ "epoch": 13.493413173652694,
2097
+ "grad_norm": 158.29722595214844,
2098
+ "learning_rate": 5.085470085470086e-06,
2099
+ "loss": 0.4145,
2100
+ "step": 2820
2101
+ },
2102
+ {
2103
+ "epoch": 13.54131736526946,
2104
+ "grad_norm": 43.535980224609375,
2105
+ "learning_rate": 5.064102564102565e-06,
2106
+ "loss": 0.5061,
2107
+ "step": 2830
2108
+ },
2109
+ {
2110
+ "epoch": 13.589221556886228,
2111
+ "grad_norm": 129.22494506835938,
2112
+ "learning_rate": 5.042735042735043e-06,
2113
+ "loss": 0.2653,
2114
+ "step": 2840
2115
+ },
2116
+ {
2117
+ "epoch": 13.637125748502994,
2118
+ "grad_norm": 19.907264709472656,
2119
+ "learning_rate": 5.021367521367522e-06,
2120
+ "loss": 0.5019,
2121
+ "step": 2850
2122
+ },
2123
+ {
2124
+ "epoch": 13.68502994011976,
2125
+ "grad_norm": 101.03581237792969,
2126
+ "learning_rate": 5e-06,
2127
+ "loss": 0.6201,
2128
+ "step": 2860
2129
+ },
2130
+ {
2131
+ "epoch": 13.732934131736528,
2132
+ "grad_norm": 54.207130432128906,
2133
+ "learning_rate": 4.978632478632479e-06,
2134
+ "loss": 0.4591,
2135
+ "step": 2870
2136
+ },
2137
+ {
2138
+ "epoch": 13.780838323353294,
2139
+ "grad_norm": 57.07173538208008,
2140
+ "learning_rate": 4.957264957264958e-06,
2141
+ "loss": 0.5418,
2142
+ "step": 2880
2143
+ },
2144
+ {
2145
+ "epoch": 13.82874251497006,
2146
+ "grad_norm": 54.304054260253906,
2147
+ "learning_rate": 4.935897435897436e-06,
2148
+ "loss": 0.6254,
2149
+ "step": 2890
2150
+ },
2151
+ {
2152
+ "epoch": 13.876646706586826,
2153
+ "grad_norm": 74.52729797363281,
2154
+ "learning_rate": 4.914529914529915e-06,
2155
+ "loss": 0.302,
2156
+ "step": 2900
2157
+ },
2158
+ {
2159
+ "epoch": 13.924550898203593,
2160
+ "grad_norm": 46.62977600097656,
2161
+ "learning_rate": 4.8931623931623934e-06,
2162
+ "loss": 0.3676,
2163
+ "step": 2910
2164
+ },
2165
+ {
2166
+ "epoch": 13.97245508982036,
2167
+ "grad_norm": 73.47917938232422,
2168
+ "learning_rate": 4.871794871794872e-06,
2169
+ "loss": 0.478,
2170
+ "step": 2920
2171
+ },
2172
+ {
2173
+ "epoch": 14.0,
2174
+ "eval_loss": 0.5613604784011841,
2175
+ "eval_macro_f1": 0.833819031115799,
2176
+ "eval_runtime": 44.8642,
2177
+ "eval_samples_per_second": 37.246,
2178
+ "eval_steps_per_second": 4.659,
2179
+ "step": 2926
2180
+ },
2181
+ {
2182
+ "epoch": 14.019161676646707,
2183
+ "grad_norm": 106.49620056152344,
2184
+ "learning_rate": 4.850427350427351e-06,
2185
+ "loss": 0.4661,
2186
+ "step": 2930
2187
+ },
2188
+ {
2189
+ "epoch": 14.067065868263473,
2190
+ "grad_norm": 23.546146392822266,
2191
+ "learning_rate": 4.829059829059829e-06,
2192
+ "loss": 0.3035,
2193
+ "step": 2940
2194
+ },
2195
+ {
2196
+ "epoch": 14.114970059880239,
2197
+ "grad_norm": 23.96906280517578,
2198
+ "learning_rate": 4.807692307692308e-06,
2199
+ "loss": 0.3686,
2200
+ "step": 2950
2201
+ },
2202
+ {
2203
+ "epoch": 14.162874251497007,
2204
+ "grad_norm": 82.27886199951172,
2205
+ "learning_rate": 4.786324786324787e-06,
2206
+ "loss": 0.4799,
2207
+ "step": 2960
2208
+ },
2209
+ {
2210
+ "epoch": 14.210778443113773,
2211
+ "grad_norm": 10.91873550415039,
2212
+ "learning_rate": 4.764957264957265e-06,
2213
+ "loss": 0.2748,
2214
+ "step": 2970
2215
+ },
2216
+ {
2217
+ "epoch": 14.258682634730539,
2218
+ "grad_norm": 38.029109954833984,
2219
+ "learning_rate": 4.743589743589744e-06,
2220
+ "loss": 0.3756,
2221
+ "step": 2980
2222
+ },
2223
+ {
2224
+ "epoch": 14.306586826347305,
2225
+ "grad_norm": 73.956787109375,
2226
+ "learning_rate": 4.722222222222222e-06,
2227
+ "loss": 0.5632,
2228
+ "step": 2990
2229
+ },
2230
+ {
2231
+ "epoch": 14.354491017964072,
2232
+ "grad_norm": 19.53325843811035,
2233
+ "learning_rate": 4.700854700854701e-06,
2234
+ "loss": 0.3685,
2235
+ "step": 3000
2236
+ },
2237
+ {
2238
+ "epoch": 14.402395209580838,
2239
+ "grad_norm": 21.934566497802734,
2240
+ "learning_rate": 4.6794871794871795e-06,
2241
+ "loss": 0.3186,
2242
+ "step": 3010
2243
+ },
2244
+ {
2245
+ "epoch": 14.450299401197604,
2246
+ "grad_norm": 48.01268768310547,
2247
+ "learning_rate": 4.658119658119659e-06,
2248
+ "loss": 0.4358,
2249
+ "step": 3020
2250
+ },
2251
+ {
2252
+ "epoch": 14.498203592814372,
2253
+ "grad_norm": 67.98200225830078,
2254
+ "learning_rate": 4.6367521367521375e-06,
2255
+ "loss": 0.2328,
2256
+ "step": 3030
2257
+ },
2258
+ {
2259
+ "epoch": 14.546107784431138,
2260
+ "grad_norm": 20.185400009155273,
2261
+ "learning_rate": 4.615384615384616e-06,
2262
+ "loss": 0.4095,
2263
+ "step": 3040
2264
+ },
2265
