Upload 8 files
Browse files- config.json +134 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +23 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1808 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ResNetForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"depths": [
|
| 6 |
+
3,
|
| 7 |
+
4,
|
| 8 |
+
6,
|
| 9 |
+
3
|
| 10 |
+
],
|
| 11 |
+
"downsample_in_bottleneck": false,
|
| 12 |
+
"downsample_in_first_stage": false,
|
| 13 |
+
"dtype": "float32",
|
| 14 |
+
"embedding_size": 64,
|
| 15 |
+
"hidden_act": "relu",
|
| 16 |
+
"hidden_sizes": [
|
| 17 |
+
256,
|
| 18 |
+
512,
|
| 19 |
+
1024,
|
| 20 |
+
2048
|
| 21 |
+
],
|
| 22 |
+
"id2label": {
|
| 23 |
+
"0": "Accessories",
|
| 24 |
+
"1": "Apparel Set",
|
| 25 |
+
"2": "Bags",
|
| 26 |
+
"3": "Bath and Body",
|
| 27 |
+
"4": "Beauty Accessories",
|
| 28 |
+
"5": "Belts",
|
| 29 |
+
"6": "Bottomwear",
|
| 30 |
+
"7": "Cufflinks",
|
| 31 |
+
"8": "Dress",
|
| 32 |
+
"9": "Eyes",
|
| 33 |
+
"10": "Eyewear",
|
| 34 |
+
"11": "Flip Flops",
|
| 35 |
+
"12": "Fragrance",
|
| 36 |
+
"13": "Free Gifts",
|
| 37 |
+
"14": "Gloves",
|
| 38 |
+
"15": "Hair",
|
| 39 |
+
"16": "Headwear",
|
| 40 |
+
"17": "Home Furnishing",
|
| 41 |
+
"18": "Innerwear",
|
| 42 |
+
"19": "Jewellery",
|
| 43 |
+
"20": "Lips",
|
| 44 |
+
"21": "Loungewear and Nightwear",
|
| 45 |
+
"22": "Makeup",
|
| 46 |
+
"23": "Mufflers",
|
| 47 |
+
"24": "Nails",
|
| 48 |
+
"25": "Perfumes",
|
| 49 |
+
"26": "Sandal",
|
| 50 |
+
"27": "Saree",
|
| 51 |
+
"28": "Scarves",
|
| 52 |
+
"29": "Shoe Accessories",
|
| 53 |
+
"30": "Shoes",
|
| 54 |
+
"31": "Skin",
|
| 55 |
+
"32": "Skin Care",
|
| 56 |
+
"33": "Socks",
|
| 57 |
+
"34": "Sports Accessories",
|
| 58 |
+
"35": "Sports Equipment",
|
| 59 |
+
"36": "Stoles",
|
| 60 |
+
"37": "Ties",
|
| 61 |
+
"38": "Topwear",
|
| 62 |
+
"39": "Umbrellas",
|
| 63 |
+
"40": "Vouchers",
|
| 64 |
+
"41": "Wallets",
|
| 65 |
+
"42": "Watches",
|
| 66 |
+
"43": "Water Bottle",
|
| 67 |
+
"44": "Wristbands"
|
| 68 |
+
},
|
| 69 |
+
"label2id": {
|
| 70 |
+
"Accessories": 0,
|
| 71 |
+
"Apparel Set": 1,
|
| 72 |
+
"Bags": 2,
|
| 73 |
+
"Bath and Body": 3,
|
| 74 |
+
"Beauty Accessories": 4,
|
| 75 |
+
"Belts": 5,
|
| 76 |
+
"Bottomwear": 6,
|
| 77 |
+
"Cufflinks": 7,
|
| 78 |
+
"Dress": 8,
|
| 79 |
+
"Eyes": 9,
|
| 80 |
+
"Eyewear": 10,
|
| 81 |
+
"Flip Flops": 11,
|
| 82 |
+
"Fragrance": 12,
|
| 83 |
+
"Free Gifts": 13,
|
| 84 |
+
"Gloves": 14,
|
| 85 |
+
"Hair": 15,
|
| 86 |
+
"Headwear": 16,
|
| 87 |
+
"Home Furnishing": 17,
|
| 88 |
+
"Innerwear": 18,
|
| 89 |
+
"Jewellery": 19,
|
| 90 |
+
"Lips": 20,
|
| 91 |
+
"Loungewear and Nightwear": 21,
|
| 92 |
+
"Makeup": 22,
|
| 93 |
+
"Mufflers": 23,
|
| 94 |
+
"Nails": 24,
|
| 95 |
+
"Perfumes": 25,
|
| 96 |
+
"Sandal": 26,
|
| 97 |
+
"Saree": 27,
|
| 98 |
+
"Scarves": 28,
|
| 99 |
+
"Shoe Accessories": 29,
|
| 100 |
+
"Shoes": 30,
|
| 101 |
+
"Skin": 31,
|
| 102 |
+
"Skin Care": 32,
|
| 103 |
+
"Socks": 33,
|
| 104 |
+
"Sports Accessories": 34,
|
| 105 |
+
"Sports Equipment": 35,
|
| 106 |
+
"Stoles": 36,
|
| 107 |
+
"Ties": 37,
|
| 108 |
+
"Topwear": 38,
|
| 109 |
+
"Umbrellas": 39,
|
| 110 |
+
"Vouchers": 40,
|
| 111 |
+
"Wallets": 41,
|
| 112 |
+
"Watches": 42,
|
| 113 |
+
"Water Bottle": 43,
|
| 114 |
+
"Wristbands": 44
|
| 115 |
+
},
|
| 116 |
+
"layer_type": "bottleneck",
|
| 117 |
+
"model_type": "resnet",
|
| 118 |
+
"num_channels": 3,
|
| 119 |
+
"out_features": [
|
| 120 |
+
"stage4"
|
| 121 |
+
],
|
| 122 |
+
"out_indices": [
|
| 123 |
+
4
|
| 124 |
+
],
|
| 125 |
+
"problem_type": "single_label_classification",
|
| 126 |
+
"stage_names": [
|
| 127 |
+
"stem",
|
| 128 |
+
"stage1",
|
| 129 |
+
"stage2",
|
| 130 |
+
"stage3",
|
| 131 |
+
"stage4"
|
| 132 |
+
],
|
| 133 |
+
"transformers_version": "5.0.0"
|
| 134 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f550a6cd4f799853a3762cc283835335442b9e40377e644c71cd214d9583022
|
| 3 |
+
size 94655444
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d56355100c76f1dc4fe7adce57757b1f8aefd718514973b50ad8113d8ee9b37d
|
| 3 |
+
size 188939979
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_pct": 0.875,
|
| 3 |
+
"data_format": "channels_first",
|
| 4 |
+
"do_normalize": true,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"image_mean": [
|
| 8 |
+
0.485,
|
| 9 |
+
0.456,
|
| 10 |
+
0.406
|
| 11 |
+
],
|
| 12 |
+
"image_processor_type": "ConvNextImageProcessorFast",
|
| 13 |
+
"image_std": [
|
| 14 |
+
0.229,
|
| 15 |
+
0.224,
|
| 16 |
+
0.225
|
| 17 |
+
],
|
| 18 |
+
"resample": 3,
|
| 19 |
+
"rescale_factor": 0.00392156862745098,
|
| 20 |
+
"size": {
|
| 21 |
+
"shortest_edge": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17cd930da9783ca70bad4b9cdeee6a06c0acea8f34645a333c93341f487f66a3
|
| 3 |
+
size 14645
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5459ac3f6c29a460ad51df3985e571a52e8101feba291f3386dd1ff324ba92f2
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1808 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 12395,
|
| 3 |
+
"best_metric": 0.9532667876588021,
|
| 4 |
+
"best_model_checkpoint": "./resnet50-fashion/checkpoint-12395",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 12395,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02016942315449778,
|
| 14 |
+
"grad_norm": 10.520899772644043,
|
| 15 |
+
"learning_rate": 4.9802339653085925e-05,
|
| 16 |
+
"loss": 3.5885751342773435,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.04033884630899556,
|
| 21 |
+
"grad_norm": 11.112278938293457,
|
| 22 |
+
"learning_rate": 4.9600645421540946e-05,
|
| 23 |
+
"loss": 3.066470947265625,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.060508269463493344,
|
| 28 |
+
"grad_norm": 6.844457149505615,
|
| 29 |
+
"learning_rate": 4.939895118999597e-05,
|
| 30 |
+
"loss": 2.5258544921875,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.08067769261799113,
|
| 35 |
+
"grad_norm": 13.254263877868652,
|
| 36 |
+
"learning_rate": 4.919725695845099e-05,
|
| 37 |
+
"loss": 2.1543016052246093,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.10084711577248891,
|
| 42 |
+
"grad_norm": 6.580667972564697,
|
| 43 |
+
"learning_rate": 4.8995562726906015e-05,
|
| 44 |
+
"loss": 1.9740011596679687,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.12101653892698669,
|
| 49 |
+
"grad_norm": 7.654535293579102,
|
| 50 |
+
"learning_rate": 4.8793868495361035e-05,
|
| 51 |
+
"loss": 1.7021075439453126,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.14118596208148446,
|
| 56 |
+
"grad_norm": 6.575769424438477,
|
| 57 |
+
"learning_rate": 4.8592174263816056e-05,
|
| 58 |
+
"loss": 1.6266166687011718,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.16135538523598225,
|
| 63 |
+
"grad_norm": 4.312820911407471,
|
| 64 |
+
"learning_rate": 4.839048003227108e-05,
|
| 65 |
+
"loss": 1.4934498596191406,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.18152480839048002,
|
| 70 |
+
"grad_norm": 4.195486068725586,
|
| 71 |
+
"learning_rate": 4.81887858007261e-05,
|
| 72 |
+
"loss": 1.33470703125,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.20169423154497781,
|
| 77 |
+
"grad_norm": 7.272072792053223,
|
| 78 |
+
"learning_rate": 4.7987091569181125e-05,
|
| 79 |
+
"loss": 1.2691184234619142,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.22186365469947558,
|
| 84 |
+
"grad_norm": 5.7398200035095215,
|
| 85 |
+
"learning_rate": 4.7785397337636145e-05,
|
| 86 |
+
"loss": 1.136511993408203,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.24203307785397338,
|
| 91 |
+
"grad_norm": 6.210187911987305,
|
| 92 |
+
"learning_rate": 4.7583703106091166e-05,
|
| 93 |
+
"loss": 1.0565355682373048,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.26220250100847114,
|
| 98 |
+
"grad_norm": 7.919003963470459,
|
| 99 |
+
"learning_rate": 4.7382008874546193e-05,
|
| 100 |
+
"loss": 0.9344426727294922,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2823719241629689,
|
| 105 |
+
"grad_norm": 3.5390396118164062,
|
| 106 |
+
"learning_rate": 4.7180314643001214e-05,
|
| 107 |
+
"loss": 0.8554486846923828,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.30254134731746674,
