Upload 9 files
Browse filesAdd tiny-vit model files for smart tocha generico
- config.json +34 -0
- optimizer.pt +3 -0
- preprocessor_config.json +17 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1069 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "WinKawaks/vit-tiny-patch16-224",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.0,
|
| 9 |
+
"hidden_size": 192,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "exc_vapor",
|
| 12 |
+
"1": "fuligem",
|
| 13 |
+
"2": "negative",
|
| 14 |
+
"3": "otimizado"
|
| 15 |
+
},
|
| 16 |
+
"image_size": 224,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 768,
|
| 19 |
+
"label2id": {
|
| 20 |
+
"exc_vapor": "0",
|
| 21 |
+
"fuligem": "1",
|
| 22 |
+
"negative": "2",
|
| 23 |
+
"otimizado": "3"
|
| 24 |
+
},
|
| 25 |
+
"layer_norm_eps": 1e-12,
|
| 26 |
+
"model_type": "vit",
|
| 27 |
+
"num_attention_heads": 3,
|
| 28 |
+
"num_channels": 3,
|
| 29 |
+
"num_hidden_layers": 12,
|
| 30 |
+
"patch_size": 16,
|
| 31 |
+
"qkv_bias": true,
|
| 32 |
+
"torch_dtype": "float32",
|
| 33 |
+
"transformers_version": "4.16.0"
|
| 34 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:008ff9bdba5a9b9a936354e86250b90f4f55923c48ac6cdf816f4b2c5112a5f0
|
| 3 |
+
size 44316513
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"do_resize": true,
|
| 4 |
+
"feature_extractor_type": "ViTFeatureExtractor",
|
| 5 |
+
"image_mean": [
|
| 6 |
+
0.5,
|
| 7 |
+
0.5,
|
| 8 |
+
0.5
|
| 9 |
+
],
|
| 10 |
+
"image_std": [
|
| 11 |
+
0.5,
|
| 12 |
+
0.5,
|
| 13 |
+
0.5
|
| 14 |
+
],
|
| 15 |
+
"resample": 2,
|
| 16 |
+
"size": 224
|
| 17 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15bb6d582f30e4f860a2af9c9e186cc46be4c1edb7d69dcfa278e84d3cae2f23
|
| 3 |
+
size 22179313
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d3f7a96453c8fd3beea5c47baeb0dc026ad208e89e2e3911bc400e4f4e42782
|
| 3 |
+
size 14503
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e29a321fc2c63ed1a38fe772c8d672ef9703b4a84397279224337d0d70238d7
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3c36d6c1e59d19d1d7556aad542f5e8c71ce179a42cc792841576420e3a0a85
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1069 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.020489266142249107,
|
| 3 |
+
"best_model_checkpoint": "./models/tiny-vit-GENERICO/checkpoint-1350",
|
| 4 |
+
"epoch": 1.967930029154519,
|
| 5 |
+
"global_step": 1350,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.0001991253644314869,
|
| 13 |
+
"loss": 2.9673,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 0.00019766763848396503,
|
| 19 |
+
"loss": 0.551,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.00019620991253644317,
|
| 25 |
+
"loss": 0.4255,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.06,
|
| 30 |
+
"learning_rate": 0.00019475218658892128,
|
| 31 |
+
"loss": 0.3714,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.07,
|
| 36 |
+
"learning_rate": 0.00019329446064139942,
|
| 37 |
+
"loss": 0.4272,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"eval_accuracy": 0.9147173489278753,
|
| 43 |
+
"eval_loss": 0.24443647265434265,
|
| 44 |
+
"eval_runtime": 286.3815,
|
| 45 |
+
"eval_samples_per_second": 14.331,
|
| 46 |
+
"eval_steps_per_second": 1.791,
|
| 47 |
+
"step": 50
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.09,
|
| 51 |
+
"learning_rate": 0.00019183673469387756,
|
| 52 |
+
"loss": 0.4202,
|
| 53 |
+
"step": 60
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.1,
|
| 57 |
+
"learning_rate": 0.0001903790087463557,
|
| 58 |
+
"loss": 0.2785,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.12,
|
| 63 |
+
"learning_rate": 0.00018892128279883382,
|
| 64 |
+
"loss": 0.2276,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.13,
|
| 69 |
+
"learning_rate": 0.00018746355685131196,
|
| 70 |
+
"loss": 0.2418,
|
| 71 |
+
"step": 90
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 0.15,
|
| 75 |
+
"learning_rate": 0.0001860058309037901,
|
| 76 |
+
"loss": 0.2214,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.15,
|
| 81 |
+