+ {
2266
+ "epoch": 14.594011976047904,
2267
+ "grad_norm": 35.661773681640625,
2268
+ "learning_rate": 4.594017094017094e-06,
2269
+ "loss": 0.6548,
2270
+ "step": 3050
2271
+ },
2272
+ {
2273
+ "epoch": 14.64191616766467,
2274
+ "grad_norm": 36.68177032470703,
2275
+ "learning_rate": 4.5726495726495725e-06,
2276
+ "loss": 0.3479,
2277
+ "step": 3060
2278
+ },
2279
+ {
2280
+ "epoch": 14.689820359281438,
2281
+ "grad_norm": 19.040000915527344,
2282
+ "learning_rate": 4.551282051282052e-06,
2283
+ "loss": 0.3784,
2284
+ "step": 3070
2285
+ },
2286
+ {
2287
+ "epoch": 14.737724550898204,
2288
+ "grad_norm": 77.58363342285156,
2289
+ "learning_rate": 4.5299145299145306e-06,
2290
+ "loss": 0.4044,
2291
+ "step": 3080
2292
+ },
2293
+ {
2294
+ "epoch": 14.78562874251497,
2295
+ "grad_norm": 98.52961730957031,
2296
+ "learning_rate": 4.508547008547009e-06,
2297
+ "loss": 0.42,
2298
+ "step": 3090
2299
+ },
2300
+ {
2301
+ "epoch": 14.833532934131737,
2302
+ "grad_norm": 167.6805419921875,
2303
+ "learning_rate": 4.487179487179488e-06,
2304
+ "loss": 0.3661,
2305
+ "step": 3100
2306
+ },
2307
+ {
2308
+ "epoch": 14.881437125748503,
2309
+ "grad_norm": 81.3060302734375,
2310
+ "learning_rate": 4.465811965811966e-06,
2311
+ "loss": 0.6895,
2312
+ "step": 3110
2313
+ },
2314
+ {
2315
+ "epoch": 14.92934131736527,
2316
+ "grad_norm": 31.589746475219727,
2317
+ "learning_rate": 4.444444444444444e-06,
2318
+ "loss": 0.3983,
2319
+ "step": 3120
2320
+ },
2321
+ {
2322
+ "epoch": 14.977245508982035,
2323
+ "grad_norm": 33.01588821411133,
2324
+ "learning_rate": 4.423076923076924e-06,
2325
+ "loss": 0.3302,
2326
+ "step": 3130
2327
+ },
2328
+ {
2329
+ "epoch": 15.0,
2330
+ "eval_loss": 0.6139780282974243,
2331
+ "eval_macro_f1": 0.850078934734469,
2332
+ "eval_runtime": 61.1653,
2333
+ "eval_samples_per_second": 27.319,
2334
+ "eval_steps_per_second": 3.417,
2335
+ "step": 3135
2336
+ },
2337
+ {
2338
+ "epoch": 15.023952095808383,
2339
+ "grad_norm": 63.095176696777344,
2340
+ "learning_rate": 4.401709401709402e-06,
2341
+ "loss": 0.492,
2342
+ "step": 3140
2343
+ },
2344
+ {
2345
+ "epoch": 15.071856287425149,
2346
+ "grad_norm": 14.580053329467773,
2347
+ "learning_rate": 4.380341880341881e-06,
2348
+ "loss": 0.5926,
2349
+ "step": 3150
2350
+ },
2351
+ {
2352
+ "epoch": 15.119760479041917,
2353
+ "grad_norm": 9.320775032043457,
2354
+ "learning_rate": 4.358974358974359e-06,
2355
+ "loss": 0.455,
2356
+ "step": 3160
2357
+ },
2358
+ {
2359
+ "epoch": 15.167664670658683,
2360
+ "grad_norm": 57.686344146728516,
2361
+ "learning_rate": 4.337606837606838e-06,
2362
+ "loss": 0.3759,
2363
+ "step": 3170
2364
+ },
2365
+ {
2366
+ "epoch": 15.215568862275449,
2367
+ "grad_norm": 4.791827201843262,
2368
+ "learning_rate": 4.316239316239317e-06,
2369
+ "loss": 0.2594,
2370
+ "step": 3180
2371
+ },
2372
+ {
2373
+ "epoch": 15.263473053892216,
2374
+ "grad_norm": 181.0989532470703,
2375
+ "learning_rate": 4.294871794871795e-06,
2376
+ "loss": 0.4456,
2377
+ "step": 3190
2378
+ },
2379
+ {
2380
+ "epoch": 15.311377245508982,
2381
+ "grad_norm": 15.999568939208984,
2382
+ "learning_rate": 4.273504273504274e-06,
2383
+ "loss": 0.3199,
2384
+ "step": 3200
2385
+ },
2386
+ {
2387
+ "epoch": 15.359281437125748,
2388
+ "grad_norm": 34.29719543457031,
2389
+ "learning_rate": 4.2521367521367524e-06,
2390
+ "loss": 0.3138,
2391
+ "step": 3210
2392
+ },
2393
+ {
2394
+ "epoch": 15.407185628742514,
2395
+ "grad_norm": 108.82241821289062,
2396
+ "learning_rate": 4.230769230769231e-06,
2397
+ "loss": 0.381,
2398
+ "step": 3220
2399
+ },
2400
+ {
2401
+ "epoch": 15.455089820359282,
2402
+ "grad_norm": 7.426156044006348,
2403
+ "learning_rate": 4.20940170940171e-06,
2404
+ "loss": 0.3027,
2405
+ "step": 3230
2406
+ },
2407
+ {
2408
+ "epoch": 15.502994011976048,
2409
+ "grad_norm": 33.00428009033203,
2410
+ "learning_rate": 4.188034188034188e-06,
2411
+ "loss": 0.2268,
2412
+ "step": 3240
2413
+ },
2414
+ {
2415
+ "epoch": 15.550898203592814,
2416
+ "grad_norm": 64.04177856445312,
2417
+ "learning_rate": 4.166666666666667e-06,
2418
+ "loss": 0.4501,
2419
+ "step": 3250
2420
+ },
2421
+ {
2422
+ "epoch": 15.598802395209582,
2423
+ "grad_norm": 28.86948585510254,
2424
+ "learning_rate": 4.145299145299146e-06,
2425
+ "loss": 0.4704,
2426
+ "step": 3260
2427
+ },
2428
+ {
2429
+ "epoch": 15.646706586826348,
2430
+ "grad_norm": 33.580509185791016,
2431
+ "learning_rate": 4.123931623931624e-06,
2432
+ "loss": 0.3373,
2433
+ "step": 3270