|
| 112 |
+
"grad_norm": 7.0212273597717285,
|
| 113 |
+
"learning_rate": 4.6978620411456235e-05,
|
| 114 |
+
"loss": 0.7601541137695312,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3227107704719645,
|
| 119 |
+
"grad_norm": 4.230292320251465,
|
| 120 |
+
"learning_rate": 4.6776926179911255e-05,
|
| 121 |
+
"loss": 0.6942082977294922,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.34288019362646227,
|
| 126 |
+
"grad_norm": 7.144539833068848,
|
| 127 |
+
"learning_rate": 4.6575231948366276e-05,
|
| 128 |
+
"loss": 0.6801434326171875,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.36304961678096004,
|
| 133 |
+
"grad_norm": 2.7382545471191406,
|
| 134 |
+
"learning_rate": 4.6373537716821303e-05,
|
| 135 |
+
"loss": 0.7084119415283203,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.38321903993545786,
|
| 140 |
+
"grad_norm": 6.787492752075195,
|
| 141 |
+
"learning_rate": 4.6171843485276324e-05,
|
| 142 |
+
"loss": 0.6313615417480469,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.40338846308995563,
|
| 147 |
+
"grad_norm": 6.617012023925781,
|
| 148 |
+
"learning_rate": 4.5970149253731345e-05,
|
| 149 |
+
"loss": 0.6253053283691407,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.4235578862444534,
|
| 154 |
+
"grad_norm": 2.82509708404541,
|
| 155 |
+
"learning_rate": 4.5768455022186365e-05,
|
| 156 |
+
"loss": 0.5697365188598633,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.44372730939895116,
|
| 161 |
+
"grad_norm": 6.462093353271484,
|
| 162 |
+
"learning_rate": 4.5566760790641386e-05,
|
| 163 |
+
"loss": 0.5618856811523437,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.463896732553449,
|
| 168 |
+
"grad_norm": 8.217955589294434,
|
| 169 |
+
"learning_rate": 4.5365066559096413e-05,
|
| 170 |
+
"loss": 0.5241514587402344,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.48406615570794675,
|
| 175 |
+
"grad_norm": 5.901000499725342,
|
| 176 |
+
"learning_rate": 4.5163372327551434e-05,
|
| 177 |
+
"loss": 0.5597282028198243,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5042355788624445,
|
| 182 |
+
"grad_norm": 5.706325054168701,
|
| 183 |
+
"learning_rate": 4.4961678096006455e-05,
|
| 184 |
+
"loss": 0.500801773071289,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.5244050020169423,
|
| 189 |
+
"grad_norm": 3.6276135444641113,
|
| 190 |
+
"learning_rate": 4.475998386446148e-05,
|
| 191 |
+
"loss": 0.4998150253295898,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5445744251714401,
|
| 196 |
+
"grad_norm": 1.5081746578216553,
|
| 197 |
+
"learning_rate": 4.45582896329165e-05,
|
| 198 |
+
"loss": 0.4404385757446289,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5647438483259378,
|
| 203 |
+
"grad_norm": 2.9040212631225586,
|
| 204 |
+
"learning_rate": 4.4356595401371523e-05,
|
| 205 |
+
"loss": 0.42787273406982423,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5849132714804357,
|
| 210 |
+
"grad_norm": 6.973745822906494,
|
| 211 |
+
"learning_rate": 4.4154901169826544e-05,
|
| 212 |
+
"loss": 0.4308282470703125,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.6050826946349335,
|
| 217 |
+
"grad_norm": 5.634387969970703,
|
| 218 |
+
"learning_rate": 4.3953206938281565e-05,
|
| 219 |
+
"loss": 0.3713912582397461,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.6252521177894312,
|
| 224 |
+
"grad_norm": 2.7692153453826904,
|
| 225 |
+
"learning_rate": 4.375151270673659e-05,
|
| 226 |
+
"loss": 0.38077377319335937,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.645421540943929,
|
| 231 |
+
"grad_norm": 3.9123497009277344,
|
| 232 |
+
"learning_rate": 4.354981847519161e-05,
|
| 233 |
+
"loss": 0.35636608123779295,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6655909640984268,
|
| 238 |
+
"grad_norm": 1.6538969278335571,
|
| 239 |
+
"learning_rate": 4.3348124243646633e-05,
|
| 240 |
+
"loss": 0.3324755096435547,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6857603872529245,
|
| 245 |
+
"grad_norm": 2.7436976432800293,
|
| 246 |
+
"learning_rate": 4.3146430012101654e-05,
|
| 247 |
+
"loss": 0.3718099594116211,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.7059298104074223,
|
| 252 |
+
"grad_norm": 2.514030933380127,
|
| 253 |
+
"learning_rate": 4.2944735780556675e-05,
|
| 254 |
+
"loss": 0.365201416015625,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.7260992335619201,
|
| 259 |
+
"grad_norm": 3.3714559078216553,
|
| 260 |
+
"learning_rate": 4.2743041549011695e-05,
|
| 261 |
+
"loss": 0.3048884391784668,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.746268656716418,
|
| 266 |
+
"grad_norm": 3.61609148979187,
|
| 267 |
+
"learning_rate": 4.254134731746672e-05,
|
| 268 |
+
"loss": 0.4202354049682617,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7664380798709157,
|
| 273 |
+
"grad_norm": 5.999142169952393,
|
| 274 |
+
"learning_rate": 4.2339653085921743e-05,
|
| 275 |
+
"loss": 0.4184662628173828,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7866075030254135,
|
| 280 |
+
"grad_norm": 7.747167587280273,
|
| 281 |
+
"learning_rate": 4.213795885437677e-05,
|
| 282 |
+
"loss": 0.2971736145019531,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.8067769261799113,
|
| 287 |
+
"grad_norm": 3.178110122680664,
|
| 288 |
+
"learning_rate": 4.193626462283179e-05,
|
| 289 |
+
"loss": 0.3129091262817383,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.826946349334409,
|
| 294 |
+
"grad_norm": 7.176916599273682,
|
| 295 |
+
"learning_rate": 4.173457039128681e-05,
|
| 296 |
+
"loss": 0.3405015182495117,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.8471157724889068,
|
| 301 |
+
"grad_norm": 2.1656532287597656,
|
| 302 |
+
"learning_rate": 4.153287615974183e-05,
|
| 303 |
+
"loss": 0.32712291717529296,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8672851956434046,
|
| 308 |
+
"grad_norm": 7.728622913360596,
|
| 309 |
+
"learning_rate": 4.1331181928196853e-05,
|
| 310 |
+
"loss": 0.3142762756347656,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8874546187979023,
|
| 315 |
+
"grad_norm": 3.9452037811279297,
|
| 316 |
+
"learning_rate": 4.1129487696651874e-05,
|
| 317 |
+
"loss": 0.3304166793823242,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.9076240419524002,
|
| 322 |
+
"grad_norm": 4.345726013183594,
|
| 323 |
+
"learning_rate": 4.09277934651069e-05,
|
| 324 |
+
"loss": 0.29476640701293944,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.927793465106898,
|
| 329 |
+
"grad_norm": 9.373312950134277,
|
| 330 |
+
"learning_rate": 4.072609923356192e-05,
|
| 331 |
+
"loss": 0.29666582107543943,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.9479628882613957,
|
| 336 |
+
"grad_norm": 15.719057083129883,
|
| 337 |
+
"learning_rate": 4.052440500201695e-05,
|
| 338 |
+
"loss": 0.31481658935546875,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9681323114158935,
|
| 343 |
+
"grad_norm": 11.424883842468262,
|
| 344 |
+
"learning_rate": 4.0322710770471963e-05,
|
| 345 |
+
"loss": 0.40077999114990237,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9883017345703913,
|
| 350 |
+
"grad_norm": 3.748368740081787,
|
| 351 |
+
"learning_rate": 4.0121016538926984e-05,
|
| 352 |
+
"loss": 0.3299874496459961,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.0,
|
| 357 |
+
"eval_accuracy": 0.9299001814882033,
|
| 358 |
+
"eval_loss": 0.27112650871276855,
|
| 359 |
+
"eval_runtime": 20.3548,
|
| 360 |
+
"eval_samples_per_second": 216.558,
|
| 361 |
+
"eval_steps_per_second": 13.559,
|
| 362 |
+
"step": 2479
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.008471157724889,
|
| 366 |
+
"grad_norm": 6.341300964355469,
|
| 367 |
+
"learning_rate": 3.991932230738201e-05,
|
| 368 |
+
"loss": 0.3089567756652832,
|
| 369 |
+
"step": 2500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.0286405808793868,
|
| 373 |
+
"grad_norm": 2.0980629920959473,
|
| 374 |
+
"learning_rate": 3.971762807583703e-05,
|
| 375 |
+
"loss": 0.3381888961791992,
|
| 376 |
+
"step": 2550
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.0488100040338846,
|
| 380 |
