Upload 8 files
Browse files- config.json +63 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +17 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1367 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "apple/mobilevit-small",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MobileViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"aspp_dropout_prob": 0.1,
|
| 7 |
+
"aspp_out_channels": 256,
|
| 8 |
+
"atrous_rates": [
|
| 9 |
+
6,
|
| 10 |
+
12,
|
| 11 |
+
18
|
| 12 |
+
],
|
| 13 |
+
"attention_probs_dropout_prob": 0.0,
|
| 14 |
+
"classifier_dropout_prob": 0.1,
|
| 15 |
+
"conv_kernel_size": 3,
|
| 16 |
+
"expand_ratio": 4.0,
|
| 17 |
+
"hidden_act": "silu",
|
| 18 |
+
"hidden_dropout_prob": 0.1,
|
| 19 |
+
"hidden_sizes": [
|
| 20 |
+
144,
|
| 21 |
+
192,
|
| 22 |
+
240
|
| 23 |
+
],
|
| 24 |
+
"id2label": {
|
| 25 |
+
"0": "Calculus",
|
| 26 |
+
"1": "Data caries",
|
| 27 |
+
"2": "Gingivitis",
|
| 28 |
+
"3": "Mouth Ulcer",
|
| 29 |
+
"4": "Tooth Discoloration",
|
| 30 |
+
"5": "hypodontia"
|
| 31 |
+
},
|
| 32 |
+
"image_size": 256,
|
| 33 |
+
"initializer_range": 0.02,
|
| 34 |
+
"label2id": {
|
| 35 |
+
"Calculus": "0",
|
| 36 |
+
"Data caries": "1",
|
| 37 |
+
"Gingivitis": "2",
|
| 38 |
+
"Mouth Ulcer": "3",
|
| 39 |
+
"Tooth Discoloration": "4",
|
| 40 |
+
"hypodontia": "5"
|
| 41 |
+
},
|
| 42 |
+
"layer_norm_eps": 1e-05,
|
| 43 |
+
"mlp_ratio": 2.0,
|
| 44 |
+
"model_type": "mobilevit",
|
| 45 |
+
"neck_hidden_sizes": [
|
| 46 |
+
16,
|
| 47 |
+
32,
|
| 48 |
+
64,
|
| 49 |
+
96,
|
| 50 |
+
128,
|
| 51 |
+
160,
|
| 52 |
+
640
|
| 53 |
+
],
|
| 54 |
+
"num_attention_heads": 4,
|
| 55 |
+
"num_channels": 3,
|
| 56 |
+
"output_stride": 32,
|
| 57 |
+
"patch_size": 2,
|
| 58 |
+
"problem_type": "single_label_classification",
|
| 59 |
+
"qkv_bias": true,
|
| 60 |
+
"semantic_loss_ignore_index": 255,
|
| 61 |
+
"torch_dtype": "float32",
|
| 62 |
+
"transformers_version": "4.48.2"
|
| 63 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e31ddb299fa4b036f4ec1036d7af3f32ea5dc650c384542ca044b31e72a078a0
|
| 3 |
+
size 19861824
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3d7ce6ffabeaf431248310b31563026f2764e38204cdbd776f4fb56b6777a74
|
| 3 |
+
size 39741050
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_size": {
|
| 3 |
+
"height": 256,
|
| 4 |
+
"width": 256
|
| 5 |
+
},
|
| 6 |
+
"do_center_crop": true,
|
| 7 |
+
"do_flip_channel_order": true,
|
| 8 |
+
"do_flip_channels": true,
|
| 9 |
+
"do_rescale": true,
|
| 10 |
+
"do_resize": true,
|
| 11 |
+
"image_processor_type": "MobileViTImageProcessor",
|
| 12 |
+
"resample": 2,
|
| 13 |
+
"rescale_factor": 0.00392156862745098,
|
| 14 |
+
"size": {
|
| 15 |
+
"shortest_edge": 288
|
| 16 |
+
}
|
| 17 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:67795c0dd68c52f18e857ca3d51ed77570af6d5b45da584143f8eafd0b2d09a7
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad0ac53b7f50cef1fd8105d195a79c4bbc05fca430948aa3c89658e98d0b705a
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1367 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.24052156507968903,
|
| 3 |
+
"best_model_checkpoint": "./oral_disease_mobilevit/checkpoint-4600",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 5830,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04288164665523156,