2434
+ },
2435
+ {
2436
+ "epoch": 15.694610778443113,
2437
+ "grad_norm": 27.37198257446289,
2438
+ "learning_rate": 4.102564102564103e-06,
2439
+ "loss": 0.3751,
2440
+ "step": 3280
2441
+ },
2442
+ {
2443
+ "epoch": 15.74251497005988,
2444
+ "grad_norm": 29.447359085083008,
2445
+ "learning_rate": 4.081196581196581e-06,
2446
+ "loss": 0.2583,
2447
+ "step": 3290
2448
+ },
2449
+ {
2450
+ "epoch": 15.790419161676647,
2451
+ "grad_norm": 53.31045913696289,
2452
+ "learning_rate": 4.05982905982906e-06,
2453
+ "loss": 0.2685,
2454
+ "step": 3300
2455
+ },
2456
+ {
2457
+ "epoch": 15.838323353293413,
2458
+ "grad_norm": 135.5189971923828,
2459
+ "learning_rate": 4.0384615384615385e-06,
2460
+ "loss": 0.6179,
2461
+ "step": 3310
2462
+ },
2463
+ {
2464
+ "epoch": 15.886227544910179,
2465
+ "grad_norm": 24.811853408813477,
2466
+ "learning_rate": 4.017094017094018e-06,
2467
+ "loss": 0.2871,
2468
+ "step": 3320
2469
+ },
2470
+ {
2471
+ "epoch": 15.934131736526947,
2472
+ "grad_norm": 44.590511322021484,
2473
+ "learning_rate": 3.9957264957264966e-06,
2474
+ "loss": 0.3918,
2475
+ "step": 3330
2476
+ },
2477
+ {
2478
+ "epoch": 15.982035928143713,
2479
+ "grad_norm": 34.29424285888672,
2480
+ "learning_rate": 3.974358974358974e-06,
2481
+ "loss": 0.4284,
2482
+ "step": 3340
2483
+ },
2484
+ {
2485
+ "epoch": 16.0,
2486
+ "eval_loss": 0.6892519593238831,
2487
+ "eval_macro_f1": 0.8416606209561599,
2488
+ "eval_runtime": 68.7411,
2489
+ "eval_samples_per_second": 24.309,
2490
+ "eval_steps_per_second": 3.04,
2491
+ "step": 3344
2492
+ },
2493
+ {
2494
+ "epoch": 16.02874251497006,
2495
+ "grad_norm": 76.20410919189453,
2496
+ "learning_rate": 3.952991452991453e-06,
2497
+ "loss": 0.2715,
2498
+ "step": 3350
2499
+ },
2500
+ {
2501
+ "epoch": 16.076646706586825,
2502
+ "grad_norm": 97.97017669677734,
2503
+ "learning_rate": 3.9316239316239315e-06,
2504
+ "loss": 0.4799,
2505
+ "step": 3360
2506
+ },
2507
+ {
2508
+ "epoch": 16.124550898203594,
2509
+ "grad_norm": 42.71623229980469,
2510
+ "learning_rate": 3.910256410256411e-06,
2511
+ "loss": 0.4008,
2512
+ "step": 3370
2513
+ },
2514
+ {
2515
+ "epoch": 16.17245508982036,
2516
+ "grad_norm": 48.501312255859375,
2517
+ "learning_rate": 3.88888888888889e-06,
2518
+ "loss": 0.4609,
2519
+ "step": 3380
2520
+ },
2521
+ {
2522
+ "epoch": 16.220359281437126,
2523
+ "grad_norm": 15.1089448928833,
2524
+ "learning_rate": 3.867521367521368e-06,
2525
+ "loss": 0.2344,
2526
+ "step": 3390
2527
+ },
2528
+ {
2529
+ "epoch": 16.268263473053892,
2530
+ "grad_norm": 87.63355255126953,
2531
+ "learning_rate": 3.846153846153847e-06,
2532
+ "loss": 0.4965,
2533
+ "step": 3400
2534
+ },
2535
+ {
2536
+ "epoch": 16.316167664670658,
2537
+ "grad_norm": 54.828556060791016,
2538
+ "learning_rate": 3.8247863247863246e-06,
2539
+ "loss": 0.3707,
2540
+ "step": 3410
2541
+ },
2542
+ {
2543
+ "epoch": 16.364071856287424,
2544
+ "grad_norm": 74.80049133300781,
2545
+ "learning_rate": 3.8034188034188036e-06,
2546
+ "loss": 0.3693,
2547
+ "step": 3420
2548
+ },
2549
+ {
2550
+ "epoch": 16.41197604790419,
2551
+ "grad_norm": 18.754371643066406,
2552
+ "learning_rate": 3.782051282051282e-06,
2553
+ "loss": 0.3593,
2554
+ "step": 3430
2555
+ },
2556
+ {
2557
+ "epoch": 16.45988023952096,
2558
+ "grad_norm": 24.015533447265625,
2559
+ "learning_rate": 3.760683760683761e-06,
2560
+ "loss": 0.215,
2561
+ "step": 3440
2562
+ },
2563
+ {
2564
+ "epoch": 16.507784431137726,
2565
+ "grad_norm": 41.305519104003906,
2566
+ "learning_rate": 3.73931623931624e-06,
2567
+ "loss": 0.3713,
2568
+ "step": 3450
2569
+ },
2570
+ {
2571
+ "epoch": 16.55568862275449,
2572
+ "grad_norm": 54.54863739013672,
2573
+ "learning_rate": 3.7179487179487184e-06,
2574
+ "loss": 0.4386,
2575
+ "step": 3460
2576
+ },
2577
+ {
2578
+ "epoch": 16.603592814371257,
2579
+ "grad_norm": 71.85780334472656,
2580
+ "learning_rate": 3.696581196581197e-06,
2581
+ "loss": 0.3638,
2582
+ "step": 3470
2583
+ },
2584
+ {
2585
+ "epoch": 16.651497005988023,
2586
+ "grad_norm": 92.95706939697266,
2587
+ "learning_rate": 3.6752136752136756e-06,
2588
+ "loss": 0.3444,
2589
+ "step": 3480
2590
+ },
2591
+ {
2592
+ "epoch": 16.69940119760479,
2593
+ "grad_norm": 67.28424072265625,
2594
+ "learning_rate": 3.653846153846154e-06,
2595
+ "loss": 0.5401,
2596
+ "step": 3490
2597
+ },
2598
+ {
2599
+ "epoch": 16.747305389221555,
2600
+ "grad_norm": 29.402442932128906,
2601
+ "learning_rate": 3.632478632478633e-06,
2602
+ "loss": 0.5194,