+
"grad_norm": 0.7737718820571899,
|
| 381 |
+
"learning_rate": 3.951593384429206e-05,
|
| 382 |
+
"loss": 0.2835183525085449,
|
| 383 |
+
"step": 2600
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.0689794271883823,
|
| 387 |
+
"grad_norm": 6.585007190704346,
|
| 388 |
+
"learning_rate": 3.931423961274708e-05,
|
| 389 |
+
"loss": 0.2929516220092773,
|
| 390 |
+
"step": 2650
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.0891488503428801,
|
| 394 |
+
"grad_norm": 2.3810129165649414,
|
| 395 |
+
"learning_rate": 3.91125453812021e-05,
|
| 396 |
+
"loss": 0.2525530433654785,
|
| 397 |
+
"step": 2700
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.1093182734973779,
|
| 401 |
+
"grad_norm": 3.5129992961883545,
|
| 402 |
+
"learning_rate": 3.891085114965712e-05,
|
| 403 |
+
"loss": 0.27019474029541013,
|
| 404 |
+
"step": 2750
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.1294876966518759,
|
| 408 |
+
"grad_norm": 5.742709159851074,
|
| 409 |
+
"learning_rate": 3.870915691811214e-05,
|
| 410 |
+
"loss": 0.2933888626098633,
|
| 411 |
+
"step": 2800
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.1496571198063736,
|
| 415 |
+
"grad_norm": 3.1736836433410645,
|
| 416 |
+
"learning_rate": 3.850746268656716e-05,
|
| 417 |
+
"loss": 0.2935880470275879,
|
| 418 |
+
"step": 2850
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.1698265429608714,
|
| 422 |
+
"grad_norm": 6.119251251220703,
|
| 423 |
+
"learning_rate": 3.830576845502219e-05,
|
| 424 |
+
"loss": 0.3064602470397949,
|
| 425 |
+
"step": 2900
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.1899959661153692,
|
| 429 |
+
"grad_norm": 3.5085670948028564,
|
| 430 |
+
"learning_rate": 3.810407422347721e-05,
|
| 431 |
+
"loss": 0.25757852554321287,
|
| 432 |
+
"step": 2950
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.210165389269867,
|
| 436 |
+
"grad_norm": 3.2313594818115234,
|
| 437 |
+
"learning_rate": 3.790237999193224e-05,
|
| 438 |
+
"loss": 0.2911284065246582,
|
| 439 |
+
"step": 3000
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2303348124243647,
|
| 443 |
+
"grad_norm": 4.116666316986084,
|
| 444 |
+
"learning_rate": 3.770068576038725e-05,
|
| 445 |
+
"loss": 0.3029876136779785,
|
| 446 |
+
"step": 3050
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.2505042355788625,
|
| 450 |
+
"grad_norm": 0.5835158824920654,
|
| 451 |
+
"learning_rate": 3.749899152884227e-05,
|
| 452 |
+
"loss": 0.22566375732421876,
|
| 453 |
+
"step": 3100
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.2706736587333602,
|
| 457 |
+
"grad_norm": 4.364255428314209,
|
| 458 |
+
"learning_rate": 3.72972972972973e-05,
|
| 459 |
+
"loss": 0.20720888137817384,
|
| 460 |
+
"step": 3150
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.290843081887858,
|
| 464 |
+
"grad_norm": 3.6221392154693604,
|
| 465 |
+
"learning_rate": 3.709560306575232e-05,
|
| 466 |
+
"loss": 0.22576507568359375,
|
| 467 |
+
"step": 3200
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.3110125050423558,
|
| 471 |
+
"grad_norm": 10.07580852508545,
|
| 472 |
+
"learning_rate": 3.689390883420734e-05,
|
| 473 |
+
"loss": 0.2189176368713379,
|
| 474 |
+
"step": 3250
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.3311819281968535,
|
| 478 |
+
"grad_norm": 6.97329568862915,
|
| 479 |
+
"learning_rate": 3.669221460266237e-05,
|
| 480 |
+
"loss": 0.338087158203125,
|
| 481 |
+
"step": 3300
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.3513513513513513,
|
| 485 |
+
"grad_norm": 1.1882565021514893,
|
| 486 |
+
"learning_rate": 3.649052037111739e-05,
|
| 487 |
+
"loss": 0.21201395034790038,
|
| 488 |
+
"step": 3350
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.371520774505849,
|
| 492 |
+
"grad_norm": 2.1187331676483154,
|
| 493 |
+
"learning_rate": 3.628882613957241e-05,
|
| 494 |
+
"loss": 0.22095136642456054,
|
| 495 |
+
"step": 3400
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.3916901976603469,
|
| 499 |
+
"grad_norm": 4.908150672912598,
|
| 500 |
+
"learning_rate": 3.608713190802743e-05,
|
| 501 |
+
"loss": 0.24210647583007813,
|
| 502 |
+
"step": 3450
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.4118596208148446,
|
| 506 |
+
"grad_norm": 2.4119579792022705,
|
| 507 |
+
"learning_rate": 3.588543767648245e-05,
|
| 508 |
+
"loss": 0.2516419792175293,
|
| 509 |
+
"step": 3500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.4320290439693424,
|
| 513 |
+
"grad_norm": 5.924858570098877,
|
| 514 |
+
"learning_rate": 3.568374344493748e-05,
|
| 515 |
+
"loss": 0.2286382484436035,
|
| 516 |
+
"step": 3550
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.4521984671238402,
|
| 520 |
+
"grad_norm": 6.0561041831970215,
|
| 521 |
+
"learning_rate": 3.54820492133925e-05,
|
| 522 |
+
"loss": 0.272211971282959,
|
| 523 |
+
"step": 3600
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 1.472367890278338,
|
| 527 |
+
"grad_norm": 11.075820922851562,
|
| 528 |
+
"learning_rate": 3.528035498184752e-05,
|
| 529 |
+
"loss": 0.18730430603027343,
|
| 530 |
+
"step": 3650
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.4925373134328357,
|
| 534 |
+
"grad_norm": 6.133443355560303,
|
| 535 |
+
"learning_rate": 3.507866075030255e-05,
|
| 536 |
+
"loss": 0.2744808006286621,
|
| 537 |
+
"step": 3700
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.5127067365873335,
|
| 541 |
+
"grad_norm": 2.7195820808410645,
|
| 542 |
+
"learning_rate": 3.487696651875756e-05,
|
| 543 |
+
"loss": 0.2309401512145996,
|
| 544 |
+
"step": 3750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.5328761597418312,
|
| 548 |
+
"grad_norm": 5.68732213973999,
|
| 549 |
+
"learning_rate": 3.467527228721259e-05,
|
| 550 |
+
"loss": 0.22877124786376954,
|
| 551 |
+
"step": 3800
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.5530455828963292,
|
| 555 |
+
"grad_norm": 5.202581882476807,
|
| 556 |
+
"learning_rate": 3.447357805566761e-05,
|
| 557 |
+
"loss": 0.27316980361938475,
|
| 558 |
+
"step": 3850
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 1.573215006050827,
|
| 562 |
+
"grad_norm": 0.48496291041374207,
|
| 563 |
+
"learning_rate": 3.427188382412263e-05,
|
| 564 |
+
"loss": 0.20376949310302733,
|
| 565 |
+
"step": 3900
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 1.5933844292053247,
|
| 569 |
+
"grad_norm": 3.1519060134887695,
|
| 570 |
+
"learning_rate": 3.407018959257766e-05,
|
| 571 |
+
"loss": 0.20258005142211913,
|
| 572 |
+
"step": 3950
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.6135538523598225,
|
| 576 |
+
"grad_norm": 0.649708092212677,
|
| 577 |
+
"learning_rate": 3.386849536103268e-05,
|
| 578 |
+
"loss": 0.2322039031982422,
|
| 579 |
+
"step": 4000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.6337232755143203,
|
| 583 |
+
"grad_norm": 2.9072558879852295,
|
| 584 |
+
"learning_rate": 3.36668011294877e-05,
|
| 585 |
+
"loss": 0.20238967895507812,
|
| 586 |
+
"step": 4050
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.653892698668818,
|
| 590 |
+
"grad_norm": 1.227184772491455,
|
| 591 |
+
"learning_rate": 3.346510689794272e-05,
|
| 592 |
+
"loss": 0.19608688354492188,
|
| 593 |
+
"step": 4100
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.6740621218233158,
|
| 597 |
+
"grad_norm": 5.7164483070373535,
|
| 598 |
+
"learning_rate": 3.326341266639774e-05,
|
| 599 |
+
"loss": 0.2062531852722168,
|
| 600 |
+
"step": 4150
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.6942315449778136,
|
| 604 |
+
"grad_norm": 4.255147457122803,
|
| 605 |
+
"learning_rate": 3.306171843485277e-05,
|
| 606 |
+
"loss": 0.24369115829467775,
|
| 607 |
+
"step": 4200
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 1.7144009681323116,
|
| 611 |
+
"grad_norm": 3.4695684909820557,
|
| 612 |
+
"learning_rate": 3.286002420330779e-05,
|
| 613 |
+
"loss": 0.19526601791381837,
|
| 614 |
+
"step": 4250
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.7345703912868093,
|
| 618 |
+
"grad_norm": 3.0551869869232178,
|
| 619 |
+
"learning_rate": 3.265832997176281e-05,
|
| 620 |
+
"loss": 0.2388477325439453,
|
| 621 |
+
"step": 4300