"eval_accuracy": 0.8379629629629629,
|
| 82 |
+
"eval_loss": 0.4828641414642334,
|
| 83 |
+
"eval_runtime": 278.5231,
|
| 84 |
+
"eval_samples_per_second": 14.735,
|
| 85 |
+
"eval_steps_per_second": 1.842,
|
| 86 |
+
"step": 100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.16,
|
| 90 |
+
"learning_rate": 0.00018454810495626821,
|
| 91 |
+
"loss": 0.2319,
|
| 92 |
+
"step": 110
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.17,
|
| 96 |
+
"learning_rate": 0.00018309037900874638,
|
| 97 |
+
"loss": 0.2103,
|
| 98 |
+
"step": 120
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 0.0001816326530612245,
|
| 103 |
+
"loss": 0.2667,
|
| 104 |
+
"step": 130
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 0.00018017492711370264,
|
| 109 |
+
"loss": 0.3514,
|
| 110 |
+
"step": 140
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"learning_rate": 0.00017871720116618075,
|
| 115 |
+
"loss": 0.3004,
|
| 116 |
+
"step": 150
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.22,
|
| 120 |
+
"eval_accuracy": 0.935672514619883,
|
| 121 |
+
"eval_loss": 0.18615978956222534,
|
| 122 |
+
"eval_runtime": 210.6691,
|
| 123 |
+
"eval_samples_per_second": 19.481,
|
| 124 |
+
"eval_steps_per_second": 2.435,
|
| 125 |
+
"step": 150
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.23,
|
| 129 |
+
"learning_rate": 0.00017725947521865892,
|
| 130 |
+
"loss": 0.2605,
|
| 131 |
+
"step": 160
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.25,
|
| 135 |
+
"learning_rate": 0.00017580174927113703,
|
| 136 |
+
"loss": 0.1776,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.26,
|
| 141 |
+
"learning_rate": 0.00017434402332361517,
|
| 142 |
+
"loss": 0.1727,
|
| 143 |
+
"step": 180
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.28,
|
| 147 |
+
"learning_rate": 0.00017288629737609328,
|
| 148 |
+
"loss": 0.191,
|
| 149 |
+
"step": 190
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.29,
|
| 153 |
+
"learning_rate": 0.00017142857142857143,
|
| 154 |
+
"loss": 0.1255,
|
| 155 |
+
"step": 200
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.29,
|
| 159 |
+
"eval_accuracy": 0.9432261208576999,
|
| 160 |
+
"eval_loss": 0.19172662496566772,
|
| 161 |
+
"eval_runtime": 227.2135,
|
| 162 |
+
"eval_samples_per_second": 18.062,
|
| 163 |
+
"eval_steps_per_second": 2.258,
|
| 164 |
+
"step": 200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.31,
|
| 168 |
+
"learning_rate": 0.00016997084548104957,
|
| 169 |
+
"loss": 0.1654,
|
| 170 |
+
"step": 210
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.32,
|
| 174 |
+
"learning_rate": 0.0001685131195335277,
|
| 175 |
+
"loss": 0.2456,
|
| 176 |
+
"step": 220
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.34,
|
| 180 |
+
"learning_rate": 0.00016705539358600585,
|
| 181 |
+
"loss": 0.2103,
|
| 182 |
+
"step": 230
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.35,
|
| 186 |
+
"learning_rate": 0.00016559766763848396,
|
| 187 |
+
"loss": 0.2826,
|
| 188 |
+
"step": 240
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.36,
|
| 192 |
+
"learning_rate": 0.00016413994169096213,
|
| 193 |
+
"loss": 0.2402,
|
| 194 |
+
"step": 250
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.36,
|
| 198 |
+
"eval_accuracy": 0.9522417153996101,
|
| 199 |
+
"eval_loss": 0.14588728547096252,
|
| 200 |
+
"eval_runtime": 228.2383,
|
| 201 |
+
"eval_samples_per_second": 17.981,
|
| 202 |
+
"eval_steps_per_second": 2.248,
|
| 203 |
+
"step": 250
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.38,
|
| 207 |
+
"learning_rate": 0.00016268221574344024,
|
| 208 |
+
"loss": 0.2109,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.39,
|
| 213 |
+
"learning_rate": 0.00016122448979591838,
|
| 214 |
+
"loss": 0.2142,
|
| 215 |
+
"step": 270
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.41,
|
| 219 |
+
"learning_rate": 0.0001597667638483965,
|
| 220 |
+
"loss": 0.1838,
|
| 221 |
+
"step": 280
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.42,
|
| 225 |
+
"learning_rate": 0.00015830903790087464,
|
| 226 |
+