|
| 13 |
+
"grad_norm": 2.5394277572631836,
|
| 14 |
+
"learning_rate": 4.957118353344769e-05,
|
| 15 |
+
"loss": 1.7703,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08576329331046312,
|
| 20 |
+
"grad_norm": 2.5862715244293213,
|
| 21 |
+
"learning_rate": 4.914236706689537e-05,
|
| 22 |
+
"loss": 1.6871,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08576329331046312,
|
| 27 |
+
"eval_accuracy": 0.577005577005577,
|
| 28 |
+
"eval_loss": 1.6024609804153442,
|
| 29 |
+
"eval_runtime": 14.1663,
|
| 30 |
+
"eval_samples_per_second": 164.545,
|
| 31 |
+
"eval_steps_per_second": 20.612,
|
| 32 |
+
"step": 100
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.12864493996569468,
|
| 36 |
+
"grad_norm": 3.443279504776001,
|
| 37 |
+
"learning_rate": 4.8713550600343055e-05,
|
| 38 |
+
"loss": 1.5656,
|
| 39 |
+
"step": 150
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.17152658662092624,
|
| 43 |
+
"grad_norm": 2.7117512226104736,
|
| 44 |
+
"learning_rate": 4.828473413379074e-05,
|
| 45 |
+
"loss": 1.4961,
|
| 46 |
+
"step": 200
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.17152658662092624,
|
| 50 |
+
"eval_accuracy": 0.5971685971685972,
|
| 51 |
+
"eval_loss": 1.3290797472000122,
|
| 52 |
+
"eval_runtime": 14.4239,
|
| 53 |
+
"eval_samples_per_second": 161.607,
|
| 54 |
+
"eval_steps_per_second": 20.244,
|
| 55 |
+
"step": 200
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.2144082332761578,
|
| 59 |
+
"grad_norm": 3.41520619392395,
|
| 60 |
+
"learning_rate": 4.7855917667238424e-05,
|
| 61 |
+
"loss": 1.3941,
|
| 62 |
+
"step": 250
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.25728987993138935,
|
| 66 |
+
"grad_norm": 2.960935354232788,
|
| 67 |
+
"learning_rate": 4.742710120068611e-05,
|
| 68 |
+
"loss": 1.2984,
|
| 69 |
+
"step": 300
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.25728987993138935,
|
| 73 |
+
"eval_accuracy": 0.6525096525096525,
|
| 74 |
+
"eval_loss": 1.1308239698410034,
|
| 75 |
+
"eval_runtime": 14.2271,
|
| 76 |
+
"eval_samples_per_second": 163.842,
|
| 77 |
+
"eval_steps_per_second": 20.524,
|
| 78 |
+
"step": 300
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.30017152658662094,
|
| 82 |
+
"grad_norm": 3.7776384353637695,
|
| 83 |
+
"learning_rate": 4.699828473413379e-05,
|
| 84 |
+
"loss": 1.2457,
|
| 85 |
+
"step": 350
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.34305317324185247,
|
| 89 |
+
"grad_norm": 6.4286088943481445,
|
| 90 |
+
"learning_rate": 4.656946826758148e-05,
|
| 91 |
+
"loss": 1.12,
|
| 92 |
+
"step": 400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.34305317324185247,
|
| 96 |
+
"eval_accuracy": 0.7232947232947233,
|
| 97 |
+
"eval_loss": 0.9413465261459351,
|
| 98 |
+
"eval_runtime": 14.2093,
|
| 99 |
+
"eval_samples_per_second": 164.048,
|
| 100 |
+
"eval_steps_per_second": 20.55,
|
| 101 |
+
"step": 400
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.38593481989708406,
|
| 105 |
+
"grad_norm": 3.400545120239258,
|
| 106 |
+
"learning_rate": 4.614065180102916e-05,
|
| 107 |
+
"loss": 1.0816,
|
| 108 |
+
"step": 450
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.4288164665523156,
|
| 112 |
+
"grad_norm": 4.862150192260742,
|
| 113 |
+
"learning_rate": 4.5711835334476845e-05,
|