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.754739814441307,
|
| 625 |
+
"grad_norm": 5.207853317260742,
|
| 626 |
+
"learning_rate": 3.2456635740217836e-05,
|
| 627 |
+
"loss": 0.22748014450073242,
|
| 628 |
+
"step": 4350
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.7749092375958049,
|
| 632 |
+
"grad_norm": 6.68976354598999,
|
| 633 |
+
"learning_rate": 3.225494150867285e-05,
|
| 634 |
+
"loss": 0.2202134132385254,
|
| 635 |
+
"step": 4400
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.7950786607503026,
|
| 639 |
+
"grad_norm": 1.4704676866531372,
|
| 640 |
+
"learning_rate": 3.205324727712787e-05,
|
| 641 |
+
"loss": 0.19456554412841798,
|
| 642 |
+
"step": 4450
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.8152480839048004,
|
| 646 |
+
"grad_norm": 7.315227508544922,
|
| 647 |
+
"learning_rate": 3.18515530455829e-05,
|
| 648 |
+
"loss": 0.3203375244140625,
|
| 649 |
+
"step": 4500
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.8354175070592982,
|
| 653 |
+
"grad_norm": 1.7864209413528442,
|
| 654 |
+
"learning_rate": 3.164985881403792e-05,
|
| 655 |
+
"loss": 0.18399269104003907,
|
| 656 |
+
"step": 4550
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.855586930213796,
|
| 660 |
+
"grad_norm": 5.953585624694824,
|
| 661 |
+
"learning_rate": 3.1448164582492946e-05,
|
| 662 |
+
"loss": 0.25314868927001954,
|
| 663 |
+
"step": 4600
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.8757563533682937,
|
| 667 |
+
"grad_norm": 4.386700630187988,
|
| 668 |
+
"learning_rate": 3.124647035094797e-05,
|
| 669 |
+
"loss": 0.19606996536254884,
|
| 670 |
+
"step": 4650
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.8959257765227915,
|
| 674 |
+
"grad_norm": 6.6788330078125,
|
| 675 |
+
"learning_rate": 3.104477611940299e-05,
|
| 676 |
+
"loss": 0.19745792388916017,
|
| 677 |
+
"step": 4700
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.9160951996772893,
|
| 681 |
+
"grad_norm": 5.657231330871582,
|
| 682 |
+
"learning_rate": 3.084308188785801e-05,
|
| 683 |
+
"loss": 0.17997068405151367,
|
| 684 |
+
"step": 4750
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.936264622831787,
|
| 688 |
+
"grad_norm": 1.3474644422531128,
|
| 689 |
+
"learning_rate": 3.064138765631303e-05,
|
| 690 |
+
"loss": 0.19095916748046876,
|
| 691 |
+
"step": 4800
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.9564340459862848,
|
| 695 |
+
"grad_norm": 4.156283378601074,
|
| 696 |
+
"learning_rate": 3.043969342476805e-05,
|
| 697 |
+
"loss": 0.21400732040405274,
|
| 698 |
+
"step": 4850
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.9766034691407826,
|
| 702 |
+
"grad_norm": 0.9628849029541016,
|
| 703 |
+
"learning_rate": 3.0237999193223077e-05,
|
| 704 |
+
"loss": 0.23152576446533202,
|
| 705 |
+
"step": 4900
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.9967728922952803,
|
| 709 |
+
"grad_norm": 8.431641578674316,
|
| 710 |
+
"learning_rate": 3.0036304961678098e-05,
|
| 711 |
+
"loss": 0.19971717834472658,
|
| 712 |
+
"step": 4950
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.0,
|
| 716 |
+
"eval_accuracy": 0.9441923774954628,
|
| 717 |
+
"eval_loss": 0.2141035795211792,
|
| 718 |
+
"eval_runtime": 20.5277,
|
| 719 |
+
"eval_samples_per_second": 214.734,
|
| 720 |
+
"eval_steps_per_second": 13.445,
|
| 721 |
+
"step": 4958
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 2.016942315449778,
|
| 725 |
+
"grad_norm": 7.967335224151611,
|
| 726 |
+
"learning_rate": 2.9834610730133122e-05,
|
| 727 |
+
"loss": 0.20258569717407227,
|
| 728 |
+
"step": 5000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 2.037111738604276,
|
| 732 |
+
"grad_norm": 1.8908740282058716,
|
| 733 |
+
"learning_rate": 2.9632916498588142e-05,
|
| 734 |
+
"loss": 0.2126533317565918,
|
| 735 |
+
"step": 5050
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 2.0572811617587736,
|
| 739 |
+
"grad_norm": 0.6554378867149353,
|
| 740 |
+
"learning_rate": 2.9431222267043163e-05,
|
| 741 |
+
"loss": 0.21582073211669922,
|
| 742 |
+
"step": 5100
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 2.0774505849132714,
|
| 746 |
+
"grad_norm": 2.4483304023742676,
|
| 747 |
+
"learning_rate": 2.9229528035498187e-05,
|
| 748 |
+
"loss": 0.21905384063720704,
|
| 749 |
+
"step": 5150
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 2.097620008067769,
|
| 753 |
+
"grad_norm": 3.9244415760040283,
|
| 754 |
+
"learning_rate": 2.9027833803953208e-05,
|
| 755 |
+
"loss": 0.16603147506713867,
|
| 756 |
+
"step": 5200
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 2.117789431222267,
|
| 760 |
+
"grad_norm": 0.17323902249336243,
|
| 761 |
+
"learning_rate": 2.882613957240823e-05,
|
| 762 |
+
"loss": 0.1652754020690918,
|
| 763 |
+
"step": 5250
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 2.1379588543767647,
|
| 767 |
+
"grad_norm": 4.670708179473877,
|
| 768 |
+
"learning_rate": 2.8624445340863252e-05,
|
| 769 |
+
"loss": 0.17886968612670898,
|
| 770 |
+
"step": 5300
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.1581282775312625,
|
| 774 |
+
"grad_norm": 1.9104596376419067,
|
| 775 |
+
"learning_rate": 2.8422751109318273e-05,
|
| 776 |
+
"loss": 0.17843175888061524,
|
| 777 |
+
"step": 5350
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 2.1782977006857602,
|
| 781 |
+
"grad_norm": 3.732868194580078,
|
| 782 |
+
"learning_rate": 2.82210568777733e-05,
|
| 783 |
+
"loss": 0.163537654876709,
|
| 784 |
+
"step": 5400
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 2.198467123840258,
|
| 788 |
+
"grad_norm": 3.159740924835205,
|
| 789 |
+
"learning_rate": 2.801936264622832e-05,
|
| 790 |
+
"loss": 0.18888442993164062,
|
| 791 |
+
"step": 5450
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.2186365469947558,
|
| 795 |
+
"grad_norm": 3.3042259216308594,
|
| 796 |
+
"learning_rate": 2.781766841468334e-05,
|
| 797 |
+
"loss": 0.1473686122894287,
|
| 798 |
+
"step": 5500
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 2.2388059701492535,
|
| 802 |
+
"grad_norm": 4.918742656707764,
|
| 803 |
+
"learning_rate": 2.7615974183138366e-05,
|
| 804 |
+
"loss": 0.26290460586547854,
|
| 805 |
+
"step": 5550
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 2.2589753933037517,
|
| 809 |
+
"grad_norm": 0.3327418267726898,
|
| 810 |
+
"learning_rate": 2.7414279951593386e-05,
|
| 811 |
+
"loss": 0.17103816986083983,
|
| 812 |
+
"step": 5600
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.279144816458249,
|
| 816 |
+
"grad_norm": 0.12691690027713776,
|
| 817 |
+
"learning_rate": 2.7212585720048407e-05,
|
| 818 |
+
"loss": 0.15552584648132325,
|
| 819 |
+
"step": 5650
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 2.2993142396127473,
|
| 823 |
+
"grad_norm": 2.5150461196899414,
|
| 824 |
+
"learning_rate": 2.701089148850343e-05,
|
| 825 |
+
"loss": 0.1537048625946045,
|
| 826 |
+
"step": 5700
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.319483662767245,
|
| 830 |
+
"grad_norm": 0.4236988127231598,
|
| 831 |
+
"learning_rate": 2.6809197256958452e-05,
|
| 832 |
+
"loss": 0.15081631660461425,
|
| 833 |
+
"step": 5750
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 2.339653085921743,
|
| 837 |
+
"grad_norm": 4.3972907066345215,
|
| 838 |
+
"learning_rate": 2.6607503025413476e-05,
|
| 839 |
+
"loss": 0.15966953277587892,
|
| 840 |
+
"step": 5800
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.3598225090762406,
|
| 844 |
+
"grad_norm": 0.20289625227451324,
|
| 845 |
+
"learning_rate": 2.6405808793868496e-05,
|
| 846 |
+
"loss": 0.20613027572631837,
|
| 847 |
+
"step": 5850
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 2.3799919322307383,
|
| 851 |
+
"grad_norm": 13.914892196655273,
|
| 852 |
+
"learning_rate": 2.6204114562323517e-05,
|
| 853 |
+
"loss": 0.1521580123901367,
|
| 854 |
+
"step": 5900
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.400161355385236,
|
| 858 |
+
"grad_norm": 1.8845982551574707,
|
| 859 |
+
"learning_rate": 2.6002420330778545e-05,
|
| 860 |
+
"loss": 0.18961977005004882,
|
| 861 |
+
"step": 5950