"loss": 0.1783,
|
| 227 |
+
"step": 290
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.44,
|
| 231 |
+
"learning_rate": 0.00015685131195335278,
|
| 232 |
+
"loss": 0.059,
|
| 233 |
+
"step": 300
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.44,
|
| 237 |
+
"eval_accuracy": 0.9485867446393762,
|
| 238 |
+
"eval_loss": 0.1581156998872757,
|
| 239 |
+
"eval_runtime": 222.9736,
|
| 240 |
+
"eval_samples_per_second": 18.406,
|
| 241 |
+
"eval_steps_per_second": 2.301,
|
| 242 |
+
"step": 300
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.45,
|
| 246 |
+
"learning_rate": 0.00015539358600583092,
|
| 247 |
+
"loss": 0.1959,
|
| 248 |
+
"step": 310
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.47,
|
| 252 |
+
"learning_rate": 0.00015393586005830903,
|
| 253 |
+
"loss": 0.0957,
|
| 254 |
+
"step": 320
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.48,
|
| 258 |
+
"learning_rate": 0.00015247813411078717,
|
| 259 |
+
"loss": 0.1786,
|
| 260 |
+
"step": 330
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.5,
|
| 264 |
+
"learning_rate": 0.0001510204081632653,
|
| 265 |
+
"loss": 0.1695,
|
| 266 |
+
"step": 340
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.51,
|
| 270 |
+
"learning_rate": 0.00014956268221574345,
|
| 271 |
+
"loss": 0.1513,
|
| 272 |
+
"step": 350
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.51,
|
| 276 |
+
"eval_accuracy": 0.9627192982456141,
|
| 277 |
+
"eval_loss": 0.11844132840633392,
|
| 278 |
+
"eval_runtime": 238.6523,
|
| 279 |
+
"eval_samples_per_second": 17.197,
|
| 280 |
+
"eval_steps_per_second": 2.15,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.52,
|
| 285 |
+
"learning_rate": 0.0001481049562682216,
|
| 286 |
+
"loss": 0.1532,
|
| 287 |
+
"step": 360
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.54,
|
| 291 |
+
"learning_rate": 0.0001466472303206997,
|
| 292 |
+
"loss": 0.125,
|
| 293 |
+
"step": 370
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.55,
|
| 297 |
+
"learning_rate": 0.00014518950437317785,
|
| 298 |
+
"loss": 0.1735,
|
| 299 |
+
"step": 380
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.57,
|
| 303 |
+
"learning_rate": 0.000143731778425656,
|
| 304 |
+
"loss": 0.1229,
|
| 305 |
+
"step": 390
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.58,
|
| 309 |
+
"learning_rate": 0.00014227405247813413,
|
| 310 |
+
"loss": 0.1444,
|
| 311 |
+
"step": 400
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.58,
|
| 315 |
+
"eval_accuracy": 0.9517543859649122,
|
| 316 |
+
"eval_loss": 0.1490934193134308,
|
| 317 |
+
"eval_runtime": 287.634,
|
| 318 |
+
"eval_samples_per_second": 14.268,
|
| 319 |
+
"eval_steps_per_second": 1.784,
|
| 320 |
+
"step": 400
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.6,
|
| 324 |
+
"learning_rate": 0.00014081632653061224,
|
| 325 |
+
"loss": 0.1257,
|
| 326 |
+
"step": 410
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.61,
|
| 330 |
+
"learning_rate": 0.00013935860058309038,
|
| 331 |
+
"loss": 0.1251,
|
| 332 |
+
"step": 420
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.63,
|
| 336 |
+
"learning_rate": 0.00013790087463556852,
|
| 337 |
+
"loss": 0.0909,
|
| 338 |
+
"step": 430
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.64,
|
| 342 |
+
"learning_rate": 0.00013644314868804666,
|
| 343 |
+
"loss": 0.1236,
|
| 344 |
+
"step": 440
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.66,
|
| 348 |
+
"learning_rate": 0.00013498542274052478,
|
| 349 |
+
"loss": 0.1385,
|
| 350 |
+
"step": 450
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.66,
|
| 354 |
+
"eval_accuracy": 0.9675925925925926,
|
| 355 |
+
"eval_loss": 0.09978482127189636,
|
| 356 |
+
"eval_runtime": 326.8281,
|
| 357 |
+
"eval_samples_per_second": 12.557,
|
| 358 |
+
"eval_steps_per_second": 1.57,
|
| 359 |
+
"step": 450
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.67,
|
| 363 |
+
"learning_rate": 0.00013352769679300292,
|
| 364 |
+
"loss": 0.1466,
|
| 365 |
+
"step": 460
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.69,
|
| 369 |
