Training in progress, epoch 2, checkpoint
Browse files- last-checkpoint/config.json +89 -0
- last-checkpoint/model.safetensors +3 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/preprocessor_config.json +23 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/trainer_state.json +1236 -0
- last-checkpoint/training_args.bin +3 -0
last-checkpoint/config.json
ADDED
|
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/swin-tiny-patch4-window7-224",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"SwinForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"depths": [
|
| 8 |
+
2,
|
| 9 |
+
2,
|
| 10 |
+
6,
|
| 11 |
+
2
|
| 12 |
+
],
|
| 13 |
+
"drop_path_rate": 0.1,
|
| 14 |
+
"embed_dim": 96,
|
| 15 |
+
"encoder_stride": 32,
|
| 16 |
+
"hidden_act": "gelu",
|
| 17 |
+
"hidden_dropout_prob": 0.0,
|
| 18 |
+
"hidden_size": 768,
|
| 19 |
+
"id2label": {
|
| 20 |
+
"0": "e",
|
| 21 |
+
"1": "es",
|
| 22 |
+
"2": "n",
|
| 23 |
+
"3": "ne",
|
| 24 |
+
"4": "nes",
|
| 25 |
+
"5": "normal",
|
| 26 |
+
"6": "ns",
|
| 27 |
+
"7": "nw",
|
| 28 |
+
"8": "nwe",
|
| 29 |
+
"9": "nwes",
|
| 30 |
+
"10": "nws",
|
| 31 |
+
"11": "s",
|
| 32 |
+
"12": "w",
|
| 33 |
+
"13": "we",
|
| 34 |
+
"14": "wes",
|
| 35 |
+
"15": "ws"
|
| 36 |
+
},
|
| 37 |
+
"image_size": 224,
|
| 38 |
+
"initializer_range": 0.02,
|
| 39 |
+
"label2id": {
|
| 40 |
+
"e": 0,
|
| 41 |
+
"es": 1,
|
| 42 |
+
"n": 2,
|
| 43 |
+
"ne": 3,
|
| 44 |
+
"nes": 4,
|
| 45 |
+
"normal": 5,
|
| 46 |
+
"ns": 6,
|
| 47 |
+
"nw": 7,
|
| 48 |
+
"nwe": 8,
|
| 49 |
+
"nwes": 9,
|
| 50 |
+
"nws": 10,
|
| 51 |
+
"s": 11,
|
| 52 |
+
"w": 12,
|
| 53 |
+
"we": 13,
|
| 54 |
+
"wes": 14,
|
| 55 |
+
"ws": 15
|
| 56 |
+
},
|
| 57 |
+
"layer_norm_eps": 1e-05,
|
| 58 |
+
"mlp_ratio": 4.0,
|
| 59 |
+
"model_type": "swin",
|
| 60 |
+
"num_channels": 3,
|
| 61 |
+
"num_heads": [
|
| 62 |
+
3,
|
| 63 |
+
6,
|
| 64 |
+
12,
|
| 65 |
+
24
|
| 66 |
+
],
|
| 67 |
+
"num_layers": 4,
|
| 68 |
+
"out_features": [
|
| 69 |
+
"stage4"
|
| 70 |
+
],
|
| 71 |
+
"out_indices": [
|
| 72 |
+
4
|
| 73 |
+
],
|
| 74 |
+
"patch_size": 4,
|
| 75 |
+
"path_norm": true,
|
| 76 |
+
"problem_type": "single_label_classification",
|
| 77 |
+
"qkv_bias": true,
|
| 78 |
+
"stage_names": [
|
| 79 |
+
"stem",
|
| 80 |
+
"stage1",
|
| 81 |
+
"stage2",
|
| 82 |
+
"stage3",
|
| 83 |
+
"stage4"
|
| 84 |
+
],
|
| 85 |
+
"torch_dtype": "float32",
|
| 86 |
+
"transformers_version": "4.47.1",
|
| 87 |
+
"use_absolute_embeddings": false,
|
| 88 |
+
"window_size": 7
|
| 89 |
+
}
|
last-checkpoint/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56d50c4af00394a47e8221cf29d52b872eff9ba711b81a9e3bff57df7ecada2b
|
| 3 |
+
size 110385904
|
last-checkpoint/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0c0ec9fc16941b41120ee6b8a63c4ef3e6018d08bc56600d663ba1ff57e540f
|
| 3 |
+
size 220436730
|
last-checkpoint/preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.485,
|
| 8 |
+
0.456,
|
| 9 |
+
0.406
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.229,
|
| 14 |
+
0.224,
|
| 15 |
+
0.225
|
| 16 |
+
],
|
| 17 |
+
"resample": 3,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e91323fc4954bfb304d75ccd1eaa491100df7060e8351a39f1d8ab42ea15ade9
|
| 3 |
+
size 14244
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18695d2fb12e6cf8ee010a2d0b8fecd135e4aa9b9c6e2717dab0c8212f0b4eb3
|
| 3 |
+
size 1064
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,1236 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.0,
|
| 3 |
+
"best_model_checkpoint": "/content/drive/MyDrive/Colab Notebooks/16_label_check_point/checkpoint-563",
|
| 4 |
+
"epoch": 2.999111111111111,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1686,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.017777777777777778,
|
| 13 |
+
"grad_norm": 10.457254409790039,
|
| 14 |
+
"learning_rate": 1.7793594306049826e-06,
|
| 15 |
+
"loss": 11.2334,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.035555555555555556,
|
| 20 |
+
"grad_norm": 6.316349506378174,
|
| 21 |
+
"learning_rate": 3.558718861209965e-06,
|
| 22 |
+
"loss": 11.1807,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05333333333333334,
|
| 27 |
+
"grad_norm": 5.264202117919922,
|
| 28 |
+
"learning_rate": 5.338078291814947e-06,
|
| 29 |
+
"loss": 11.1463,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07111111111111111,
|
| 34 |
+
"grad_norm": 5.205317497253418,
|
| 35 |
+
"learning_rate": 7.11743772241993e-06,
|
| 36 |
+
"loss": 11.0929,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08888888888888889,
|
| 41 |
+
"grad_norm": 4.696351528167725,
|
| 42 |
+
"learning_rate": 8.896797153024912e-06,
|
| 43 |
+
"loss": 11.1015,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10666666666666667,
|
| 48 |
+
"grad_norm": 5.9699320793151855,
|
| 49 |
+