2603
+ "step": 3500
2604
+ },
2605
+ {
2606
+ "epoch": 16.795209580838325,
2607
+ "grad_norm": 38.80514144897461,
2608
+ "learning_rate": 3.6111111111111115e-06,
2609
+ "loss": 0.4502,
2610
+ "step": 3510
2611
+ },
2612
+ {
2613
+ "epoch": 16.84311377245509,
2614
+ "grad_norm": 10.038030624389648,
2615
+ "learning_rate": 3.58974358974359e-06,
2616
+ "loss": 0.1391,
2617
+ "step": 3520
2618
+ },
2619
+ {
2620
+ "epoch": 16.891017964071857,
2621
+ "grad_norm": 53.43631362915039,
2622
+ "learning_rate": 3.5683760683760687e-06,
2623
+ "loss": 0.2745,
2624
+ "step": 3530
2625
+ },
2626
+ {
2627
+ "epoch": 16.938922155688623,
2628
+ "grad_norm": 82.64073181152344,
2629
+ "learning_rate": 3.5470085470085473e-06,
2630
+ "loss": 0.2056,
2631
+ "step": 3540
2632
+ },
2633
+ {
2634
+ "epoch": 16.98682634730539,
2635
+ "grad_norm": 57.446128845214844,
2636
+ "learning_rate": 3.5256410256410263e-06,
2637
+ "loss": 0.5293,
2638
+ "step": 3550
2639
+ },
2640
+ {
2641
+ "epoch": 17.0,
2642
+ "eval_loss": 0.7794991731643677,
2643
+ "eval_macro_f1": 0.8272296780330446,
2644
+ "eval_runtime": 42.3227,
2645
+ "eval_samples_per_second": 39.482,
2646
+ "eval_steps_per_second": 4.938,
2647
+ "step": 3553
2648
+ },
2649
+ {
2650
+ "epoch": 17.033532934131735,
2651
+ "grad_norm": 47.826210021972656,
2652
+ "learning_rate": 3.5042735042735045e-06,
2653
+ "loss": 0.5691,
2654
+ "step": 3560
2655
+ },
2656
+ {
2657
+ "epoch": 17.081437125748504,
2658
+ "grad_norm": 51.37601852416992,
2659
+ "learning_rate": 3.482905982905983e-06,
2660
+ "loss": 0.3646,
2661
+ "step": 3570
2662
+ },
2663
+ {
2664
+ "epoch": 17.12934131736527,
2665
+ "grad_norm": 5.729410171508789,
2666
+ "learning_rate": 3.4615384615384617e-06,
2667
+ "loss": 0.2953,
2668
+ "step": 3580
2669
+ },
2670
+ {
2671
+ "epoch": 17.177245508982036,
2672
+ "grad_norm": 74.28618621826172,
2673
+ "learning_rate": 3.4401709401709403e-06,
2674
+ "loss": 0.4689,
2675
+ "step": 3590
2676
+ },
2677
+ {
2678
+ "epoch": 17.225149700598802,
2679
+ "grad_norm": 1.512823224067688,
2680
+ "learning_rate": 3.4188034188034193e-06,
2681
+ "loss": 0.2678,
2682
+ "step": 3600
2683
+ },
2684
+ {
2685
+ "epoch": 17.273053892215568,
2686
+ "grad_norm": 80.0973129272461,
2687
+ "learning_rate": 3.397435897435898e-06,
2688
+ "loss": 0.3399,
2689
+ "step": 3610
2690
+ },
2691
+ {
2692
+ "epoch": 17.320958083832334,
2693
+ "grad_norm": 46.759395599365234,
2694
+ "learning_rate": 3.3760683760683765e-06,
2695
+ "loss": 0.3441,
2696
+ "step": 3620
2697
+ },
2698
+ {
2699
+ "epoch": 17.368862275449104,
2700
+ "grad_norm": 41.87525939941406,
2701
+ "learning_rate": 3.3547008547008547e-06,
2702
+ "loss": 0.2421,
2703
+ "step": 3630
2704
+ },
2705
+ {
2706
+ "epoch": 17.41676646706587,
2707
+ "grad_norm": 31.34505844116211,
2708
+ "learning_rate": 3.3333333333333333e-06,
2709
+ "loss": 0.2768,
2710
+ "step": 3640
2711
+ },
2712
+ {
2713
+ "epoch": 17.464670658682635,
2714
+ "grad_norm": 71.10737609863281,
2715
+ "learning_rate": 3.311965811965812e-06,
2716
+ "loss": 0.4662,
2717
+ "step": 3650
2718
+ },
2719
+ {
2720
+ "epoch": 17.5125748502994,
2721
+ "grad_norm": 70.69206237792969,
2722
+ "learning_rate": 3.290598290598291e-06,
2723
+ "loss": 0.3731,
2724
+ "step": 3660
2725
+ },
2726
+ {
2727
+ "epoch": 17.560479041916167,
2728
+ "grad_norm": 58.422847747802734,
2729
+ "learning_rate": 3.2692307692307696e-06,
2730
+ "loss": 0.5217,
2731
+ "step": 3670
2732
+ },
2733
+ {
2734
+ "epoch": 17.608383233532933,
2735
+ "grad_norm": 11.82396125793457,
2736
+ "learning_rate": 3.247863247863248e-06,
2737
+ "loss": 0.2558,
2738
+ "step": 3680
2739
+ },
2740
+ {
2741
+ "epoch": 17.6562874251497,
2742
+ "grad_norm": 144.99168395996094,
2743
+ "learning_rate": 3.2264957264957268e-06,
2744
+ "loss": 0.3243,
2745
+ "step": 3690
2746
+ },
2747
+ {
2748
+ "epoch": 17.704191616766465,
2749
+ "grad_norm": 59.86516571044922,
2750
+ "learning_rate": 3.205128205128206e-06,
2751
+ "loss": 0.4804,
2752
+ "step": 3700
2753
+ },
2754
+ {
2755
+ "epoch": 17.752095808383235,
2756
+ "grad_norm": 62.745296478271484,
2757
+ "learning_rate": 3.183760683760684e-06,
2758
+ "loss": 0.4333,
2759
+ "step": 3710
2760
+ },
2761
+ {
2762
+ "epoch": 17.8,
2763
+ "grad_norm": 30.099609375,
2764
+ "learning_rate": 3.1623931623931626e-06,
2765
+ "loss": 0.2316,
2766
+ "step": 3720
2767
+ },
2768
+ {
2769
+ "epoch": 17.847904191616767,
2770
+ "grad_norm": 53.383583068847656,
2771
+ "learning_rate": 3.141025641025641e-06,
2772