+
"learning_rate": 0.00013206997084548106,
|
| 370 |
+
"loss": 0.1286,
|
| 371 |
+
"step": 470
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.7,
|
| 375 |
+
"learning_rate": 0.00013061224489795917,
|
| 376 |
+
"loss": 0.1099,
|
| 377 |
+
"step": 480
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.71,
|
| 381 |
+
"learning_rate": 0.00012915451895043734,
|
| 382 |
+
"loss": 0.1559,
|
| 383 |
+
"step": 490
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.73,
|
| 387 |
+
"learning_rate": 0.00012769679300291545,
|
| 388 |
+
"loss": 0.1263,
|
| 389 |
+
"step": 500
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.73,
|
| 393 |
+
"eval_accuracy": 0.9624756335282652,
|
| 394 |
+
"eval_loss": 0.11673439294099808,
|
| 395 |
+
"eval_runtime": 291.2631,
|
| 396 |
+
"eval_samples_per_second": 14.09,
|
| 397 |
+
"eval_steps_per_second": 1.761,
|
| 398 |
+
"step": 500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.74,
|
| 402 |
+
"learning_rate": 0.0001262390670553936,
|
| 403 |
+
"loss": 0.2311,
|
| 404 |
+
"step": 510
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.76,
|
| 408 |
+
"learning_rate": 0.0001247813411078717,
|
| 409 |
+
"loss": 0.103,
|
| 410 |
+
"step": 520
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.77,
|
| 414 |
+
"learning_rate": 0.00012332361516034988,
|
| 415 |
+
"loss": 0.0728,
|
| 416 |
+
"step": 530
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.79,
|
| 420 |
+
"learning_rate": 0.000121865889212828,
|
| 421 |
+
"loss": 0.1353,
|
| 422 |
+
"step": 540
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.8,
|
| 426 |
+
"learning_rate": 0.00012040816326530613,
|
| 427 |
+
"loss": 0.1307,
|
| 428 |
+
"step": 550
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.8,
|
| 432 |
+
"eval_accuracy": 0.9471247563352827,
|
| 433 |
+
"eval_loss": 0.16630055010318756,
|
| 434 |
+
"eval_runtime": 312.9758,
|
| 435 |
+
"eval_samples_per_second": 13.113,
|
| 436 |
+
"eval_steps_per_second": 1.639,
|
| 437 |
+
"step": 550
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.82,
|
| 441 |
+
"learning_rate": 0.00011895043731778426,
|
| 442 |
+
"loss": 0.1407,
|
| 443 |
+
"step": 560
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.83,
|
| 447 |
+
"learning_rate": 0.00011749271137026238,
|
| 448 |
+
"loss": 0.081,
|
| 449 |
+
"step": 570
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.85,
|
| 453 |
+
"learning_rate": 0.00011603498542274054,
|
| 454 |
+
"loss": 0.1272,
|
| 455 |
+
"step": 580
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.86,
|
| 459 |
+
"learning_rate": 0.00011457725947521867,
|
| 460 |
+
"loss": 0.1224,
|
| 461 |
+
"step": 590
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.87,
|
| 465 |
+
"learning_rate": 0.0001131195335276968,
|
| 466 |
+
"loss": 0.1517,
|
| 467 |
+
"step": 600
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.87,
|
| 471 |
+
"eval_accuracy": 0.9451754385964912,
|
| 472 |
+
"eval_loss": 0.18923337757587433,
|
| 473 |
+
"eval_runtime": 328.1373,
|
| 474 |
+
"eval_samples_per_second": 12.507,
|
| 475 |
+
"eval_steps_per_second": 1.563,
|
| 476 |
+
"step": 600
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.89,
|
| 480 |
+
"learning_rate": 0.00011166180758017492,
|
| 481 |
+
"loss": 0.1336,
|
| 482 |
+
"step": 610
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.9,
|
| 486 |
+
"learning_rate": 0.00011020408163265306,
|
| 487 |
+
"loss": 0.1245,
|
| 488 |
+
"step": 620
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.92,
|
| 492 |
+
"learning_rate": 0.0001087463556851312,
|
| 493 |
+
"loss": 0.0762,
|
| 494 |
+
"step": 630
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.93,
|
| 498 |
+
"learning_rate": 0.00010728862973760934,
|
| 499 |
+
"loss": 0.1901,
|
| 500 |
+
"step": 640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.95,
|
| 504 |
+
"learning_rate": 0.00010583090379008747,
|
| 505 |
+
"loss": 0.1051,
|
| 506 |
+
"step": 650
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.95,
|
| 510 |
+
"eval_accuracy": 0.973196881091618,
|
| 511 |
+