"learning_rate": 1.0676156583629894e-05,
|
| 50 |
+
"loss": 11.0795,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12444444444444444,
|
| 55 |
+
"grad_norm": 7.235191822052002,
|
| 56 |
+
"learning_rate": 1.2455516014234877e-05,
|
| 57 |
+
"loss": 11.046,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14222222222222222,
|
| 62 |
+
"grad_norm": 14.865583419799805,
|
| 63 |
+
"learning_rate": 1.423487544483986e-05,
|
| 64 |
+
"loss": 11.0113,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 17.412696838378906,
|
| 70 |
+
"learning_rate": 1.601423487544484e-05,
|
| 71 |
+
"loss": 10.8502,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.17777777777777778,
|
| 76 |
+
"grad_norm": 20.41786003112793,
|
| 77 |
+
"learning_rate": 1.7793594306049825e-05,
|
| 78 |
+
"loss": 10.5185,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19555555555555557,
|
| 83 |
+
"grad_norm": 24.58249282836914,
|
| 84 |
+
"learning_rate": 1.9572953736654805e-05,
|
| 85 |
+
"loss": 9.8301,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21333333333333335,
|
| 90 |
+
"grad_norm": 43.27064895629883,
|
| 91 |
+
"learning_rate": 2.135231316725979e-05,
|
| 92 |
+
"loss": 8.8636,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2311111111111111,
|
| 97 |
+
"grad_norm": 46.2359733581543,
|
| 98 |
+
"learning_rate": 2.313167259786477e-05,
|
| 99 |
+
"loss": 7.4813,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24888888888888888,
|
| 104 |
+
"grad_norm": 71.36530303955078,
|
| 105 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 106 |
+
"loss": 6.0609,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.26666666666666666,
|
| 111 |
+
"grad_norm": 61.219024658203125,
|
| 112 |
+
"learning_rate": 2.669039145907473e-05,
|
| 113 |
+
"loss": 4.5934,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28444444444444444,
|
| 118 |
+
"grad_norm": 52.64271545410156,
|
| 119 |
+
"learning_rate": 2.846975088967972e-05,
|
| 120 |
+
"loss": 3.3049,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3022222222222222,
|
| 125 |
+
"grad_norm": 40.045623779296875,
|
| 126 |
+
"learning_rate": 3.02491103202847e-05,
|
| 127 |
+
"loss": 2.0759,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32,
|
| 132 |
+
"grad_norm": 32.63826370239258,
|
| 133 |
+
"learning_rate": 3.202846975088968e-05,
|
| 134 |
+
"loss": 1.2791,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3377777777777778,
|
| 139 |
+
"grad_norm": 39.16119384765625,
|
| 140 |
+
"learning_rate": 3.380782918149467e-05,
|
| 141 |
+
"loss": 0.9052,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.35555555555555557,
|
| 146 |
+
"grad_norm": 36.06990051269531,
|
| 147 |
+
"learning_rate": 3.558718861209965e-05,
|
| 148 |
+
"loss": 0.6508,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37333333333333335,
|
| 153 |
+
"grad_norm": 22.00220489501953,
|
| 154 |
+
"learning_rate": 3.736654804270463e-05,
|
| 155 |
+
"loss": 0.6293,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39111111111111113,
|
| 160 |
+
"grad_norm": 27.334341049194336,
|
| 161 |
+
"learning_rate": 3.914590747330961e-05,
|
| 162 |
+
"loss": 0.5774,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4088888888888889,
|
| 167 |
+
"grad_norm": 21.130746841430664,
|
| 168 |
+
"learning_rate": 4.09252669039146e-05,
|
| 169 |
+
"loss": 0.4318,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4266666666666667,
|
| 174 |
+
"grad_norm": 21.37102508544922,
|
| 175 |
+
"learning_rate": 4.270462633451958e-05,
|
| 176 |
+
"loss": 0.414,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4444444444444444,
|
| 181 |
+
"grad_norm": 86.8355712890625,
|
| 182 |
+
"learning_rate": 4.448398576512456e-05,
|
| 183 |
+
"loss": 0.4338,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4622222222222222,
|
| 188 |
+
"grad_norm": 9.731348037719727,
|
| 189 |
+
"learning_rate": 4.626334519572954e-05,
|
| 190 |
+
"loss": 0.3978,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.48,
|
| 195 |
+
"grad_norm": 23.59229278564453,
|
| 196 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 197 |
+
"loss": 0.2971,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.49777777777777776,
|
| 202 |
+
"grad_norm": 10.458292961120605,
|
| 203 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 204 |
+
"loss": 0.3015,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5155555555555555,
|
| 209 |
+
"grad_norm": 13.044571876525879,
|
| 210 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 211 |
+
"loss": 0.2013,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5333333333333333,
|
| 216 |
+
"grad_norm": 30.592700958251953,
|
| 217 |
+
"learning_rate": 4.962435745353895e-05,
|
| 218 |
+
"loss": 0.182,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5511111111111111,
|
| 223 |
+
"grad_norm": 26.582555770874023,