+ "loss": 0.5045,
2773
+ "step": 3730
2774
+ },
2775
+ {
2776
+ "epoch": 17.895808383233533,
2777
+ "grad_norm": 9.510607719421387,
2778
+ "learning_rate": 3.11965811965812e-06,
2779
+ "loss": 0.2016,
2780
+ "step": 3740
2781
+ },
2782
+ {
2783
+ "epoch": 17.9437125748503,
2784
+ "grad_norm": 41.073726654052734,
2785
+ "learning_rate": 3.098290598290599e-06,
2786
+ "loss": 0.4775,
2787
+ "step": 3750
2788
+ },
2789
+ {
2790
+ "epoch": 17.991616766467065,
2791
+ "grad_norm": 18.685773849487305,
2792
+ "learning_rate": 3.0769230769230774e-06,
2793
+ "loss": 0.3327,
2794
+ "step": 3760
2795
+ },
2796
+ {
2797
+ "epoch": 18.0,
2798
+ "eval_loss": 0.6723179817199707,
2799
+ "eval_macro_f1": 0.8484331568013577,
2800
+ "eval_runtime": 38.5624,
2801
+ "eval_samples_per_second": 43.332,
2802
+ "eval_steps_per_second": 5.42,
2803
+ "step": 3762
2804
+ },
2805
+ {
2806
+ "epoch": 18.038323353293414,
2807
+ "grad_norm": 28.627132415771484,
2808
+ "learning_rate": 3.055555555555556e-06,
2809
+ "loss": 0.3131,
2810
+ "step": 3770
2811
+ },
2812
+ {
2813
+ "epoch": 18.08622754491018,
2814
+ "grad_norm": 22.602691650390625,
2815
+ "learning_rate": 3.0341880341880342e-06,
2816
+ "loss": 0.1872,
2817
+ "step": 3780
2818
+ },
2819
+ {
2820
+ "epoch": 18.134131736526946,
2821
+ "grad_norm": 70.41796112060547,
2822
+ "learning_rate": 3.012820512820513e-06,
2823
+ "loss": 0.2575,
2824
+ "step": 3790
2825
+ },
2826
+ {
2827
+ "epoch": 18.182035928143712,
2828
+ "grad_norm": 52.98373031616211,
2829
+ "learning_rate": 2.9914529914529914e-06,
2830
+ "loss": 0.366,
2831
+ "step": 3800
2832
+ },
2833
+ {
2834
+ "epoch": 18.229940119760478,
2835
+ "grad_norm": 76.51351928710938,
2836
+ "learning_rate": 2.9700854700854705e-06,
2837
+ "loss": 0.3965,
2838
+ "step": 3810
2839
+ },
2840
+ {
2841
+ "epoch": 18.277844311377244,
2842
+ "grad_norm": 57.413387298583984,
2843
+ "learning_rate": 2.948717948717949e-06,
2844
+ "loss": 0.1353,
2845
+ "step": 3820
2846
+ },
2847
+ {
2848
+ "epoch": 18.325748502994013,
2849
+ "grad_norm": 38.45537185668945,
2850
+ "learning_rate": 2.9273504273504277e-06,
2851
+ "loss": 0.2277,
2852
+ "step": 3830
2853
+ },
2854
+ {
2855
+ "epoch": 18.37365269461078,
2856
+ "grad_norm": 76.71987915039062,
2857
+ "learning_rate": 2.9059829059829063e-06,
2858
+ "loss": 0.4852,
2859
+ "step": 3840
2860
+ },
2861
+ {
2862
+ "epoch": 18.421556886227545,
2863
+ "grad_norm": 13.536568641662598,
2864
+ "learning_rate": 2.8846153846153845e-06,
2865
+ "loss": 0.2461,
2866
+ "step": 3850
2867
+ },
2868
+ {
2869
+ "epoch": 18.46946107784431,
2870
+ "grad_norm": 63.02874755859375,
2871
+ "learning_rate": 2.8632478632478635e-06,
2872
+ "loss": 0.3511,
2873
+ "step": 3860
2874
+ },
2875
+ {
2876
+ "epoch": 18.517365269461077,
2877
+ "grad_norm": 98.7140884399414,
2878
+ "learning_rate": 2.841880341880342e-06,
2879
+ "loss": 0.2256,
2880
+ "step": 3870
2881
+ },
2882
+ {
2883
+ "epoch": 18.565269461077843,
2884
+ "grad_norm": 35.94133758544922,
2885
+ "learning_rate": 2.8205128205128207e-06,
2886
+ "loss": 0.1746,
2887
+ "step": 3880
2888
+ },
2889
+ {
2890
+ "epoch": 18.61317365269461,
2891
+ "grad_norm": 87.98931121826172,
2892
+ "learning_rate": 2.7991452991452993e-06,
2893
+ "loss": 0.5013,
2894
+ "step": 3890
2895
+ },
2896
+ {
2897
+ "epoch": 18.66107784431138,
2898
+ "grad_norm": 1.2747302055358887,
2899
+ "learning_rate": 2.7777777777777783e-06,
2900
+ "loss": 0.3938,
2901
+ "step": 3900
2902
+ },
2903
+ {
2904
+ "epoch": 18.708982035928145,
2905
+ "grad_norm": 0.7769640684127808,
2906
+ "learning_rate": 2.756410256410257e-06,
2907
+ "loss": 0.5851,
2908
+ "step": 3910
2909
+ },
2910
+ {
2911
+ "epoch": 18.75688622754491,
2912
+ "grad_norm": 65.76081085205078,
2913
+ "learning_rate": 2.7350427350427355e-06,
2914
+ "loss": 0.4233,
2915
+ "step": 3920
2916
+ },
2917
+ {
2918
+ "epoch": 18.804790419161677,
2919
+ "grad_norm": 4.068265438079834,
2920
+ "learning_rate": 2.7136752136752137e-06,
2921
+ "loss": 0.3455,
2922
+ "step": 3930
2923
+ },
2924
+ {
2925
+ "epoch": 18.852694610778443,
2926
+ "grad_norm": 1912.1357421875,
2927
+ "learning_rate": 2.6923076923076923e-06,
2928
+ "loss": 0.6044,
2929
+ "step": 3940
2930
+ },
2931
+ {
2932
+ "epoch": 18.90059880239521,
2933
+ "grad_norm": 20.534839630126953,
2934
+ "learning_rate": 2.670940170940171e-06,
2935
+ "loss": 0.1716,
2936
+ "step": 3950
2937
+ },
2938
+ {
2939
+ "epoch": 18.948502994011974,
2940
+ "grad_norm": 77.07267761230469,
2941