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 2.420330778539734,
|
| 865 |
+
"grad_norm": 6.2201433181762695,
|
| 866 |
+
"learning_rate": 2.5800726099233562e-05,
|
| 867 |
+
"loss": 0.17040407180786132,
|
| 868 |
+
"step": 6000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.4405002016942317,
|
| 872 |
+
"grad_norm": 1.7300294637680054,
|
| 873 |
+
"learning_rate": 2.5599031867688582e-05,
|
| 874 |
+
"loss": 0.15661882400512694,
|
| 875 |
+
"step": 6050
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 2.4606696248487294,
|
| 879 |
+
"grad_norm": 0.38582950830459595,
|
| 880 |
+
"learning_rate": 2.539733763614361e-05,
|
| 881 |
+
"loss": 0.18057403564453126,
|
| 882 |
+
"step": 6100
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.480839048003227,
|
| 886 |
+
"grad_norm": 3.5432305335998535,
|
| 887 |
+
"learning_rate": 2.519564340459863e-05,
|
| 888 |
+
"loss": 0.21012399673461915,
|
| 889 |
+
"step": 6150
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 2.501008471157725,
|
| 893 |
+
"grad_norm": 3.0747153759002686,
|
| 894 |
+
"learning_rate": 2.499394917305365e-05,
|
| 895 |
+
"loss": 0.1722999954223633,
|
| 896 |
+
"step": 6200
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.5211778943122227,
|
| 900 |
+
"grad_norm": 1.0176174640655518,
|
| 901 |
+
"learning_rate": 2.4792254941508675e-05,
|
| 902 |
+
"loss": 0.19214908599853517,
|
| 903 |
+
"step": 6250
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 2.5413473174667205,
|
| 907 |
+
"grad_norm": 1.0402295589447021,
|
| 908 |
+
"learning_rate": 2.45905607099637e-05,
|
| 909 |
+
"loss": 0.14209456443786622,
|
| 910 |
+
"step": 6300
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.5615167406212183,
|
| 914 |
+
"grad_norm": 0.44618988037109375,
|
| 915 |
+
"learning_rate": 2.4388866478418716e-05,
|
| 916 |
+
"loss": 0.14631601333618163,
|
| 917 |
+
"step": 6350
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 2.581686163775716,
|
| 921 |
+
"grad_norm": 0.32430288195610046,
|
| 922 |
+
"learning_rate": 2.418717224687374e-05,
|
| 923 |
+
"loss": 0.15652142524719237,
|
| 924 |
+
"step": 6400
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 2.601855586930214,
|
| 928 |
+
"grad_norm": 4.26512336730957,
|
| 929 |
+
"learning_rate": 2.3985478015328765e-05,
|
| 930 |
+
"loss": 0.17958589553833007,
|
| 931 |
+
"step": 6450
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 2.6220250100847116,
|
| 935 |
+
"grad_norm": 5.863807678222656,
|
| 936 |
+
"learning_rate": 2.3783783783783785e-05,
|
| 937 |
+
"loss": 0.12491923332214355,
|
| 938 |
+
"step": 6500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.6421944332392093,
|
| 942 |
+
"grad_norm": 1.6168506145477295,
|
| 943 |
+
"learning_rate": 2.3582089552238806e-05,
|
| 944 |
+
"loss": 0.16195316314697267,
|
| 945 |
+
"step": 6550
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.662363856393707,
|
| 949 |
+
"grad_norm": 0.24692083895206451,
|
| 950 |
+
"learning_rate": 2.338039532069383e-05,
|
| 951 |
+
"loss": 0.21834095001220702,
|
| 952 |
+
"step": 6600
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 2.682533279548205,
|
| 956 |
+
"grad_norm": 2.46010684967041,
|
| 957 |
+
"learning_rate": 2.317870108914885e-05,
|
| 958 |
+
"loss": 0.14829426765441894,
|
| 959 |
+
"step": 6650
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 2.7027027027027026,
|
| 963 |
+
"grad_norm": 2.4015772342681885,
|
| 964 |
+
"learning_rate": 2.2977006857603875e-05,
|
| 965 |
+
"loss": 0.19773597717285157,
|
| 966 |
+
"step": 6700
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 2.7228721258572004,
|
| 970 |
+
"grad_norm": 0.30142679810523987,
|
| 971 |
+
"learning_rate": 2.2775312626058895e-05,
|
| 972 |
+
"loss": 0.19015130996704102,
|
| 973 |
+
"step": 6750
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 2.743041549011698,
|
| 977 |
+
"grad_norm": 3.2308762073516846,
|
| 978 |
+
"learning_rate": 2.257361839451392e-05,
|
| 979 |
+
"loss": 0.15893199920654297,
|
| 980 |
+
"step": 6800
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.763210972166196,
|
| 984 |
+
"grad_norm": 4.456058979034424,
|
| 985 |
+
"learning_rate": 2.237192416296894e-05,
|
| 986 |
+
"loss": 0.21003019332885742,
|
| 987 |
+
"step": 6850
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 2.7833803953206937,
|
| 991 |
+
"grad_norm": 0.072988361120224,
|
| 992 |
+
"learning_rate": 2.2170229931423964e-05,
|
| 993 |
+
"loss": 0.17291305541992188,
|
| 994 |
+
"step": 6900
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 2.8035498184751915,
|
| 998 |
+
"grad_norm": 1.3505617380142212,
|
| 999 |
+
"learning_rate": 2.1968535699878985e-05,
|
| 1000 |
+
"loss": 0.16227161407470703,
|
| 1001 |
+
"step": 6950
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 2.8237192416296892,
|
| 1005 |
+
"grad_norm": 6.348002910614014,
|
| 1006 |
+
"learning_rate": 2.1766841468334005e-05,
|
| 1007 |
+
"loss": 0.18089292526245118,
|
| 1008 |
+
"step": 7000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 2.8438886647841874,
|
| 1012 |
+
"grad_norm": 4.224061965942383,
|
| 1013 |
+
"learning_rate": 2.156514723678903e-05,
|
| 1014 |
+
"loss": 0.21152067184448242,
|
| 1015 |
+
"step": 7050
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 2.8640580879386848,
|
| 1019 |
+
"grad_norm": 6.137605667114258,
|
| 1020 |
+
"learning_rate": 2.1363453005244053e-05,
|
| 1021 |
+
"loss": 0.1759657669067383,
|
| 1022 |
+
"step": 7100
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.884227511093183,
|
| 1026 |
+
"grad_norm": 0.8793334364891052,
|
| 1027 |
+
"learning_rate": 2.1161758773699074e-05,
|
| 1028 |
+
"loss": 0.1867251205444336,
|
| 1029 |
+
"step": 7150
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 2.9043969342476803,
|
| 1033 |
+
"grad_norm": 6.664191246032715,
|
| 1034 |
+
"learning_rate": 2.0960064542154095e-05,
|
| 1035 |
+
"loss": 0.16023876190185546,
|
| 1036 |
+
"step": 7200
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 2.9245663574021785,
|
| 1040 |
+
"grad_norm": 7.107497692108154,
|
| 1041 |
+
"learning_rate": 2.075837031060912e-05,
|
| 1042 |
+
"loss": 0.16876220703125,
|
| 1043 |
+
"step": 7250
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 2.944735780556676,
|
| 1047 |
+
"grad_norm": 6.290689468383789,
|
| 1048 |
+
"learning_rate": 2.0556676079064143e-05,
|
| 1049 |
+
"loss": 0.14115575790405274,
|
| 1050 |
+
"step": 7300
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 2.964905203711174,
|
| 1054 |
+
"grad_norm": 2.726917266845703,
|
| 1055 |
+
"learning_rate": 2.035498184751916e-05,
|
| 1056 |
+
"loss": 0.15181530952453615,
|
| 1057 |
+
"step": 7350
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 2.9850746268656714,
|
| 1061 |
+
"grad_norm": 0.1575377732515335,
|
| 1062 |
+
"learning_rate": 2.0153287615974184e-05,
|
| 1063 |
+
"loss": 0.204818058013916,
|
| 1064 |
+
"step": 7400
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 3.0,
|
| 1068 |
+
"eval_accuracy": 0.9469147005444646,
|
| 1069 |
+
"eval_loss": 0.20167165994644165,
|
| 1070 |
+
"eval_runtime": 20.7163,
|
| 1071 |
+
"eval_samples_per_second": 212.78,
|
| 1072 |
+
"eval_steps_per_second": 13.323,
|
| 1073 |
+
"step": 7437
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.0052440500201696,
|
| 1077 |
+
"grad_norm": 4.361955165863037,
|
| 1078 |
+
"learning_rate": 1.9951593384429208e-05,
|
| 1079 |
+
"loss": 0.1282938003540039,
|
| 1080 |
+
"step": 7450
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.0254134731746674,
|
| 1084 |
+
"grad_norm": 2.4413537979125977,
|
| 1085 |
+
"learning_rate": 1.974989915288423e-05,
|
| 1086 |
+
"loss": 0.14994653701782226,
|
| 1087 |
+
"step": 7500
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.045582896329165,
|
| 1091 |
+
"grad_norm": 0.9888622760772705,
|
| 1092 |
+
"learning_rate": 1.954820492133925e-05,
|
| 1093 |
+
"loss": 0.18875495910644532,
|
| 1094 |
+
"step": 7550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.065752319483663,
|
| 1098 |
+
"grad_norm": 1.7957770824432373,
|
| 1099 |
+
"learning_rate": 1.9346510689794273e-05,
|
| 1100 |
+
"loss": 0.16345306396484374,