|
| 224 |
+
"learning_rate": 4.9426650850138396e-05,
|
| 225 |
+
"loss": 0.2276,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5688888888888889,
|
| 230 |
+
"grad_norm": 20.704526901245117,
|
| 231 |
+
"learning_rate": 4.9228944246737844e-05,
|
| 232 |
+
"loss": 0.1836,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5866666666666667,
|
| 237 |
+
"grad_norm": 29.400476455688477,
|
| 238 |
+
"learning_rate": 4.903123764333729e-05,
|
| 239 |
+
"loss": 0.2921,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6044444444444445,
|
| 244 |
+
"grad_norm": 23.031789779663086,
|
| 245 |
+
"learning_rate": 4.8833531039936733e-05,
|
| 246 |
+
"loss": 0.0848,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6222222222222222,
|
| 251 |
+
"grad_norm": 0.9449447393417358,
|
| 252 |
+
"learning_rate": 4.863582443653618e-05,
|
| 253 |
+
"loss": 0.0741,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.64,
|
| 258 |
+
"grad_norm": 20.334348678588867,
|
| 259 |
+
"learning_rate": 4.843811783313563e-05,
|
| 260 |
+
"loss": 0.1798,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6577777777777778,
|
| 265 |
+
"grad_norm": 13.533489227294922,
|
| 266 |
+
"learning_rate": 4.824041122973508e-05,
|
| 267 |
+
"loss": 0.1007,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6755555555555556,
|
| 272 |
+
"grad_norm": 1.5982263088226318,
|
| 273 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 274 |
+
"loss": 0.1016,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6933333333333334,
|
| 279 |
+
"grad_norm": 2.323336362838745,
|
| 280 |
+
"learning_rate": 4.784499802293397e-05,
|
| 281 |
+
"loss": 0.1131,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7111111111111111,
|
| 286 |
+
"grad_norm": 24.458276748657227,
|
| 287 |
+
"learning_rate": 4.7647291419533415e-05,
|
| 288 |
+
"loss": 0.069,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7288888888888889,
|
| 293 |
+
"grad_norm": 29.530794143676758,
|
| 294 |
+
"learning_rate": 4.7449584816132864e-05,
|
| 295 |
+
"loss": 0.2415,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7466666666666667,
|
| 300 |
+
"grad_norm": 2.0636749267578125,
|
| 301 |
+
"learning_rate": 4.725187821273231e-05,
|
| 302 |
+
"loss": 0.1173,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7644444444444445,
|
| 307 |
+
"grad_norm": 2.225900888442993,
|
| 308 |
+
"learning_rate": 4.705417160933175e-05,
|
| 309 |
+
"loss": 0.0918,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7822222222222223,
|
| 314 |
+
"grad_norm": 7.136375904083252,
|
| 315 |
+
"learning_rate": 4.68564650059312e-05,
|
| 316 |
+
"loss": 0.1589,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8,
|
| 321 |
+
"grad_norm": 2.6402971744537354,
|
| 322 |
+
"learning_rate": 4.665875840253064e-05,
|
| 323 |
+
"loss": 0.2547,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8177777777777778,
|
| 328 |
+
"grad_norm": 12.275321960449219,
|
| 329 |
+
"learning_rate": 4.64610517991301e-05,
|
| 330 |
+
"loss": 0.1082,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8355555555555556,
|
| 335 |
+
"grad_norm": 0.6636475920677185,
|
| 336 |
+
"learning_rate": 4.626334519572954e-05,
|
| 337 |
+
"loss": 0.0986,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8533333333333334,
|
| 342 |
+
"grad_norm": 27.784332275390625,
|
| 343 |
+
"learning_rate": 4.606563859232899e-05,
|
| 344 |
+
"loss": 0.1056,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8711111111111111,
|
| 349 |
+
"grad_norm": 22.047527313232422,
|
| 350 |
+
"learning_rate": 4.586793198892843e-05,
|
| 351 |
+
"loss": 0.1654,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8888888888888888,
|
| 356 |
+
"grad_norm": 15.613493919372559,
|
| 357 |
+
"learning_rate": 4.5670225385527876e-05,
|
| 358 |
+
"loss": 0.1517,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9066666666666666,
|
| 363 |
+
"grad_norm": 20.039813995361328,
|
| 364 |
+
"learning_rate": 4.5472518782127324e-05,
|
| 365 |
+
"loss": 0.1283,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9244444444444444,
|
| 370 |
+
"grad_norm": 21.469423294067383,
|
| 371 |
+
"learning_rate": 4.527481217872677e-05,
|
| 372 |
+
"loss": 0.0707,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9422222222222222,
|
| 377 |
+
"grad_norm": 38.251953125,
|
| 378 |
+
"learning_rate": 4.5077105575326214e-05,
|
| 379 |
+
"loss": 0.1116,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.96,
|
| 384 |
+
"grad_norm": 12.317658424377441,
|
| 385 |
+
"learning_rate": 4.487939897192566e-05,
|
| 386 |
+
"loss": 0.138,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9777777777777777,
|
| 391 |
+
"grad_norm": 5.918329238891602,
|
| 392 |
+
"learning_rate": 4.468169236852511e-05,
|
| 393 |
+
"loss": 0.1053,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9955555555555555,