| 114 |
+
"loss": 1.0873,
|
| 115 |
+
"step": 500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.4288164665523156,
|
| 119 |
+
"eval_accuracy": 0.7610467610467611,
|
| 120 |
+
"eval_loss": 0.8114073872566223,
|
| 121 |
+
"eval_runtime": 14.2607,
|
| 122 |
+
"eval_samples_per_second": 163.456,
|
| 123 |
+
"eval_steps_per_second": 20.476,
|
| 124 |
+
"step": 500
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.4716981132075472,
|
| 128 |
+
"grad_norm": 3.4492220878601074,
|
| 129 |
+
"learning_rate": 4.528301886792453e-05,
|
| 130 |
+
"loss": 0.9814,
|
| 131 |
+
"step": 550
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.5145797598627787,
|
| 135 |
+
"grad_norm": 3.7629287242889404,
|
| 136 |
+
"learning_rate": 4.4854202401372214e-05,
|
| 137 |
+
"loss": 1.0012,
|
| 138 |
+
"step": 600
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.5145797598627787,
|
| 142 |
+
"eval_accuracy": 0.7919347919347919,
|
| 143 |
+
"eval_loss": 0.6951708197593689,
|
| 144 |
+
"eval_runtime": 13.8986,
|
| 145 |
+
"eval_samples_per_second": 167.714,
|
| 146 |
+
"eval_steps_per_second": 21.009,
|
| 147 |
+
"step": 600
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.5574614065180102,
|
| 151 |
+
"grad_norm": 4.381726264953613,
|
| 152 |
+
"learning_rate": 4.4425385934819905e-05,
|
| 153 |
+
"loss": 0.9516,
|
| 154 |
+
"step": 650
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.6003430531732419,
|
| 158 |
+
"grad_norm": 3.579646348953247,
|
| 159 |
+
"learning_rate": 4.399656946826758e-05,
|
| 160 |
+
"loss": 0.8379,
|
| 161 |
+
"step": 700
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.6003430531732419,
|
| 165 |
+
"eval_accuracy": 0.8125268125268126,
|
| 166 |
+
"eval_loss": 0.6312650442123413,
|
| 167 |
+
"eval_runtime": 13.8871,
|
| 168 |
+
"eval_samples_per_second": 167.853,
|
| 169 |
+
"eval_steps_per_second": 21.027,
|
| 170 |
+
"step": 700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6432246998284734,
|
| 174 |
+
"grad_norm": 5.985883712768555,
|
| 175 |
+
"learning_rate": 4.356775300171527e-05,
|
| 176 |
+
"loss": 0.8269,
|
| 177 |
+
"step": 750
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6861063464837049,
|
| 181 |
+
"grad_norm": 5.968649864196777,
|
| 182 |
+
"learning_rate": 4.313893653516296e-05,
|
| 183 |
+
"loss": 0.8627,
|
| 184 |
+
"step": 800
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6861063464837049,
|
| 188 |
+
"eval_accuracy": 0.8181038181038182,
|
| 189 |
+
"eval_loss": 0.5455917716026306,
|
| 190 |
+
"eval_runtime": 13.9813,
|
| 191 |
+
"eval_samples_per_second": 166.723,
|
| 192 |
+
"eval_steps_per_second": 20.885,
|
| 193 |
+
"step": 800
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.7289879931389366,
|
| 197 |
+
"grad_norm": 5.903740882873535,
|
| 198 |
+
"learning_rate": 4.2710120068610635e-05,
|
| 199 |
+
"loss": 0.8055,
|
| 200 |
+
"step": 850
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.7718696397941681,
|
| 204 |
+
"grad_norm": 6.328656196594238,
|
| 205 |
+
"learning_rate": 4.228130360205832e-05,
|
| 206 |
+
"loss": 0.8252,
|
| 207 |
+
"step": 900
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.7718696397941681,
|
| 211 |
+
"eval_accuracy": 0.8314028314028314,
|
| 212 |
+
"eval_loss": 0.5072463750839233,
|
| 213 |
+
"eval_runtime": 13.8768,