|
| 1101 |
+
"step": 7600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.0859217426381607,
|
| 1105 |
+
"grad_norm": 10.348389625549316,
|
| 1106 |
+
"learning_rate": 1.9144816458249294e-05,
|
| 1107 |
+
"loss": 0.16646900177001953,
|
| 1108 |
+
"step": 7650
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.1060911657926584,
|
| 1112 |
+
"grad_norm": 1.3166776895523071,
|
| 1113 |
+
"learning_rate": 1.8943122226704318e-05,
|
| 1114 |
+
"loss": 0.11046904563903809,
|
| 1115 |
+
"step": 7700
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.126260588947156,
|
| 1119 |
+
"grad_norm": 4.864676475524902,
|
| 1120 |
+
"learning_rate": 1.874142799515934e-05,
|
| 1121 |
+
"loss": 0.14641575813293456,
|
| 1122 |
+
"step": 7750
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.146430012101654,
|
| 1126 |
+
"grad_norm": 0.590563178062439,
|
| 1127 |
+
"learning_rate": 1.8539733763614363e-05,
|
| 1128 |
+
"loss": 0.1471682071685791,
|
| 1129 |
+
"step": 7800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.1665994352561517,
|
| 1133 |
+
"grad_norm": 2.9072699546813965,
|
| 1134 |
+
"learning_rate": 1.8338039532069383e-05,
|
| 1135 |
+
"loss": 0.14556562423706054,
|
| 1136 |
+
"step": 7850
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.1867688584106495,
|
| 1140 |
+
"grad_norm": 1.3315081596374512,
|
| 1141 |
+
"learning_rate": 1.8136345300524407e-05,
|
| 1142 |
+
"loss": 0.09854414939880371,
|
| 1143 |
+
"step": 7900
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.2069382815651473,
|
| 1147 |
+
"grad_norm": 5.771864414215088,
|
| 1148 |
+
"learning_rate": 1.7934651068979428e-05,
|
| 1149 |
+
"loss": 0.18490333557128907,
|
| 1150 |
+
"step": 7950
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.227107704719645,
|
| 1154 |
+
"grad_norm": 1.54212486743927,
|
| 1155 |
+
"learning_rate": 1.773295683743445e-05,
|
| 1156 |
+
"loss": 0.11037788391113282,
|
| 1157 |
+
"step": 8000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.247277127874143,
|
| 1161 |
+
"grad_norm": 0.1329752802848816,
|
| 1162 |
+
"learning_rate": 1.7531262605889473e-05,
|
| 1163 |
+
"loss": 0.150218448638916,
|
| 1164 |
+
"step": 8050
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.2674465510286406,
|
| 1168 |
+
"grad_norm": 4.867387771606445,
|
| 1169 |
+
"learning_rate": 1.7329568374344497e-05,
|
| 1170 |
+
"loss": 0.17623374938964845,
|
| 1171 |
+
"step": 8100
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.2876159741831383,
|
| 1175 |
+
"grad_norm": 4.709106922149658,
|
| 1176 |
+
"learning_rate": 1.7127874142799517e-05,
|
| 1177 |
+
"loss": 0.1503092098236084,
|
| 1178 |
+
"step": 8150
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.307785397337636,
|
| 1182 |
+
"grad_norm": 2.4587321281433105,
|
| 1183 |
+
"learning_rate": 1.6926179911254538e-05,
|
| 1184 |
+
"loss": 0.17221961975097655,
|
| 1185 |
+
"step": 8200
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.327954820492134,
|
| 1189 |
+
"grad_norm": 0.18433253467082977,
|
| 1190 |
+
"learning_rate": 1.6724485679709562e-05,
|
| 1191 |
+
"loss": 0.15626630783081055,
|
| 1192 |
+
"step": 8250
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.3481242436466316,
|
| 1196 |
+
"grad_norm": 5.018091678619385,
|
| 1197 |
+
"learning_rate": 1.6522791448164586e-05,
|
| 1198 |
+
"loss": 0.11743727684020996,
|
| 1199 |
+
"step": 8300
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.3682936668011294,
|
| 1203 |
+
"grad_norm": 5.374708652496338,
|
| 1204 |
+
"learning_rate": 1.6321097216619603e-05,
|
| 1205 |
+
"loss": 0.1508438777923584,
|
| 1206 |
+
"step": 8350
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.388463089955627,
|
| 1210 |
+
"grad_norm": 2.271517753601074,
|
| 1211 |
+
"learning_rate": 1.6119402985074627e-05,
|
| 1212 |
+
"loss": 0.13781851768493653,
|
| 1213 |
+
"step": 8400
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.408632513110125,
|
| 1217 |
+
"grad_norm": 5.559137344360352,
|
| 1218 |
+
"learning_rate": 1.591770875352965e-05,
|
| 1219 |
+
"loss": 0.11309442520141602,
|
| 1220 |
+
"step": 8450
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.4288019362646227,
|
| 1224 |
+
"grad_norm": 0.41697025299072266,
|
| 1225 |
+
"learning_rate": 1.5716014521984672e-05,
|
| 1226 |
+
"loss": 0.1275617790222168,
|
| 1227 |
+
"step": 8500
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.4489713594191205,
|
| 1231 |
+
"grad_norm": 2.4398200511932373,
|
| 1232 |
+
"learning_rate": 1.5514320290439693e-05,
|
| 1233 |
+
"loss": 0.11772675514221191,
|
| 1234 |
+
"step": 8550
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.4691407825736182,
|
| 1238 |
+
"grad_norm": 5.258758544921875,
|
| 1239 |
+
"learning_rate": 1.5312626058894717e-05,
|
| 1240 |
+
"loss": 0.1622784423828125,
|
| 1241 |
+
"step": 8600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.489310205728116,
|
| 1245 |
+
"grad_norm": 3.7077977657318115,
|
| 1246 |
+
"learning_rate": 1.5110931827349739e-05,
|
| 1247 |
+
"loss": 0.17116220474243163,
|
| 1248 |
+
"step": 8650
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.509479628882614,
|
| 1252 |
+
"grad_norm": 0.45965898036956787,
|
| 1253 |
+
"learning_rate": 1.4909237595804761e-05,
|
| 1254 |
+
"loss": 0.16231245040893555,
|
| 1255 |
+
"step": 8700
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.5296490520371115,
|
| 1259 |
+
"grad_norm": 1.7838913202285767,
|
| 1260 |
+
"learning_rate": 1.4707543364259782e-05,
|
| 1261 |
+
"loss": 0.1521257972717285,
|
| 1262 |
+
"step": 8750
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.5498184751916098,
|
| 1266 |
+
"grad_norm": 0.8296908736228943,
|
| 1267 |
+
"learning_rate": 1.4505849132714804e-05,
|
| 1268 |
+
"loss": 0.1450238609313965,
|
| 1269 |
+
"step": 8800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.569987898346107,
|
| 1273 |
+
"grad_norm": 2.603304862976074,
|
| 1274 |
+
"learning_rate": 1.4304154901169828e-05,
|
| 1275 |
+
"loss": 0.18242509841918944,
|
| 1276 |
+
"step": 8850
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.5901573215006053,
|
| 1280 |
+
"grad_norm": 0.5085131525993347,
|
| 1281 |
+
"learning_rate": 1.410246066962485e-05,
|
| 1282 |
+
"loss": 0.1511705684661865,
|
| 1283 |
+
"step": 8900
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.6103267446551026,
|
| 1287 |
+
"grad_norm": 0.39615824818611145,
|
| 1288 |
+
"learning_rate": 1.3900766438079871e-05,
|
| 1289 |
+
"loss": 0.18586685180664062,
|
| 1290 |
+
"step": 8950
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.630496167809601,
|
| 1294 |
+
"grad_norm": 3.9840400218963623,
|
| 1295 |
+
"learning_rate": 1.3699072206534894e-05,
|
| 1296 |
+
"loss": 0.12510306358337403,
|
| 1297 |
+
"step": 9000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.650665590964098,
|
| 1301 |
+
"grad_norm": 1.856611728668213,
|
| 1302 |
+
"learning_rate": 1.3497377974989916e-05,
|
| 1303 |
+
"loss": 0.13308097839355468,
|
| 1304 |
+
"step": 9050
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.6708350141185964,
|
| 1308 |
+
"grad_norm": 3.8172595500946045,
|
| 1309 |
+
"learning_rate": 1.329568374344494e-05,
|
| 1310 |
+
"loss": 0.15418212890625,
|
| 1311 |
+
"step": 9100
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.691004437273094,
|
| 1315 |
+
"grad_norm": 1.3330752849578857,
|
| 1316 |
+
"learning_rate": 1.3093989511899959e-05,
|
| 1317 |
+
"loss": 0.11185816764831542,
|
| 1318 |
+
"step": 9150
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.711173860427592,
|
| 1322 |
+
"grad_norm": 5.0063652992248535,
|
| 1323 |
+
"learning_rate": 1.2892295280354983e-05,
|
| 1324 |
+
"loss": 0.14064055442810058,
|
| 1325 |
+
"step": 9200
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.7313432835820897,
|
| 1329 |
+
"grad_norm": 5.983749866485596,
|
| 1330 |
+
"learning_rate": 1.2690601048810005e-05,
|
| 1331 |
+
"loss": 0.1424097728729248,
|
| 1332 |
+
"step": 9250
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.7515127067365874,
|
| 1336 |
+
"grad_norm": 0.6912927627563477,
|
| 1337 |
+
"learning_rate": 1.2488906817265026e-05,
|
| 1338 |
+
"loss": 0.1355896282196045,