|
| 398 |
+
"grad_norm": 12.31584358215332,
|
| 399 |
+
"learning_rate": 4.448398576512456e-05,
|
| 400 |
+
"loss": 0.1354,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0,
|
| 405 |
+
"eval_accuracy": 1.0,
|
| 406 |
+
"eval_loss": 8.097552927210927e-05,
|
| 407 |
+
"eval_runtime": 74.3955,
|
| 408 |
+
"eval_samples_per_second": 107.533,
|
| 409 |
+
"eval_steps_per_second": 3.36,
|
| 410 |
+
"step": 563
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.0124444444444445,
|
| 414 |
+
"grad_norm": 10.894996643066406,
|
| 415 |
+
"learning_rate": 4.4286279161724006e-05,
|
| 416 |
+
"loss": 0.1134,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0302222222222222,
|
| 421 |
+
"grad_norm": 16.037940979003906,
|
| 422 |
+
"learning_rate": 4.408857255832345e-05,
|
| 423 |
+
"loss": 0.0917,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.048,
|
| 428 |
+
"grad_norm": 0.164012148976326,
|
| 429 |
+
"learning_rate": 4.3890865954922896e-05,
|
| 430 |
+
"loss": 0.0816,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0657777777777777,
|
| 435 |
+
"grad_norm": 0.05601898953318596,
|
| 436 |
+
"learning_rate": 4.3693159351522344e-05,
|
| 437 |
+
"loss": 0.0778,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.0835555555555556,
|
| 442 |
+
"grad_norm": 9.240290641784668,
|
| 443 |
+
"learning_rate": 4.349545274812179e-05,
|
| 444 |
+
"loss": 0.0893,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1013333333333333,
|
| 449 |
+
"grad_norm": 8.448566436767578,
|
| 450 |
+
"learning_rate": 4.3297746144721233e-05,
|
| 451 |
+
"loss": 0.0905,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.1191111111111112,
|
| 456 |
+
"grad_norm": 1.9537012577056885,
|
| 457 |
+
"learning_rate": 4.310003954132068e-05,
|
| 458 |
+
"loss": 0.1561,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.1368888888888888,
|
| 463 |
+
"grad_norm": 25.13317108154297,
|
| 464 |
+
"learning_rate": 4.290233293792013e-05,
|
| 465 |
+
"loss": 0.0856,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.1546666666666667,
|
| 470 |
+
"grad_norm": 1.3837047815322876,
|
| 471 |
+
"learning_rate": 4.270462633451958e-05,
|
| 472 |
+
"loss": 0.0962,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.1724444444444444,
|
| 477 |
+
"grad_norm": 46.560367584228516,
|
| 478 |
+
"learning_rate": 4.250691973111902e-05,
|
| 479 |
+
"loss": 0.1009,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.1902222222222223,
|
| 484 |
+
"grad_norm": 42.40678787231445,
|
| 485 |
+
"learning_rate": 4.230921312771847e-05,
|
| 486 |
+
"loss": 0.1027,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.208,
|
| 491 |
+
"grad_norm": 1.2823681831359863,
|
| 492 |
+
"learning_rate": 4.211150652431791e-05,
|
| 493 |
+
"loss": 0.0609,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.2257777777777779,
|
| 498 |
+
"grad_norm": 3.9979896545410156,
|
| 499 |
+
"learning_rate": 4.1913799920917364e-05,
|
| 500 |
+
"loss": 0.0884,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.2435555555555555,
|
| 505 |
+
"grad_norm": 23.4843692779541,
|
| 506 |
+
"learning_rate": 4.1716093317516805e-05,
|
| 507 |
+
"loss": 0.049,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.2613333333333334,
|
| 512 |
+
"grad_norm": 0.9539620876312256,
|
| 513 |
+
"learning_rate": 4.151838671411625e-05,
|
| 514 |
+
"loss": 0.0481,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.279111111111111,
|
| 519 |
+
"grad_norm": 0.23285511136054993,
|
| 520 |
+
"learning_rate": 4.13206801107157e-05,
|
| 521 |
+
"loss": 0.0388,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.2968888888888888,
|
| 526 |
+
"grad_norm": 24.404285430908203,
|
| 527 |
+
"learning_rate": 4.112297350731515e-05,
|
| 528 |
+
"loss": 0.0425,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.3146666666666667,
|
| 533 |
+
"grad_norm": 3.143155813217163,
|
| 534 |
+
"learning_rate": 4.09252669039146e-05,
|
| 535 |
+
"loss": 0.0313,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.3324444444444445,
|
| 540 |
+
"grad_norm": 0.3580031991004944,
|
| 541 |
+
"learning_rate": 4.072756030051404e-05,
|
| 542 |
+
"loss": 0.0466,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.3502222222222222,
|
| 547 |
+
"grad_norm": 39.96261978149414,
|
| 548 |
+
"learning_rate": 4.052985369711349e-05,
|
| 549 |
+
"loss": 0.0706,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.3679999999999999,
|
| 554 |
+
"grad_norm": 22.012971878051758,
|
| 555 |
+
"learning_rate": 4.033214709371293e-05,
|
| 556 |
+
"loss": 0.1249,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.3857777777777778,
|
| 561 |
+
"grad_norm": 0.20229700207710266,
|
| 562 |
+
"learning_rate": 4.013444049031238e-05,
|
| 563 |
+
"loss": 0.1089,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.4035555555555557,