"eval_loss": 0.07433269917964935,
|
| 512 |
+
"eval_runtime": 295.1825,
|
| 513 |
+
"eval_samples_per_second": 13.903,
|
| 514 |
+
"eval_steps_per_second": 1.738,
|
| 515 |
+
"step": 650
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.96,
|
| 519 |
+
"learning_rate": 0.0001043731778425656,
|
| 520 |
+
"loss": 0.0844,
|
| 521 |
+
"step": 660
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.98,
|
| 525 |
+
"learning_rate": 0.00010291545189504375,
|
| 526 |
+
"loss": 0.0764,
|
| 527 |
+
"step": 670
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.99,
|
| 531 |
+
"learning_rate": 0.00010145772594752188,
|
| 532 |
+
"loss": 0.0786,
|
| 533 |
+
"step": 680
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.01,
|
| 537 |
+
"learning_rate": 0.0001,
|
| 538 |
+
"loss": 0.0626,
|
| 539 |
+
"step": 690
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.02,
|
| 543 |
+
"learning_rate": 9.854227405247813e-05,
|
| 544 |
+
"loss": 0.04,
|
| 545 |
+
"step": 700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.02,
|
| 549 |
+
"eval_accuracy": 0.979775828460039,
|
| 550 |
+
"eval_loss": 0.06021393463015556,
|
| 551 |
+
"eval_runtime": 294.0751,
|
| 552 |
+
"eval_samples_per_second": 13.956,
|
| 553 |
+
"eval_steps_per_second": 1.744,
|
| 554 |
+
"step": 700
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.03,
|
| 558 |
+
"learning_rate": 9.708454810495627e-05,
|
| 559 |
+
"loss": 0.0473,
|
| 560 |
+
"step": 710
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.05,
|
| 564 |
+
"learning_rate": 9.56268221574344e-05,
|
| 565 |
+
"loss": 0.101,
|
| 566 |
+
"step": 720
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.06,
|
| 570 |
+
"learning_rate": 9.416909620991254e-05,
|
| 571 |
+
"loss": 0.0604,
|
| 572 |
+
"step": 730
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.08,
|
| 576 |
+
"learning_rate": 9.271137026239067e-05,
|
| 577 |
+
"loss": 0.0917,
|
| 578 |
+
"step": 740
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.09,
|
| 582 |
+
"learning_rate": 9.125364431486881e-05,
|
| 583 |
+
"loss": 0.047,
|
| 584 |
+
"step": 750
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.09,
|
| 588 |
+
"eval_accuracy": 0.976364522417154,
|
| 589 |
+
"eval_loss": 0.07237815111875534,
|
| 590 |
+
"eval_runtime": 314.914,
|
| 591 |
+
"eval_samples_per_second": 13.032,
|
| 592 |
+
"eval_steps_per_second": 1.629,
|
| 593 |
+
"step": 750
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.11,
|
| 597 |
+
"learning_rate": 8.979591836734695e-05,
|
| 598 |
+
"loss": 0.0216,
|
| 599 |
+
"step": 760
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.12,
|
| 603 |
+
"learning_rate": 8.833819241982508e-05,
|
| 604 |
+
"loss": 0.0568,
|
| 605 |
+
"step": 770
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.14,
|
| 609 |
+
"learning_rate": 8.688046647230322e-05,
|
| 610 |
+
"loss": 0.0664,
|
| 611 |
+
"step": 780
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.15,
|
| 615 |
+
"learning_rate": 8.542274052478134e-05,
|
| 616 |
+
"loss": 0.0488,
|
| 617 |
+
"step": 790
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 1.17,
|
| 621 |
+
"learning_rate": 8.396501457725948e-05,
|
| 622 |
+
"loss": 0.0893,
|
| 623 |
+
"step": 800
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 1.17,
|
| 627 |
+
"eval_accuracy": 0.9834307992202729,
|
| 628 |
+
"eval_loss": 0.05777524411678314,
|
| 629 |
+
"eval_runtime": 310.6824,
|
| 630 |
+
"eval_samples_per_second": 13.21,
|
| 631 |
+
"eval_steps_per_second": 1.651,
|
| 632 |
+
"step": 800
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.18,
|
| 636 |
+
"learning_rate": 8.250728862973761e-05,
|
| 637 |
+
"loss": 0.0414,
|
| 638 |
+
"step": 810
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.2,
|
| 642 |
+
"learning_rate": 8.104956268221575e-05,
|
| 643 |
+
"loss": 0.0306,
|
| 644 |
+
"step": 820
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.21,
|
| 648 |
+
"learning_rate": 7.959183673469388e-05,
|
| 649 |
+
"loss": 0.0164,
|
| 650 |
+
"step": 830
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.22,
|
| 654 |