|
| 1339 |
+
"step": 9300
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.771682129891085,
|
| 1343 |
+
"grad_norm": 0.789875864982605,
|
| 1344 |
+
"learning_rate": 1.228721258572005e-05,
|
| 1345 |
+
"loss": 0.1569892692565918,
|
| 1346 |
+
"step": 9350
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.791851553045583,
|
| 1350 |
+
"grad_norm": 25.235803604125977,
|
| 1351 |
+
"learning_rate": 1.208551835417507e-05,
|
| 1352 |
+
"loss": 0.16975107192993164,
|
| 1353 |
+
"step": 9400
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.8120209762000807,
|
| 1357 |
+
"grad_norm": 10.807896614074707,
|
| 1358 |
+
"learning_rate": 1.1883824122630095e-05,
|
| 1359 |
+
"loss": 0.12455953598022461,
|
| 1360 |
+
"step": 9450
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.8321903993545785,
|
| 1364 |
+
"grad_norm": 1.9222596883773804,
|
| 1365 |
+
"learning_rate": 1.1682129891085115e-05,
|
| 1366 |
+
"loss": 0.13488195419311524,
|
| 1367 |
+
"step": 9500
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.8523598225090763,
|
| 1371 |
+
"grad_norm": 2.0477261543273926,
|
| 1372 |
+
"learning_rate": 1.1480435659540138e-05,
|
| 1373 |
+
"loss": 0.1264381980895996,
|
| 1374 |
+
"step": 9550
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.872529245663574,
|
| 1378 |
+
"grad_norm": 0.35994377732276917,
|
| 1379 |
+
"learning_rate": 1.127874142799516e-05,
|
| 1380 |
+
"loss": 0.1540088176727295,
|
| 1381 |
+
"step": 9600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.892698668818072,
|
| 1385 |
+
"grad_norm": 4.0362019538879395,
|
| 1386 |
+
"learning_rate": 1.1077047196450182e-05,
|
| 1387 |
+
"loss": 0.15407589912414552,
|
| 1388 |
+
"step": 9650
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.9128680919725696,
|
| 1392 |
+
"grad_norm": 0.061721403151750565,
|
| 1393 |
+
"learning_rate": 1.0875352964905205e-05,
|
| 1394 |
+
"loss": 0.11756512641906738,
|
| 1395 |
+
"step": 9700
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.9330375151270673,
|
| 1399 |
+
"grad_norm": 9.173218727111816,
|
| 1400 |
+
"learning_rate": 1.0673658733360227e-05,
|
| 1401 |
+
"loss": 0.12152119636535645,
|
| 1402 |
+
"step": 9750
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.953206938281565,
|
| 1406 |
+
"grad_norm": 1.0651715993881226,
|
| 1407 |
+
"learning_rate": 1.0471964501815248e-05,
|
| 1408 |
+
"loss": 0.13009054183959962,
|
| 1409 |
+
"step": 9800
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.973376361436063,
|
| 1413 |
+
"grad_norm": 8.925455093383789,
|
| 1414 |
+
"learning_rate": 1.0270270270270272e-05,
|
| 1415 |
+
"loss": 0.13634368896484375,
|
| 1416 |
+
"step": 9850
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.9935457845905606,
|
| 1420 |
+
"grad_norm": 7.055464267730713,
|
| 1421 |
+
"learning_rate": 1.0068576038725292e-05,
|
| 1422 |
+
"loss": 0.12510560989379882,
|
| 1423 |
+
"step": 9900
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 4.0,
|
| 1427 |
+
"eval_accuracy": 0.9498638838475499,
|
| 1428 |
+
"eval_loss": 0.19086875021457672,
|
| 1429 |
+
"eval_runtime": 20.6378,
|
| 1430 |
+
"eval_samples_per_second": 213.589,
|
| 1431 |
+
"eval_steps_per_second": 13.374,
|
| 1432 |
+
"step": 9916
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 4.013715207745059,
|
| 1436 |
+
"grad_norm": 1.8108693361282349,
|
| 1437 |
+
"learning_rate": 9.866881807180316e-06,
|
| 1438 |
+
"loss": 0.13720301628112794,
|
| 1439 |
+
"step": 9950
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 4.033884630899556,
|
| 1443 |
+
"grad_norm": 1.0627018213272095,
|
| 1444 |
+
"learning_rate": 9.665187575635337e-06,
|
| 1445 |
+
"loss": 0.13952345848083497,
|
| 1446 |
+
"step": 10000
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 4.054054054054054,
|
| 1450 |
+
"grad_norm": 0.9664111733436584,
|
| 1451 |
+
"learning_rate": 9.46349334409036e-06,
|
| 1452 |
+
"loss": 0.15156702995300292,
|
| 1453 |
+
"step": 10050
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 4.074223477208552,
|
| 1457 |
+
"grad_norm": 43.628841400146484,
|
| 1458 |
+
"learning_rate": 9.261799112545382e-06,
|
| 1459 |
+
"loss": 0.12908251762390135,
|
| 1460 |
+
"step": 10100
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 4.09439290036305,
|
| 1464 |
+
"grad_norm": 1.320526123046875,
|
| 1465 |
+
"learning_rate": 9.060104881000404e-06,
|
| 1466 |
+
"loss": 0.13279842376708983,
|
| 1467 |
+
"step": 10150
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 4.114562323517547,
|
| 1471 |
+
"grad_norm": 2.7222747802734375,
|
| 1472 |
+
"learning_rate": 8.858410649455426e-06,
|
| 1473 |
+
"loss": 0.0985030746459961,
|
| 1474 |
+
"step": 10200
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 4.1347317466720455,
|
| 1478 |
+
"grad_norm": 14.308988571166992,
|
| 1479 |
+
"learning_rate": 8.656716417910449e-06,
|
| 1480 |
+
"loss": 0.14445188522338867,
|
| 1481 |
+
"step": 10250
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 4.154901169826543,
|
| 1485 |
+
"grad_norm": 11.831380844116211,
|
| 1486 |
+
"learning_rate": 8.45502218636547e-06,
|
| 1487 |
+
"loss": 0.1422930145263672,
|
| 1488 |
+
"step": 10300
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 4.175070592981041,
|
| 1492 |
+
"grad_norm": 0.28589361906051636,
|
| 1493 |
+
"learning_rate": 8.253327954820494e-06,
|
| 1494 |
+
"loss": 0.12077164649963379,
|
| 1495 |
+
"step": 10350
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 4.195240016135538,
|
| 1499 |
+
"grad_norm": 1.5173996686935425,
|
| 1500 |
+
"learning_rate": 8.051633723275514e-06,
|
| 1501 |
+
"loss": 0.09480426788330078,
|
| 1502 |
+
"step": 10400
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 4.2154094392900365,
|
| 1506 |
+
"grad_norm": 3.5482044219970703,
|
| 1507 |
+
"learning_rate": 7.849939491730538e-06,
|
| 1508 |
+
"loss": 0.13487308502197265,
|
| 1509 |
+
"step": 10450
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 4.235578862444534,
|
| 1513 |
+
"grad_norm": 4.181003093719482,
|
| 1514 |
+
"learning_rate": 7.648245260185559e-06,
|
| 1515 |
+
"loss": 0.10856437683105469,
|
| 1516 |
+
"step": 10500
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 4.255748285599032,
|
| 1520 |
+
"grad_norm": 3.127241849899292,
|
| 1521 |
+
"learning_rate": 7.446551028640582e-06,
|
| 1522 |
+
"loss": 0.11362249374389649,
|
| 1523 |
+
"step": 10550
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 4.275917708753529,
|
| 1527 |
+
"grad_norm": 16.710121154785156,
|
| 1528 |
+
"learning_rate": 7.2448567970956035e-06,
|
| 1529 |
+
"loss": 0.16090269088745118,
|
| 1530 |
+
"step": 10600
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 4.296087131908028,
|
| 1534 |
+
"grad_norm": 7.735565662384033,
|
| 1535 |
+
"learning_rate": 7.043162565550626e-06,
|
| 1536 |
+
"loss": 0.1192931079864502,
|
| 1537 |
+
"step": 10650
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 4.316256555062525,
|
| 1541 |
+
"grad_norm": 1.4519258737564087,
|
| 1542 |
+
"learning_rate": 6.841468334005647e-06,
|
| 1543 |
+
"loss": 0.1244455623626709,
|
| 1544 |
+
"step": 10700
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 4.336425978217023,
|
| 1548 |
+
"grad_norm": 0.2084610015153885,
|
| 1549 |
+
"learning_rate": 6.6397741024606705e-06,
|
| 1550 |
+
"loss": 0.12691824913024902,
|
| 1551 |
+
"step": 10750
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 4.3565954013715205,
|
| 1555 |
+
"grad_norm": 5.141778469085693,
|
| 1556 |
+
"learning_rate": 6.438079870915692e-06,
|
| 1557 |
+
"loss": 0.15231054306030273,
|
| 1558 |
+
"step": 10800
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 4.376764824526019,
|
| 1562 |
+
"grad_norm": 4.018757343292236,
|
| 1563 |
+
"learning_rate": 6.236385639370714e-06,
|
| 1564 |
+
"loss": 0.13898849487304688,
|
| 1565 |
+
"step": 10850
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 4.396934247680516,
|
| 1569 |
+
"grad_norm": 2.3020405769348145,
|
| 1570 |
+
"learning_rate": 6.034691407825737e-06,
|
| 1571 |
+
"loss": 0.11184153556823731,
|
| 1572 |
+
"step": 10900
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 4.417103670835014,