|
| 568 |
+
"grad_norm": 36.518348693847656,
|
| 569 |
+
"learning_rate": 3.9936733886911825e-05,
|
| 570 |
+
"loss": 0.0514,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.4213333333333333,
|
| 575 |
+
"grad_norm": 0.011868222616612911,
|
| 576 |
+
"learning_rate": 3.973902728351127e-05,
|
| 577 |
+
"loss": 0.0298,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.439111111111111,
|
| 582 |
+
"grad_norm": 23.455787658691406,
|
| 583 |
+
"learning_rate": 3.9541320680110714e-05,
|
| 584 |
+
"loss": 0.0365,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.456888888888889,
|
| 589 |
+
"grad_norm": 0.5454981923103333,
|
| 590 |
+
"learning_rate": 3.934361407671016e-05,
|
| 591 |
+
"loss": 0.0603,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.4746666666666668,
|
| 596 |
+
"grad_norm": 0.2658223509788513,
|
| 597 |
+
"learning_rate": 3.914590747330961e-05,
|
| 598 |
+
"loss": 0.0364,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4924444444444445,
|
| 603 |
+
"grad_norm": 32.6451301574707,
|
| 604 |
+
"learning_rate": 3.894820086990906e-05,
|
| 605 |
+
"loss": 0.1105,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5102222222222221,
|
| 610 |
+
"grad_norm": 0.37322714924812317,
|
| 611 |
+
"learning_rate": 3.87504942665085e-05,
|
| 612 |
+
"loss": 0.0102,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.528,
|
| 617 |
+
"grad_norm": 1.2302775382995605,
|
| 618 |
+
"learning_rate": 3.855278766310795e-05,
|
| 619 |
+
"loss": 0.093,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.545777777777778,
|
| 624 |
+
"grad_norm": 0.11981203407049179,
|
| 625 |
+
"learning_rate": 3.8355081059707396e-05,
|
| 626 |
+
"loss": 0.0721,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.5635555555555556,
|
| 631 |
+
"grad_norm": 0.09180541336536407,
|
| 632 |
+
"learning_rate": 3.8157374456306844e-05,
|
| 633 |
+
"loss": 0.0408,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.5813333333333333,
|
| 638 |
+
"grad_norm": 29.872051239013672,
|
| 639 |
+
"learning_rate": 3.7959667852906285e-05,
|
| 640 |
+
"loss": 0.0393,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5991111111111111,
|
| 645 |
+
"grad_norm": 1.0710923671722412,
|
| 646 |
+
"learning_rate": 3.7761961249505734e-05,
|
| 647 |
+
"loss": 0.0023,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.616888888888889,
|
| 652 |
+
"grad_norm": 20.386173248291016,
|
| 653 |
+
"learning_rate": 3.756425464610518e-05,
|
| 654 |
+
"loss": 0.0605,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.6346666666666667,
|
| 659 |
+
"grad_norm": 0.03605956584215164,
|
| 660 |
+
"learning_rate": 3.736654804270463e-05,
|
| 661 |
+
"loss": 0.0059,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.6524444444444444,
|
| 666 |
+
"grad_norm": 0.38812369108200073,
|
| 667 |
+
"learning_rate": 3.716884143930408e-05,
|
| 668 |
+
"loss": 0.0819,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.6702222222222223,
|
| 673 |
+
"grad_norm": 27.319766998291016,
|
| 674 |
+
"learning_rate": 3.697113483590352e-05,
|
| 675 |
+
"loss": 0.0201,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.688,
|
| 680 |
+
"grad_norm": 20.58792495727539,
|
| 681 |
+
"learning_rate": 3.677342823250297e-05,
|
| 682 |
+
"loss": 0.0678,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.7057777777777776,
|
| 687 |
+
"grad_norm": 0.33605310320854187,
|
| 688 |
+
"learning_rate": 3.6575721629102416e-05,
|
| 689 |
+
"loss": 0.0027,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.7235555555555555,
|
| 694 |
+
"grad_norm": 0.13025720417499542,
|
| 695 |
+
"learning_rate": 3.6378015025701864e-05,
|
| 696 |
+
"loss": 0.0029,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.7413333333333334,
|
| 701 |
+
"grad_norm": 31.07040023803711,
|
| 702 |
+
"learning_rate": 3.6180308422301305e-05,
|
| 703 |
+
"loss": 0.0626,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.759111111111111,
|
| 708 |
+
"grad_norm": 0.055677346885204315,
|
| 709 |
+
"learning_rate": 3.598260181890075e-05,
|
| 710 |
+
"loss": 0.0079,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.7768888888888887,
|
| 715 |
+
"grad_norm": 0.012918527238070965,
|
| 716 |
+
"learning_rate": 3.5784895215500194e-05,
|
| 717 |
+
"loss": 0.0308,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.7946666666666666,
|
| 722 |
+
"grad_norm": 12.896405220031738,
|
| 723 |
+
"learning_rate": 3.558718861209965e-05,
|
| 724 |
+
"loss": 0.0551,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.8124444444444445,
|
| 729 |
+
"grad_norm": 0.0037423851899802685,
|
| 730 |
+
"learning_rate": 3.538948200869909e-05,
|
| 731 |
+
"loss": 0.0437,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.8302222222222222,
|
| 736 |
+
"grad_norm": 0.2329370528459549,
|
| 737 |
+
"learning_rate": 3.519177540529854e-05,