+
"learning_rate": 7.8134110787172e-05,
|
| 655 |
+
"loss": 0.0465,
|
| 656 |
+
"step": 840
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.24,
|
| 660 |
+
"learning_rate": 7.667638483965015e-05,
|
| 661 |
+
"loss": 0.0254,
|
| 662 |
+
"step": 850
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.24,
|
| 666 |
+
"eval_accuracy": 0.9790448343079922,
|
| 667 |
+
"eval_loss": 0.09394790232181549,
|
| 668 |
+
"eval_runtime": 292.1931,
|
| 669 |
+
"eval_samples_per_second": 14.046,
|
| 670 |
+
"eval_steps_per_second": 1.756,
|
| 671 |
+
"step": 850
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.25,
|
| 675 |
+
"learning_rate": 7.521865889212827e-05,
|
| 676 |
+
"loss": 0.049,
|
| 677 |
+
"step": 860
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.27,
|
| 681 |
+
"learning_rate": 7.376093294460641e-05,
|
| 682 |
+
"loss": 0.0482,
|
| 683 |
+
"step": 870
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.28,
|
| 687 |
+
"learning_rate": 7.230320699708455e-05,
|
| 688 |
+
"loss": 0.1339,
|
| 689 |
+
"step": 880
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.3,
|
| 693 |
+
"learning_rate": 7.08454810495627e-05,
|
| 694 |
+
"loss": 0.0658,
|
| 695 |
+
"step": 890
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.31,
|
| 699 |
+
"learning_rate": 6.938775510204082e-05,
|
| 700 |
+
"loss": 0.1217,
|
| 701 |
+
"step": 900
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.31,
|
| 705 |
+
"eval_accuracy": 0.979775828460039,
|
| 706 |
+
"eval_loss": 0.07414576411247253,
|
| 707 |
+
"eval_runtime": 296.0615,
|
| 708 |
+
"eval_samples_per_second": 13.862,
|
| 709 |
+
"eval_steps_per_second": 1.733,
|
| 710 |
+
"step": 900
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.33,
|
| 714 |
+
"learning_rate": 6.793002915451895e-05,
|
| 715 |
+
"loss": 0.0514,
|
| 716 |
+
"step": 910
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.34,
|
| 720 |
+
"learning_rate": 6.647230320699709e-05,
|
| 721 |
+
"loss": 0.0903,
|
| 722 |
+
"step": 920
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.36,
|
| 726 |
+
"learning_rate": 6.501457725947522e-05,
|
| 727 |
+
"loss": 0.0495,
|
| 728 |
+
"step": 930
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.37,
|
| 732 |
+
"learning_rate": 6.355685131195336e-05,
|
| 733 |
+
"loss": 0.0214,
|
| 734 |
+
"step": 940
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.38,
|
| 738 |
+
"learning_rate": 6.209912536443149e-05,
|
| 739 |
+
"loss": 0.081,
|
| 740 |
+
"step": 950
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.38,
|
| 744 |
+
"eval_accuracy": 0.9839181286549707,
|
| 745 |
+
"eval_loss": 0.05142759159207344,
|
| 746 |
+
"eval_runtime": 325.6943,
|
| 747 |
+
"eval_samples_per_second": 12.601,
|
| 748 |
+
"eval_steps_per_second": 1.575,
|
| 749 |
+
"step": 950
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.4,
|
| 753 |
+
"learning_rate": 6.0641399416909626e-05,
|
| 754 |
+
"loss": 0.1152,
|
| 755 |
+
"step": 960
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.41,
|
| 759 |
+
"learning_rate": 5.918367346938776e-05,
|
| 760 |
+
"loss": 0.0445,
|
| 761 |
+
"step": 970
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.43,
|
| 765 |
+
"learning_rate": 5.77259475218659e-05,
|
| 766 |
+
"loss": 0.0555,
|
| 767 |
+
"step": 980
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.44,
|
| 771 |
+
"learning_rate": 5.626822157434403e-05,
|
| 772 |
+
"loss": 0.0717,
|
| 773 |
+
"step": 990
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.46,
|
| 777 |
+
"learning_rate": 5.4810495626822155e-05,
|
| 778 |
+
"loss": 0.0607,
|
| 779 |
+
"step": 1000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.46,
|
| 783 |
+
"eval_accuracy": 0.9846491228070176,
|
| 784 |
+
"eval_loss": 0.040749311447143555,
|
| 785 |
+
"eval_runtime": 293.4556,
|
| 786 |
+
"eval_samples_per_second": 13.985,
|
| 787 |
+
"eval_steps_per_second": 1.748,
|
| 788 |
+
"step": 1000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.47,
|
| 792 |
+
"learning_rate": 5.3352769679300295e-05,
|
| 793 |
+
"loss": 0.0147,
|
| 794 |
+