|
| 1576 |
+
"grad_norm": 0.3617361783981323,
|
| 1577 |
+
"learning_rate": 5.832997176280758e-06,
|
| 1578 |
+
"loss": 0.12001494407653808,
|
| 1579 |
+
"step": 10950
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 4.4372730939895115,
|
| 1583 |
+
"grad_norm": 3.145230531692505,
|
| 1584 |
+
"learning_rate": 5.6313029447357805e-06,
|
| 1585 |
+
"loss": 0.111842679977417,
|
| 1586 |
+
"step": 11000
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 4.45744251714401,
|
| 1590 |
+
"grad_norm": 6.321645736694336,
|
| 1591 |
+
"learning_rate": 5.429608713190803e-06,
|
| 1592 |
+
"loss": 0.1446524429321289,
|
| 1593 |
+
"step": 11050
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 4.477611940298507,
|
| 1597 |
+
"grad_norm": 0.19845882058143616,
|
| 1598 |
+
"learning_rate": 5.227914481645825e-06,
|
| 1599 |
+
"loss": 0.10471660614013673,
|
| 1600 |
+
"step": 11100
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 4.497781363453005,
|
| 1604 |
+
"grad_norm": 14.234234809875488,
|
| 1605 |
+
"learning_rate": 5.0262202501008476e-06,
|
| 1606 |
+
"loss": 0.13982812881469728,
|
| 1607 |
+
"step": 11150
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 4.5179507866075035,
|
| 1611 |
+
"grad_norm": 6.142904281616211,
|
| 1612 |
+
"learning_rate": 4.824526018555869e-06,
|
| 1613 |
+
"loss": 0.14638319969177246,
|
| 1614 |
+
"step": 11200
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 4.538120209762001,
|
| 1618 |
+
"grad_norm": 10.167677879333496,
|
| 1619 |
+
"learning_rate": 4.622831787010891e-06,
|
| 1620 |
+
"loss": 0.12575898170471192,
|
| 1621 |
+
"step": 11250
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 4.558289632916498,
|
| 1625 |
+
"grad_norm": 0.4165036082267761,
|
| 1626 |
+
"learning_rate": 4.421137555465914e-06,
|
| 1627 |
+
"loss": 0.10566370964050292,
|
| 1628 |
+
"step": 11300
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 4.578459056070996,
|
| 1632 |
+
"grad_norm": 0.9815388321876526,
|
| 1633 |
+
"learning_rate": 4.219443323920936e-06,
|
| 1634 |
+
"loss": 0.1358126163482666,
|
| 1635 |
+
"step": 11350
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 4.598628479225495,
|
| 1639 |
+
"grad_norm": 3.1989729404449463,
|
| 1640 |
+
"learning_rate": 4.017749092375958e-06,
|
| 1641 |
+
"loss": 0.1418777275085449,
|
| 1642 |
+
"step": 11400
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 4.618797902379992,
|
| 1646 |
+
"grad_norm": 2.143892288208008,
|
| 1647 |
+
"learning_rate": 3.81605486083098e-06,
|
| 1648 |
+
"loss": 0.13869763374328614,
|
| 1649 |
+
"step": 11450
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 4.63896732553449,
|
| 1653 |
+
"grad_norm": 1.0389790534973145,
|
| 1654 |
+
"learning_rate": 3.6143606292860022e-06,
|
| 1655 |
+
"loss": 0.20401397705078125,
|
| 1656 |
+
"step": 11500
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 4.659136748688987,
|
| 1660 |
+
"grad_norm": 8.121844291687012,
|
| 1661 |
+
"learning_rate": 3.4126663977410246e-06,
|
| 1662 |
+
"loss": 0.13985713005065917,
|
| 1663 |
+
"step": 11550
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 4.679306171843486,
|
| 1667 |
+
"grad_norm": 18.19819450378418,
|
| 1668 |
+
"learning_rate": 3.210972166196047e-06,
|
| 1669 |
+
"loss": 0.12457592010498048,
|
| 1670 |
+
"step": 11600
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 4.699475594997983,
|
| 1674 |
+
"grad_norm": 0.9452546238899231,
|
| 1675 |
+
"learning_rate": 3.009277934651069e-06,
|
| 1676 |
+
"loss": 0.14423842430114747,
|
| 1677 |
+
"step": 11650
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 4.719645018152481,
|
| 1681 |
+
"grad_norm": 0.4873313307762146,
|
| 1682 |
+
"learning_rate": 2.807583703106091e-06,
|
| 1683 |
+
"loss": 0.11125797271728516,
|
| 1684 |
+
"step": 11700
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 4.7398144413069785,
|
| 1688 |
+
"grad_norm": 0.23056188225746155,
|
| 1689 |
+
"learning_rate": 2.6058894715611135e-06,
|
| 1690 |
+
"loss": 0.10489352226257324,
|
| 1691 |
+
"step": 11750
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 4.759983864461477,
|
| 1695 |
+
"grad_norm": 4.3498921394348145,
|
| 1696 |
+
"learning_rate": 2.4041952400161354e-06,
|
| 1697 |
+
"loss": 0.11880455017089844,
|
| 1698 |
+
"step": 11800
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 4.780153287615974,
|
| 1702 |
+
"grad_norm": 6.197351932525635,
|
| 1703 |
+
"learning_rate": 2.2025010084711578e-06,
|
| 1704 |
+
"loss": 0.11629629135131836,
|
| 1705 |
+
"step": 11850
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 4.800322710770472,
|
| 1709 |
+
"grad_norm": 1.4869271516799927,
|
| 1710 |
+
"learning_rate": 2.0008067769261797e-06,
|
| 1711 |
+
"loss": 0.06002850532531738,
|
| 1712 |
+
"step": 11900
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 4.82049213392497,
|
| 1716 |
+
"grad_norm": 0.11449511349201202,
|
| 1717 |
+
"learning_rate": 1.799112545381202e-06,
|
| 1718 |
+
"loss": 0.09682181358337402,
|
| 1719 |
+
"step": 11950
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 4.840661557079468,
|
| 1723 |
+
"grad_norm": 0.537159264087677,
|
| 1724 |
+
"learning_rate": 1.5974183138362242e-06,
|
| 1725 |
+
"loss": 0.1060289478302002,
|
| 1726 |
+
"step": 12000
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 4.860830980233965,
|
| 1730 |
+
"grad_norm": 1.7972829341888428,
|
| 1731 |
+
"learning_rate": 1.3957240822912465e-06,
|
| 1732 |
+
"loss": 0.1122105884552002,
|
| 1733 |
+
"step": 12050
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 4.881000403388463,
|
| 1737 |
+
"grad_norm": 0.5588096976280212,
|
| 1738 |
+
"learning_rate": 1.1940298507462686e-06,
|
| 1739 |
+
"loss": 0.20083948135375976,
|
| 1740 |
+
"step": 12100
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 4.901169826542961,
|
| 1744 |
+
"grad_norm": 0.34725531935691833,
|
| 1745 |
+
"learning_rate": 9.92335619201291e-07,
|
| 1746 |
+
"loss": 0.1103892421722412,
|
| 1747 |
+
"step": 12150
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 4.921339249697459,
|
| 1751 |
+
"grad_norm": 4.558995246887207,
|
| 1752 |
+
"learning_rate": 7.906413876563131e-07,
|
| 1753 |
+
"loss": 0.09908536911010742,
|
| 1754 |
+
"step": 12200
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 4.941508672851956,
|
| 1758 |
+
"grad_norm": 8.438501358032227,
|
| 1759 |
+
"learning_rate": 5.889471561113352e-07,
|
| 1760 |
+
"loss": 0.13646589279174803,
|
| 1761 |
+
"step": 12250
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 4.961678096006454,
|
| 1765 |
+
"grad_norm": 4.4897847175598145,
|
| 1766 |
+
"learning_rate": 3.872529245663574e-07,
|
| 1767 |
+
"loss": 0.11858689308166503,
|
| 1768 |
+
"step": 12300
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 4.981847519160952,
|
| 1772 |
+
"grad_norm": 3.2453246116638184,
|
| 1773 |
+
"learning_rate": 1.8555869302137961e-07,
|
| 1774 |
+
"loss": 0.11941192626953125,
|
| 1775 |
+
"step": 12350
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 5.0,
|
| 1779 |
+
"eval_accuracy": 0.9532667876588021,
|
| 1780 |
+
"eval_loss": 0.1782979518175125,
|
| 1781 |
+
"eval_runtime": 20.7187,
|
| 1782 |
+
"eval_samples_per_second": 212.755,
|
| 1783 |
+
"eval_steps_per_second": 13.321,
|
| 1784 |
+
"step": 12395
|
| 1785 |
+
}
|
| 1786 |
+
],
|
| 1787 |
+
"logging_steps": 50,
|
| 1788 |
+
"max_steps": 12395,
|
| 1789 |
+
"num_input_tokens_seen": 0,
|
| 1790 |
+
"num_train_epochs": 5,
|
| 1791 |
+
"save_steps": 500,
|
| 1792 |
+
"stateful_callbacks": {
|
| 1793 |
+
"TrainerControl": {
|
| 1794 |
+
"args": {
|
| 1795 |
+
"should_epoch_stop": false,
|
| 1796 |
+
"should_evaluate": false,
|
| 1797 |
+
"should_log": false,
|
| 1798 |
+
"should_save": true,
|
| 1799 |
+
"should_training_stop": true
|
| 1800 |
+
},
|
| 1801 |
+
"attributes": {}
|
| 1802 |
+
}
|
| 1803 |
+
},
|
| 1804 |
+
"total_flos": 4.227186605693829e+18,
|
| 1805 |
+
"train_batch_size": 16,
|
| 1806 |
+
"trial_name": null,
|
| 1807 |
+
"trial_params": null
|
| 1808 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1d0467e8a53923748f6b194d5dfdb02ead40650cb9b4ba236a02fe631ff6f03
|
| 3 |
+
size 5201
|