+ "learning_rate": 2.64957264957265e-06,
2942
+ "loss": 0.2453,
2943
+ "step": 3960
2944
+ },
2945
+ {
2946
+ "epoch": 18.996407185628744,
2947
+ "grad_norm": 89.15924072265625,
2948
+ "learning_rate": 2.6282051282051286e-06,
2949
+ "loss": 0.5155,
2950
+ "step": 3970
2951
+ },
2952
+ {
2953
+ "epoch": 19.0,
2954
+ "eval_loss": 0.7802789807319641,
2955
+ "eval_macro_f1": 0.8321765860916699,
2956
+ "eval_runtime": 41.5052,
2957
+ "eval_samples_per_second": 40.26,
2958
+ "eval_steps_per_second": 5.036,
2959
+ "step": 3971
2960
+ },
2961
+ {
2962
+ "epoch": 19.04311377245509,
2963
+ "grad_norm": 119.42920684814453,
2964
+ "learning_rate": 2.606837606837607e-06,
2965
+ "loss": 0.4741,
2966
+ "step": 3980
2967
+ },
2968
+ {
2969
+ "epoch": 19.091017964071856,
2970
+ "grad_norm": 48.56228256225586,
2971
+ "learning_rate": 2.5854700854700858e-06,
2972
+ "loss": 0.3726,
2973
+ "step": 3990
2974
+ },
2975
+ {
2976
+ "epoch": 19.138922155688622,
2977
+ "grad_norm": 22.457124710083008,
2978
+ "learning_rate": 2.564102564102564e-06,
2979
+ "loss": 0.3174,
2980
+ "step": 4000
2981
+ },
2982
+ {
2983
+ "epoch": 19.186826347305388,
2984
+ "grad_norm": 56.32841873168945,
2985
+ "learning_rate": 2.542735042735043e-06,
2986
+ "loss": 0.2022,
2987
+ "step": 4010
2988
+ },
2989
+ {
2990
+ "epoch": 19.234730538922157,
2991
+ "grad_norm": 28.47975730895996,
2992
+ "learning_rate": 2.5213675213675216e-06,
2993
+ "loss": 0.2525,
2994
+ "step": 4020
2995
+ },
2996
+ {
2997
+ "epoch": 19.282634730538923,
2998
+ "grad_norm": 17.601634979248047,
2999
+ "learning_rate": 2.5e-06,
3000
+ "loss": 0.2316,
3001
+ "step": 4030
3002
+ },
3003
+ {
3004
+ "epoch": 19.33053892215569,
3005
+ "grad_norm": 47.431663513183594,
3006
+ "learning_rate": 2.478632478632479e-06,
3007
+ "loss": 0.3101,
3008
+ "step": 4040
3009
+ },
3010
+ {
3011
+ "epoch": 19.378443113772455,
3012
+ "grad_norm": 77.91453552246094,
3013
+ "learning_rate": 2.4572649572649574e-06,
3014
+ "loss": 0.4441,
3015
+ "step": 4050
3016
+ },
3017
+ {
3018
+ "epoch": 19.42634730538922,
3019
+ "grad_norm": 78.02772521972656,
3020
+ "learning_rate": 2.435897435897436e-06,
3021
+ "loss": 0.3817,
3022
+ "step": 4060
3023
+ },
3024
+ {
3025
+ "epoch": 19.474251497005987,
3026
+ "grad_norm": 20.506948471069336,
3027
+ "learning_rate": 2.4145299145299146e-06,
3028
+ "loss": 0.2292,
3029
+ "step": 4070
3030
+ },
3031
+ {
3032
+ "epoch": 19.522155688622753,
3033
+ "grad_norm": 16.479249954223633,
3034
+ "learning_rate": 2.3931623931623937e-06,
3035
+ "loss": 0.5324,
3036
+ "step": 4080
3037
+ },
3038
+ {
3039
+ "epoch": 19.57005988023952,
3040
+ "grad_norm": 30.338481903076172,
3041
+ "learning_rate": 2.371794871794872e-06,
3042
+ "loss": 0.2052,
3043
+ "step": 4090
3044
+ },
3045
+ {
3046
+ "epoch": 19.61796407185629,
3047
+ "grad_norm": 57.33430480957031,
3048
+ "learning_rate": 2.3504273504273504e-06,
3049
+ "loss": 0.2965,
3050
+ "step": 4100
3051
+ },
3052
+ {
3053
+ "epoch": 19.665868263473055,
3054
+ "grad_norm": 52.898719787597656,
3055
+ "learning_rate": 2.3290598290598295e-06,
3056
+ "loss": 0.2153,
3057
+ "step": 4110
3058
+ },
3059
+ {
3060
+ "epoch": 19.71377245508982,
3061
+ "grad_norm": 13.877717971801758,
3062
+ "learning_rate": 2.307692307692308e-06,
3063
+ "loss": 0.1302,
3064
+ "step": 4120
3065
+ },
3066
+ {
3067
+ "epoch": 19.761676646706587,
3068
+ "grad_norm": 82.69320678710938,
3069
+ "learning_rate": 2.2863247863247863e-06,
3070
+ "loss": 0.1914,
3071
+ "step": 4130
3072
+ },
3073
+ {
3074
+ "epoch": 19.809580838323352,
3075
+ "grad_norm": 69.18468475341797,
3076
+ "learning_rate": 2.2649572649572653e-06,
3077
+ "loss": 0.2875,
3078
+ "step": 4140
3079
+ },
3080
+ {
3081
+ "epoch": 19.85748502994012,
3082
+ "grad_norm": 60.78540802001953,
3083
+ "learning_rate": 2.243589743589744e-06,
3084
+ "loss": 0.1801,
3085
+ "step": 4150
3086
+ },
3087
+ {
3088
+ "epoch": 19.905389221556888,
3089
+ "grad_norm": 32.22920227050781,
3090
+ "learning_rate": 2.222222222222222e-06,
3091
+ "loss": 0.3202,
3092
+ "step": 4160
3093
+ },
3094
+ {
3095
+ "epoch": 19.953293413173654,
3096
+ "grad_norm": 65.0261459350586,
3097
+ "learning_rate": 2.200854700854701e-06,
3098
+ "loss": 0.1849,
3099
+ "step": 4170
3100
+ },
3101
+ {
3102
+ "epoch": 20.0,
3103
+ "grad_norm": 0.3435074985027313,
3104
+ "learning_rate": 2.1794871794871797e-06,
3105
+ "loss": 0.3044,
3106
+ "step": 4180
3107
+ },
3108
+ {
3109
+ "epoch": 20.0,
3110
+ "eval_loss": 0.7713411450386047,
3111
+ "eval_macro_f1": 0.844977734683617,