"step": 1010
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.49,
|
| 798 |
+
"learning_rate": 5.189504373177842e-05,
|
| 799 |
+
"loss": 0.042,
|
| 800 |
+
"step": 1020
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.5,
|
| 804 |
+
"learning_rate": 5.043731778425656e-05,
|
| 805 |
+
"loss": 0.0513,
|
| 806 |
+
"step": 1030
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.52,
|
| 810 |
+
"learning_rate": 4.89795918367347e-05,
|
| 811 |
+
"loss": 0.0177,
|
| 812 |
+
"step": 1040
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.53,
|
| 816 |
+
"learning_rate": 4.752186588921283e-05,
|
| 817 |
+
"loss": 0.0985,
|
| 818 |
+
"step": 1050
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.53,
|
| 822 |
+
"eval_accuracy": 0.9890350877192983,
|
| 823 |
+
"eval_loss": 0.0381675623357296,
|
| 824 |
+
"eval_runtime": 285.8353,
|
| 825 |
+
"eval_samples_per_second": 14.358,
|
| 826 |
+
"eval_steps_per_second": 1.795,
|
| 827 |
+
"step": 1050
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 1.55,
|
| 831 |
+
"learning_rate": 4.6064139941690965e-05,
|
| 832 |
+
"loss": 0.0308,
|
| 833 |
+
"step": 1060
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 1.56,
|
| 837 |
+
"learning_rate": 4.46064139941691e-05,
|
| 838 |
+
"loss": 0.0297,
|
| 839 |
+
"step": 1070
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.57,
|
| 843 |
+
"learning_rate": 4.314868804664723e-05,
|
| 844 |
+
"loss": 0.1126,
|
| 845 |
+
"step": 1080
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.59,
|
| 849 |
+
"learning_rate": 4.1690962099125366e-05,
|
| 850 |
+
"loss": 0.0457,
|
| 851 |
+
"step": 1090
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.6,
|
| 855 |
+
"learning_rate": 4.02332361516035e-05,
|
| 856 |
+
"loss": 0.0492,
|
| 857 |
+
"step": 1100
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.6,
|
| 861 |
+
"eval_accuracy": 0.9883040935672515,
|
| 862 |
+
"eval_loss": 0.03809972107410431,
|
| 863 |
+
"eval_runtime": 315.6444,
|
| 864 |
+
"eval_samples_per_second": 13.002,
|
| 865 |
+
"eval_steps_per_second": 1.625,
|
| 866 |
+
"step": 1100
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.62,
|
| 870 |
+
"learning_rate": 3.8775510204081634e-05,
|
| 871 |
+
"loss": 0.0364,
|
| 872 |
+
"step": 1110
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.63,
|
| 876 |
+
"learning_rate": 3.731778425655977e-05,
|
| 877 |
+
"loss": 0.0521,
|
| 878 |
+
"step": 1120
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.65,
|
| 882 |
+
"learning_rate": 3.58600583090379e-05,
|
| 883 |
+
"loss": 0.0393,
|
| 884 |
+
"step": 1130
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.66,
|
| 888 |
+
"learning_rate": 3.4402332361516035e-05,
|
| 889 |
+
"loss": 0.0541,
|
| 890 |
+
"step": 1140
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.68,
|
| 894 |
+
"learning_rate": 3.294460641399417e-05,
|
| 895 |
+
"loss": 0.0504,
|
| 896 |
+
"step": 1150
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.68,
|
| 900 |
+
"eval_accuracy": 0.9880604288499025,
|
| 901 |
+
"eval_loss": 0.03482714295387268,
|
| 902 |
+
"eval_runtime": 330.0497,
|
| 903 |
+
"eval_samples_per_second": 12.434,
|
| 904 |
+
"eval_steps_per_second": 1.554,
|
| 905 |
+
"step": 1150
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.69,
|
| 909 |
+
"learning_rate": 3.148688046647231e-05,
|
| 910 |
+
"loss": 0.0493,
|
| 911 |
+
"step": 1160
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.71,
|
| 915 |
+
"learning_rate": 3.0029154518950437e-05,
|
| 916 |
+
"loss": 0.034,
|
| 917 |
+
"step": 1170
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1.72,
|
| 921 |
+
"learning_rate": 2.857142857142857e-05,
|
| 922 |
+
"loss": 0.0598,
|
| 923 |
+
"step": 1180
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.73,
|
| 927 |
+
"learning_rate": 2.7113702623906705e-05,
|
| 928 |
+
"loss": 0.0415,
|
| 929 |
+
"step": 1190
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.75,
|
| 933 |
+
"learning_rate": 2.5655976676384842e-05,
|
| 934 |
+
"loss": 0.0354,
|
| 935 |
+
"step": 1200