|
| 738 |
+
"loss": 0.0281,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.8479999999999999,
|
| 743 |
+
"grad_norm": 0.027233602479100227,
|
| 744 |
+
"learning_rate": 3.499406880189798e-05,
|
| 745 |
+
"loss": 0.0136,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8657777777777778,
|
| 750 |
+
"grad_norm": 0.1924924999475479,
|
| 751 |
+
"learning_rate": 3.4796362198497435e-05,
|
| 752 |
+
"loss": 0.056,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.8835555555555556,
|
| 757 |
+
"grad_norm": 0.02651727944612503,
|
| 758 |
+
"learning_rate": 3.4598655595096876e-05,
|
| 759 |
+
"loss": 0.0647,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.9013333333333333,
|
| 764 |
+
"grad_norm": 38.13447570800781,
|
| 765 |
+
"learning_rate": 3.4400948991696325e-05,
|
| 766 |
+
"loss": 0.0216,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.919111111111111,
|
| 771 |
+
"grad_norm": 11.278813362121582,
|
| 772 |
+
"learning_rate": 3.420324238829577e-05,
|
| 773 |
+
"loss": 0.0395,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9368888888888889,
|
| 778 |
+
"grad_norm": 0.5866456031799316,
|
| 779 |
+
"learning_rate": 3.4005535784895214e-05,
|
| 780 |
+
"loss": 0.0433,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.9546666666666668,
|
| 785 |
+
"grad_norm": 2.6308796405792236,
|
| 786 |
+
"learning_rate": 3.380782918149467e-05,
|
| 787 |
+
"loss": 0.0118,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.9724444444444444,
|
| 792 |
+
"grad_norm": 0.1729055494070053,
|
| 793 |
+
"learning_rate": 3.361012257809411e-05,
|
| 794 |
+
"loss": 0.0251,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.9902222222222221,
|
| 799 |
+
"grad_norm": 10.363531112670898,
|
| 800 |
+
"learning_rate": 3.341241597469356e-05,
|
| 801 |
+
"loss": 0.0656,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.0,
|
| 806 |
+
"eval_accuracy": 1.0,
|
| 807 |
+
"eval_loss": 2.5387274945387617e-05,
|
| 808 |
+
"eval_runtime": 75.7395,
|
| 809 |
+
"eval_samples_per_second": 105.625,
|
| 810 |
+
"eval_steps_per_second": 3.301,
|
| 811 |
+
"step": 1126
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.010666666666667,
|
| 815 |
+
"grad_norm": 29.318574905395508,
|
| 816 |
+
"learning_rate": 3.3214709371293e-05,
|
| 817 |
+
"loss": 0.2308,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.0284444444444443,
|
| 822 |
+
"grad_norm": 67.69547271728516,
|
| 823 |
+
"learning_rate": 3.301700276789245e-05,
|
| 824 |
+
"loss": 0.0604,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.046222222222222,
|
| 829 |
+
"grad_norm": 0.18736213445663452,
|
| 830 |
+
"learning_rate": 3.2819296164491896e-05,
|
| 831 |
+
"loss": 0.0546,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.064,
|
| 836 |
+
"grad_norm": 15.664798736572266,
|
| 837 |
+
"learning_rate": 3.2621589561091344e-05,
|
| 838 |
+
"loss": 0.0304,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.081777777777778,
|
| 843 |
+
"grad_norm": 0.13788799941539764,
|
| 844 |
+
"learning_rate": 3.2423882957690785e-05,
|
| 845 |
+
"loss": 0.0029,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.0995555555555554,
|
| 850 |
+
"grad_norm": 4.969343185424805,
|
| 851 |
+
"learning_rate": 3.2226176354290234e-05,
|
| 852 |
+
"loss": 0.0099,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.1173333333333333,
|
| 857 |
+
"grad_norm": 1.4300692081451416,
|
| 858 |
+
"learning_rate": 3.202846975088968e-05,
|
| 859 |
+
"loss": 0.0181,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.135111111111111,
|
| 864 |
+
"grad_norm": 0.019833851605653763,
|
| 865 |
+
"learning_rate": 3.183076314748913e-05,
|
| 866 |
+
"loss": 0.0107,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.152888888888889,
|
| 871 |
+
"grad_norm": 0.01474601961672306,
|
| 872 |
+
"learning_rate": 3.163305654408857e-05,
|
| 873 |
+
"loss": 0.0169,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.1706666666666665,
|
| 878 |
+
"grad_norm": 0.004971742630004883,
|
| 879 |
+
"learning_rate": 3.143534994068802e-05,
|
| 880 |
+
"loss": 0.0254,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.1884444444444444,
|
| 885 |
+
"grad_norm": 0.42502257227897644,
|
| 886 |
+
"learning_rate": 3.123764333728747e-05,
|
| 887 |
+
"loss": 0.0623,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.2062222222222223,
|
| 892 |
+
"grad_norm": 30.118955612182617,
|
| 893 |
+
"learning_rate": 3.1039936733886916e-05,
|
| 894 |
+
"loss": 0.0456,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.224,
|
| 899 |
+
"grad_norm": 7.1990132331848145,
|
| 900 |
+
"learning_rate": 3.0842230130486364e-05,
|
| 901 |
+
"loss": 0.0468,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.2417777777777776,
|
| 906 |
+
"grad_norm": 0.021625256165862083,
|
| 907 |
+
"learning_rate": 3.0644523527085805e-05,