3112
+ "eval_runtime": 50.3873,
3113
+ "eval_samples_per_second": 33.163,
3114
+ "eval_steps_per_second": 4.148,
3115
+ "step": 4180
3116
+ },
3117
+ {
3118
+ "epoch": 20.047904191616766,
3119
+ "grad_norm": 40.881629943847656,
3120
+ "learning_rate": 2.1581196581196583e-06,
3121
+ "loss": 0.3328,
3122
+ "step": 4190
3123
+ },
3124
+ {
3125
+ "epoch": 20.095808383233532,
3126
+ "grad_norm": 19.17068862915039,
3127
+ "learning_rate": 2.136752136752137e-06,
3128
+ "loss": 0.3443,
3129
+ "step": 4200
3130
+ },
3131
+ {
3132
+ "epoch": 20.143712574850298,
3133
+ "grad_norm": 33.69413375854492,
3134
+ "learning_rate": 2.1153846153846155e-06,
3135
+ "loss": 0.3359,
3136
+ "step": 4210
3137
+ },
3138
+ {
3139
+ "epoch": 20.191616766467067,
3140
+ "grad_norm": 17.18857765197754,
3141
+ "learning_rate": 2.094017094017094e-06,
3142
+ "loss": 0.3028,
3143
+ "step": 4220
3144
+ },
3145
+ {
3146
+ "epoch": 20.239520958083833,
3147
+ "grad_norm": 57.64948272705078,
3148
+ "learning_rate": 2.072649572649573e-06,
3149
+ "loss": 0.3474,
3150
+ "step": 4230
3151
+ },
3152
+ {
3153
+ "epoch": 20.2874251497006,
3154
+ "grad_norm": 0.19221803545951843,
3155
+ "learning_rate": 2.0512820512820513e-06,
3156
+ "loss": 0.3006,
3157
+ "step": 4240
3158
+ },
3159
+ {
3160
+ "epoch": 20.335329341317365,
3161
+ "grad_norm": 7.332968235015869,
3162
+ "learning_rate": 2.02991452991453e-06,
3163
+ "loss": 0.3067,
3164
+ "step": 4250
3165
+ },
3166
+ {
3167
+ "epoch": 20.38323353293413,
3168
+ "grad_norm": 6.971332550048828,
3169
+ "learning_rate": 2.008547008547009e-06,
3170
+ "loss": 0.2523,
3171
+ "step": 4260
3172
+ },
3173
+ {
3174
+ "epoch": 20.431137724550897,
3175
+ "grad_norm": 16.12879180908203,
3176
+ "learning_rate": 1.987179487179487e-06,
3177
+ "loss": 0.1617,
3178
+ "step": 4270
3179
+ },
3180
+ {
3181
+ "epoch": 20.479041916167663,
3182
+ "grad_norm": 4.716946125030518,
3183
+ "learning_rate": 1.9658119658119658e-06,
3184
+ "loss": 0.2161,
3185
+ "step": 4280
3186
+ },
3187
+ {
3188
+ "epoch": 20.526946107784433,
3189
+ "grad_norm": 29.626829147338867,
3190
+ "learning_rate": 1.944444444444445e-06,
3191
+ "loss": 0.2236,
3192
+ "step": 4290
3193
+ },
3194
+ {
3195
+ "epoch": 20.5748502994012,
3196
+ "grad_norm": 5.962003231048584,
3197
+ "learning_rate": 1.9230769230769234e-06,
3198
+ "loss": 0.1348,
3199
+ "step": 4300
3200
+ },
3201
+ {
3202
+ "epoch": 20.622754491017965,
3203
+ "grad_norm": 22.97109603881836,
3204
+ "learning_rate": 1.9017094017094018e-06,
3205
+ "loss": 0.3083,
3206
+ "step": 4310
3207
+ },
3208
+ {
3209
+ "epoch": 20.67065868263473,
3210
+ "grad_norm": 45.447818756103516,
3211
+ "learning_rate": 1.8803418803418804e-06,
3212
+ "loss": 0.3335,
3213
+ "step": 4320
3214
+ },
3215
+ {
3216
+ "epoch": 20.718562874251496,
3217
+ "grad_norm": 205.6171417236328,
3218
+ "learning_rate": 1.8589743589743592e-06,
3219
+ "loss": 0.3472,
3220
+ "step": 4330
3221
+ },
3222
+ {
3223
+ "epoch": 20.766467065868262,
3224
+ "grad_norm": 41.59774398803711,
3225
+ "learning_rate": 1.8376068376068378e-06,
3226
+ "loss": 0.2514,
3227
+ "step": 4340
3228
+ },
3229
+ {
3230
+ "epoch": 20.81437125748503,
3231
+ "grad_norm": 37.663997650146484,
3232
+ "learning_rate": 1.8162393162393164e-06,
3233
+ "loss": 0.3328,
3234
+ "step": 4350
3235
+ },
3236
+ {
3237
+ "epoch": 20.862275449101798,
3238
+ "grad_norm": 5.474306106567383,
3239
+ "learning_rate": 1.794871794871795e-06,
3240
+ "loss": 0.3874,
3241
+ "step": 4360
3242
+ },
3243
+ {
3244
+ "epoch": 20.910179640718564,
3245
+ "grad_norm": 17.301982879638672,
3246
+ "learning_rate": 1.7735042735042736e-06,
3247
+ "loss": 0.2811,
3248
+ "step": 4370
3249
+ },
3250
+ {
3251
+ "epoch": 20.95808383233533,
3252
+ "grad_norm": 51.472537994384766,
3253
+ "learning_rate": 1.7521367521367522e-06,
3254
+ "loss": 0.4697,
3255
+ "step": 4380
3256
+ },
3257
+ {
3258
+ "epoch": 21.0,
3259
+ "eval_loss": 0.7450286746025085,
3260
+ "eval_macro_f1": 0.827007153949272,
3261
+ "eval_runtime": 57.1204,
3262
+ "eval_samples_per_second": 29.254,
3263
+ "eval_steps_per_second": 3.659,
3264
+ "step": 4389
3265
+ },
3266
+ {
3267
+ "epoch": 21.004790419161676,
3268
+ "grad_norm": 86.52445983886719,
3269
+ "learning_rate": 1.7307692307692308e-06,
3270
+ "loss": 0.4239,
3271
+ "step": 4390
3272
+ },
3273
+ {
3274
+ "epoch": 21.052694610778442,
3275
+ "grad_norm": 68.6434326171875,
3276
+ "learning_rate": 1.7094017094017097e-06,
3277
+ "loss": 0.3987,
3278
+ "step": 4400
3279
+ },
3280