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.75,
|
| 939 |
+
"eval_accuracy": 0.9902534113060428,
|
| 940 |
+
"eval_loss": 0.025796251371502876,
|
| 941 |
+
"eval_runtime": 316.0263,
|
| 942 |
+
"eval_samples_per_second": 12.986,
|
| 943 |
+
"eval_steps_per_second": 1.623,
|
| 944 |
+
"step": 1200
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.76,
|
| 948 |
+
"learning_rate": 2.4198250728862976e-05,
|
| 949 |
+
"loss": 0.027,
|
| 950 |
+
"step": 1210
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.78,
|
| 954 |
+
"learning_rate": 2.2740524781341106e-05,
|
| 955 |
+
"loss": 0.031,
|
| 956 |
+
"step": 1220
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.79,
|
| 960 |
+
"learning_rate": 2.1282798833819244e-05,
|
| 961 |
+
"loss": 0.0448,
|
| 962 |
+
"step": 1230
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.81,
|
| 966 |
+
"learning_rate": 1.9825072886297377e-05,
|
| 967 |
+
"loss": 0.0077,
|
| 968 |
+
"step": 1240
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.82,
|
| 972 |
+
"learning_rate": 1.836734693877551e-05,
|
| 973 |
+
"loss": 0.0604,
|
| 974 |
+
"step": 1250
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.82,
|
| 978 |
+
"eval_accuracy": 0.990009746588694,
|
| 979 |
+
"eval_loss": 0.029700743034482002,
|
| 980 |
+
"eval_runtime": 261.4803,
|
| 981 |
+
"eval_samples_per_second": 15.695,
|
| 982 |
+
"eval_steps_per_second": 1.962,
|
| 983 |
+
"step": 1250
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.84,
|
| 987 |
+
"learning_rate": 1.6909620991253645e-05,
|
| 988 |
+
"loss": 0.0278,
|
| 989 |
+
"step": 1260
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.85,
|
| 993 |
+
"learning_rate": 1.545189504373178e-05,
|
| 994 |
+
"loss": 0.0219,
|
| 995 |
+
"step": 1270
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 1.87,
|
| 999 |
+
"learning_rate": 1.3994169096209913e-05,
|
| 1000 |
+
"loss": 0.0366,
|
| 1001 |
+
"step": 1280
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.88,
|
| 1005 |
+
"learning_rate": 1.2536443148688048e-05,
|
| 1006 |
+
"loss": 0.0307,
|
| 1007 |
+
"step": 1290
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.9,
|
| 1011 |
+
"learning_rate": 1.1078717201166182e-05,
|
| 1012 |
+
"loss": 0.0918,
|
| 1013 |
+
"step": 1300
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 1.9,
|
| 1017 |
+
"eval_accuracy": 0.990009746588694,
|
| 1018 |
+
"eval_loss": 0.026875579729676247,
|
| 1019 |
+
"eval_runtime": 245.5691,
|
| 1020 |
+
"eval_samples_per_second": 16.712,
|
| 1021 |
+
"eval_steps_per_second": 2.089,
|
| 1022 |
+
"step": 1300
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.91,
|
| 1026 |
+
"learning_rate": 9.620991253644314e-06,
|
| 1027 |
+
"loss": 0.0259,
|
| 1028 |
+
"step": 1310
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.92,
|
| 1032 |
+
"learning_rate": 8.163265306122448e-06,
|
| 1033 |
+
"loss": 0.0613,
|
| 1034 |
+
"step": 1320
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.94,
|
| 1038 |
+
"learning_rate": 6.705539358600584e-06,
|
| 1039 |
+
"loss": 0.0326,
|
| 1040 |
+
"step": 1330
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.95,
|
| 1044 |
+
"learning_rate": 5.247813411078718e-06,
|
| 1045 |
+
"loss": 0.0315,
|
| 1046 |
+
"step": 1340
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.97,
|
| 1050 |
+
"learning_rate": 3.7900874635568516e-06,
|
| 1051 |
+
"loss": 0.0555,
|
| 1052 |
+
"step": 1350
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.97,
|
| 1056 |
+
"eval_accuracy": 0.9926900584795322,
|
| 1057 |
+
"eval_loss": 0.020489266142249107,
|
| 1058 |
+
"eval_runtime": 250.5348,
|
| 1059 |
+
"eval_samples_per_second": 16.381,
|
| 1060 |
+
"eval_steps_per_second": 2.048,
|
| 1061 |
+
"step": 1350
|
| 1062 |
+
}
|
| 1063 |
+
],
|
| 1064 |
+
"max_steps": 1372,
|
| 1065 |
+
"num_train_epochs": 2,
|
| 1066 |
+
"total_flos": 2.154507190843392e+17,
|
| 1067 |
+
"trial_name": null,
|
| 1068 |
+
"trial_params": null
|
| 1069 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fa42306194177cd8eb6086fba4da1da9ef41852781637e070056b98d6ed0c74
|
| 3 |
+
size 3055
|