|
| 908 |
+
"loss": 0.0375,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.2595555555555555,
|
| 913 |
+
"grad_norm": 0.08051316440105438,
|
| 914 |
+
"learning_rate": 3.044681692368525e-05,
|
| 915 |
+
"loss": 0.0097,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.2773333333333334,
|
| 920 |
+
"grad_norm": 38.26923751831055,
|
| 921 |
+
"learning_rate": 3.02491103202847e-05,
|
| 922 |
+
"loss": 0.0247,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.295111111111111,
|
| 927 |
+
"grad_norm": 1.5090163946151733,
|
| 928 |
+
"learning_rate": 3.0051403716884146e-05,
|
| 929 |
+
"loss": 0.0207,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.3128888888888888,
|
| 934 |
+
"grad_norm": 14.290279388427734,
|
| 935 |
+
"learning_rate": 2.985369711348359e-05,
|
| 936 |
+
"loss": 0.0041,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.3306666666666667,
|
| 941 |
+
"grad_norm": 0.025663571432232857,
|
| 942 |
+
"learning_rate": 2.9655990510083035e-05,
|
| 943 |
+
"loss": 0.0013,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.3484444444444446,
|
| 948 |
+
"grad_norm": 13.073821067810059,
|
| 949 |
+
"learning_rate": 2.9458283906682484e-05,
|
| 950 |
+
"loss": 0.006,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.3662222222222224,
|
| 955 |
+
"grad_norm": 0.03257250785827637,
|
| 956 |
+
"learning_rate": 2.9260577303281932e-05,
|
| 957 |
+
"loss": 0.0383,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.384,
|
| 962 |
+
"grad_norm": 2.0085232257843018,
|
| 963 |
+
"learning_rate": 2.906287069988138e-05,
|
| 964 |
+
"loss": 0.0552,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.401777777777778,
|
| 969 |
+
"grad_norm": 19.939329147338867,
|
| 970 |
+
"learning_rate": 2.8865164096480825e-05,
|
| 971 |
+
"loss": 0.0714,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.4195555555555557,
|
| 976 |
+
"grad_norm": 0.028250480070710182,
|
| 977 |
+
"learning_rate": 2.866745749308027e-05,
|
| 978 |
+
"loss": 0.0227,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.437333333333333,
|
| 983 |
+
"grad_norm": 0.6903110146522522,
|
| 984 |
+
"learning_rate": 2.846975088967972e-05,
|
| 985 |
+
"loss": 0.0085,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.455111111111111,
|
| 990 |
+
"grad_norm": 0.22341494262218475,
|
| 991 |
+
"learning_rate": 2.8272044286279166e-05,
|
| 992 |
+
"loss": 0.0093,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.472888888888889,
|
| 997 |
+
"grad_norm": 0.015194721519947052,
|
| 998 |
+
"learning_rate": 2.807433768287861e-05,
|
| 999 |
+
"loss": 0.0413,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.490666666666667,
|
| 1004 |
+
"grad_norm": 0.035975806415081024,
|
| 1005 |
+
"learning_rate": 2.7876631079478055e-05,
|
| 1006 |
+
"loss": 0.0013,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.5084444444444447,
|
| 1011 |
+
"grad_norm": 0.32504796981811523,
|
| 1012 |
+
"learning_rate": 2.76789244760775e-05,
|
| 1013 |
+
"loss": 0.0547,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.526222222222222,
|
| 1018 |
+
"grad_norm": 0.34928593039512634,
|
| 1019 |
+
"learning_rate": 2.748121787267695e-05,
|
| 1020 |
+
"loss": 0.0556,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.544,
|
| 1025 |
+
"grad_norm": 0.045603252947330475,
|
| 1026 |
+
"learning_rate": 2.7283511269276396e-05,
|
| 1027 |
+
"loss": 0.0156,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.561777777777778,
|
| 1032 |
+
"grad_norm": 11.350424766540527,
|
| 1033 |
+
"learning_rate": 2.708580466587584e-05,
|
| 1034 |
+
"loss": 0.011,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.5795555555555554,
|
| 1039 |
+
"grad_norm": 0.07788264751434326,
|
| 1040 |
+
"learning_rate": 2.6888098062475286e-05,
|
| 1041 |
+
"loss": 0.0607,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.5973333333333333,
|
| 1046 |
+
"grad_norm": 0.06617221236228943,
|
| 1047 |
+
"learning_rate": 2.669039145907473e-05,
|
| 1048 |
+
"loss": 0.0102,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.615111111111111,
|
| 1053 |
+
"grad_norm": 34.64754867553711,
|
| 1054 |
+
"learning_rate": 2.6492684855674182e-05,
|
| 1055 |
+
"loss": 0.0268,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.632888888888889,
|
| 1060 |
+
"grad_norm": 9.72877311706543,
|
| 1061 |
+
"learning_rate": 2.6294978252273626e-05,
|
| 1062 |
+
"loss": 0.0051,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.6506666666666665,
|
| 1067 |
+
"grad_norm": 21.619274139404297,
|
| 1068 |
+
"learning_rate": 2.609727164887307e-05,
|
| 1069 |
+
"loss": 0.0092,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.6684444444444444,
|
| 1074 |
+
"grad_norm": 4.081634521484375,
|
| 1075 |
+
"learning_rate": 2.589956504547252e-05,
|
| 1076 |
+