+ {
3281
+ "epoch": 21.10059880239521,
3282
+ "grad_norm": 45.44715881347656,
3283
+ "learning_rate": 1.6880341880341883e-06,
3284
+ "loss": 0.2822,
3285
+ "step": 4410
3286
+ },
3287
+ {
3288
+ "epoch": 21.148502994011977,
3289
+ "grad_norm": 104.67868041992188,
3290
+ "learning_rate": 1.6666666666666667e-06,
3291
+ "loss": 0.1772,
3292
+ "step": 4420
3293
+ },
3294
+ {
3295
+ "epoch": 21.196407185628743,
3296
+ "grad_norm": 35.29579162597656,
3297
+ "learning_rate": 1.6452991452991455e-06,
3298
+ "loss": 0.2434,
3299
+ "step": 4430
3300
+ },
3301
+ {
3302
+ "epoch": 21.24431137724551,
3303
+ "grad_norm": 77.53510284423828,
3304
+ "learning_rate": 1.623931623931624e-06,
3305
+ "loss": 0.2378,
3306
+ "step": 4440
3307
+ },
3308
+ {
3309
+ "epoch": 21.292215568862275,
3310
+ "grad_norm": 67.75172424316406,
3311
+ "learning_rate": 1.602564102564103e-06,
3312
+ "loss": 0.3755,
3313
+ "step": 4450
3314
+ },
3315
+ {
3316
+ "epoch": 21.34011976047904,
3317
+ "grad_norm": 8.150789260864258,
3318
+ "learning_rate": 1.5811965811965813e-06,
3319
+ "loss": 0.2303,
3320
+ "step": 4460
3321
+ },
3322
+ {
3323
+ "epoch": 21.388023952095807,
3324
+ "grad_norm": 57.804100036621094,
3325
+ "learning_rate": 1.55982905982906e-06,
3326
+ "loss": 0.3705,
3327
+ "step": 4470
3328
+ },
3329
+ {
3330
+ "epoch": 21.435928143712573,
3331
+ "grad_norm": 12.233683586120605,
3332
+ "learning_rate": 1.5384615384615387e-06,
3333
+ "loss": 0.1344,
3334
+ "step": 4480
3335
+ },
3336
+ {
3337
+ "epoch": 21.483832335329343,
3338
+ "grad_norm": 48.35832595825195,
3339
+ "learning_rate": 1.5170940170940171e-06,
3340
+ "loss": 0.1988,
3341
+ "step": 4490
3342
+ },
3343
+ {
3344
+ "epoch": 21.53173652694611,
3345
+ "grad_norm": 112.48056030273438,
3346
+ "learning_rate": 1.4957264957264957e-06,
3347
+ "loss": 0.2977,
3348
+ "step": 4500
3349
+ },
3350
+ {
3351
+ "epoch": 21.579640718562874,
3352
+ "grad_norm": 97.60857391357422,
3353
+ "learning_rate": 1.4743589743589745e-06,
3354
+ "loss": 0.2382,
3355
+ "step": 4510
3356
+ },
3357
+ {
3358
+ "epoch": 21.62754491017964,
3359
+ "grad_norm": 4.879365921020508,
3360
+ "learning_rate": 1.4529914529914531e-06,
3361
+ "loss": 0.2374,
3362
+ "step": 4520
3363
+ },
3364
+ {
3365
+ "epoch": 21.675449101796406,
3366
+ "grad_norm": 60.12843704223633,
3367
+ "learning_rate": 1.4316239316239317e-06,
3368
+ "loss": 0.1913,
3369
+ "step": 4530
3370
+ },
3371
+ {
3372
+ "epoch": 21.723353293413172,
3373
+ "grad_norm": 39.34722900390625,
3374
+ "learning_rate": 1.4102564102564104e-06,
3375
+ "loss": 0.1366,
3376
+ "step": 4540
3377
+ },
3378
+ {
3379
+ "epoch": 21.771257485029942,
3380
+ "grad_norm": 0.2388223111629486,
3381
+ "learning_rate": 1.3888888888888892e-06,
3382
+ "loss": 0.1015,
3383
+ "step": 4550
3384
+ },
3385
+ {
3386
+ "epoch": 21.819161676646708,
3387
+ "grad_norm": 63.66411209106445,
3388
+ "learning_rate": 1.3675213675213678e-06,
3389
+ "loss": 0.3127,
3390
+ "step": 4560
3391
+ },
3392
+ {
3393
+ "epoch": 21.867065868263474,
3394
+ "grad_norm": 4.219015598297119,
3395
+ "learning_rate": 1.3461538461538462e-06,
3396
+ "loss": 0.3047,
3397
+ "step": 4570
3398
+ },
3399
+ {
3400
+ "epoch": 21.91497005988024,
3401
+ "grad_norm": 55.83110427856445,
3402
+ "learning_rate": 1.324786324786325e-06,
3403
+ "loss": 0.2811,
3404
+ "step": 4580
3405
+ },
3406
+ {
3407
+ "epoch": 21.962874251497006,
3408
+ "grad_norm": 34.60869216918945,
3409
+ "learning_rate": 1.3034188034188036e-06,
3410
+ "loss": 0.1733,
3411
+ "step": 4590
3412
+ },
3413
+ {
3414
+ "epoch": 22.0,
3415
+ "eval_loss": 0.745612621307373,
3416
+ "eval_macro_f1": 0.8567568925713644,
3417
+ "eval_runtime": 53.3616,
3418
+ "eval_samples_per_second": 31.315,
3419
+ "eval_steps_per_second": 3.917,
3420
+ "step": 4598
3421
+ }
3422
+ ],
3423
+ "logging_steps": 10,
3424
+ "max_steps": 5200,
3425
+ "num_input_tokens_seen": 0,
3426
+ "num_train_epochs": 25,
3427
+ "save_steps": 500,
3428
+ "stateful_callbacks": {
3429
+ "TrainerControl": {
3430
+ "args": {
3431
+ "should_epoch_stop": false,
3432
+ "should_evaluate": false,
3433
+ "should_log": false,
3434
+ "should_save": true,
3435
+ "should_training_stop": false
3436
+ },
3437
+ "attributes": {}
3438
+ }
3439
+ },
3440
+ "total_flos": 5.273478692599549e+19,
3441
+ "train_batch_size": 8,
3442
+ "trial_name": null,
3443
+ "trial_params": null
3444
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07d2927ed52956b28603c3b8ad5a638f86a4500d220b3ad8e422301e3cccaed6
3
+ size 5304