"loss": 0.0369,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.6862222222222223,
|
| 1081 |
+
"grad_norm": 0.009345272555947304,
|
| 1082 |
+
"learning_rate": 2.5701858442071967e-05,
|
| 1083 |
+
"loss": 0.018,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.7039999999999997,
|
| 1088 |
+
"grad_norm": 73.03565979003906,
|
| 1089 |
+
"learning_rate": 2.5504151838671416e-05,
|
| 1090 |
+
"loss": 0.0945,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.7217777777777776,
|
| 1095 |
+
"grad_norm": 1.0712828636169434,
|
| 1096 |
+
"learning_rate": 2.530644523527086e-05,
|
| 1097 |
+
"loss": 0.0156,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.7395555555555555,
|
| 1102 |
+
"grad_norm": 0.023015221580863,
|
| 1103 |
+
"learning_rate": 2.5108738631870305e-05,
|
| 1104 |
+
"loss": 0.0408,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.7573333333333334,
|
| 1109 |
+
"grad_norm": 3.0739543437957764,
|
| 1110 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 1111 |
+
"loss": 0.0384,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.7751111111111113,
|
| 1116 |
+
"grad_norm": 0.017695285379886627,
|
| 1117 |
+
"learning_rate": 2.4713325425069198e-05,
|
| 1118 |
+
"loss": 0.034,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.7928888888888888,
|
| 1123 |
+
"grad_norm": 0.013055549003183842,
|
| 1124 |
+
"learning_rate": 2.4515618821668646e-05,
|
| 1125 |
+
"loss": 0.074,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.8106666666666666,
|
| 1130 |
+
"grad_norm": 6.4839582443237305,
|
| 1131 |
+
"learning_rate": 2.431791221826809e-05,
|
| 1132 |
+
"loss": 0.0016,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.8284444444444445,
|
| 1137 |
+
"grad_norm": 0.17747992277145386,
|
| 1138 |
+
"learning_rate": 2.412020561486754e-05,
|
| 1139 |
+
"loss": 0.0579,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.846222222222222,
|
| 1144 |
+
"grad_norm": 0.07140109688043594,
|
| 1145 |
+
"learning_rate": 2.3922499011466984e-05,
|
| 1146 |
+
"loss": 0.0303,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.864,
|
| 1151 |
+
"grad_norm": 0.0027039784472435713,
|
| 1152 |
+
"learning_rate": 2.3724792408066432e-05,
|
| 1153 |
+
"loss": 0.0337,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.8817777777777778,
|
| 1158 |
+
"grad_norm": 0.015552740544080734,
|
| 1159 |
+
"learning_rate": 2.3527085804665877e-05,
|
| 1160 |
+
"loss": 0.0131,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.8995555555555557,
|
| 1165 |
+
"grad_norm": 0.014052975922822952,
|
| 1166 |
+
"learning_rate": 2.332937920126532e-05,
|
| 1167 |
+
"loss": 0.0434,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.9173333333333336,
|
| 1172 |
+
"grad_norm": 0.7165421843528748,
|
| 1173 |
+
"learning_rate": 2.313167259786477e-05,
|
| 1174 |
+
"loss": 0.0069,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.935111111111111,
|
| 1179 |
+
"grad_norm": 0.020382430404424667,
|
| 1180 |
+
"learning_rate": 2.2933965994464214e-05,
|
| 1181 |
+
"loss": 0.0419,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.952888888888889,
|
| 1186 |
+
"grad_norm": 0.07127852737903595,
|
| 1187 |
+
"learning_rate": 2.2736259391063662e-05,
|
| 1188 |
+
"loss": 0.0303,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.970666666666667,
|
| 1193 |
+
"grad_norm": 2.5529978275299072,
|
| 1194 |
+
"learning_rate": 2.2538552787663107e-05,
|
| 1195 |
+
"loss": 0.0063,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.9884444444444442,
|
| 1200 |
+
"grad_norm": 0.0018354392377659678,
|
| 1201 |
+
"learning_rate": 2.2340846184262555e-05,
|
| 1202 |
+
"loss": 0.0323,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.999111111111111,
|
| 1207 |
+
"eval_accuracy": 1.0,
|
| 1208 |
+
"eval_loss": 6.3951174524845555e-06,
|
| 1209 |
+
"eval_runtime": 39.8564,
|
| 1210 |
+
"eval_samples_per_second": 200.721,
|
| 1211 |
+
"eval_steps_per_second": 6.273,
|
| 1212 |
+
"step": 1686
|
| 1213 |
+
}
|
| 1214 |
+
],
|
| 1215 |
+
"logging_steps": 10,
|
| 1216 |
+
"max_steps": 2810,
|
| 1217 |
+
"num_input_tokens_seen": 0,
|
| 1218 |
+
"num_train_epochs": 5,
|
| 1219 |
+
"save_steps": 500,
|
| 1220 |
+
"stateful_callbacks": {
|
| 1221 |
+
"TrainerControl": {
|
| 1222 |
+
"args": {
|
| 1223 |
+
"should_epoch_stop": false,
|
| 1224 |
+
"should_evaluate": false,
|
| 1225 |
+
"should_log": false,
|
| 1226 |
+
"should_save": true,
|
| 1227 |
+
"should_training_stop": false
|
| 1228 |
+
},
|
| 1229 |
+
"attributes": {}
|
| 1230 |
+
}
|
| 1231 |
+
},
|
| 1232 |
+
"total_flos": 5.364628275991413e+18,
|
| 1233 |
+
"train_batch_size": 32,
|
| 1234 |
+
"trial_name": null,
|
| 1235 |
+
"trial_params": null
|
| 1236 |
+
}
|
last-checkpoint/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a8175a9ad6281337906169a5a7d648239e46227467b846a8a98a2327fc311cd
|
| 3 |
+
size 5432
|