|
| 214 |
+
"eval_samples_per_second": 167.979,
|
| 215 |
+
"eval_steps_per_second": 21.042,
|
| 216 |
+
"step": 900
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.8147512864493996,
|
| 220 |
+
"grad_norm": 4.45211124420166,
|
| 221 |
+
"learning_rate": 4.185248713550601e-05,
|
| 222 |
+
"loss": 0.7173,
|
| 223 |
+
"step": 950
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.8576329331046312,
|
| 227 |
+
"grad_norm": 5.749237537384033,
|
| 228 |
+
"learning_rate": 4.142367066895369e-05,
|
| 229 |
+
"loss": 0.7223,
|
| 230 |
+
"step": 1000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.8576329331046312,
|
| 234 |
+
"eval_accuracy": 0.8378378378378378,
|
| 235 |
+
"eval_loss": 0.46284219622612,
|
| 236 |
+
"eval_runtime": 14.2987,
|
| 237 |
+
"eval_samples_per_second": 163.022,
|
| 238 |
+
"eval_steps_per_second": 20.421,
|
| 239 |
+
"step": 1000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.9005145797598628,
|
| 243 |
+
"grad_norm": 16.25695037841797,
|
| 244 |
+
"learning_rate": 4.099485420240137e-05,
|
| 245 |
+
"loss": 0.7153,
|
| 246 |
+
"step": 1050
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.9433962264150944,
|
| 250 |
+
"grad_norm": 7.836055755615234,
|
| 251 |
+
"learning_rate": 4.0566037735849064e-05,
|
| 252 |
+
"loss": 0.6617,
|
| 253 |
+
"step": 1100
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.9433962264150944,
|
| 257 |
+
"eval_accuracy": 0.8468468468468469,
|
| 258 |
+
"eval_loss": 0.43263953924179077,
|
| 259 |
+
"eval_runtime": 13.9551,
|
| 260 |
+
"eval_samples_per_second": 167.035,
|
| 261 |
+
"eval_steps_per_second": 20.924,
|
| 262 |
+
"step": 1100
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.9862778730703259,
|
| 266 |
+
"grad_norm": 8.199559211730957,
|
| 267 |
+
"learning_rate": 4.013722126929674e-05,
|
| 268 |
+
"loss": 0.7156,
|
| 269 |
+
"step": 1150
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.0291595197255574,
|
| 273 |
+
"grad_norm": 8.776689529418945,
|
| 274 |
+
"learning_rate": 3.9708404802744425e-05,
|
| 275 |
+
"loss": 0.7784,
|
| 276 |
+
"step": 1200
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.0291595197255574,
|
| 280 |
+
"eval_accuracy": 0.8511368511368511,
|
| 281 |
+
"eval_loss": 0.42261621356010437,
|
| 282 |
+
"eval_runtime": 14.1288,
|
| 283 |
+
"eval_samples_per_second": 164.982,
|
| 284 |
+
"eval_steps_per_second": 20.667,
|
| 285 |
+
"step": 1200
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.072041166380789,
|
| 289 |
+
"grad_norm": 5.693437099456787,
|
| 290 |
+
"learning_rate": 3.9279588336192116e-05,
|
| 291 |
+
"loss": 0.6163,
|
| 292 |
+
"step": 1250
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.1149228130360207,
|
| 296 |
+
"grad_norm": 16.413869857788086,
|
| 297 |
+
"learning_rate": 3.8850771869639794e-05,
|
| 298 |
+
"loss": 0.6086,
|
| 299 |
+
"step": 1300
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.1149228130360207,
|
| 303 |
+
"eval_accuracy": 0.8592878592878593,
|
| 304 |
+
"eval_loss": 0.3927444517612457,
|
| 305 |
+
"eval_runtime": 13.8914,
|
| 306 |
+
"eval_samples_per_second": 167.801,
|
| 307 |
+
"eval_steps_per_second": 21.02,
|
| 308 |
+
"step": 1300
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.1578044596912522,
|
| 312 |
+
"grad_norm": 4.260317325592041,
|
| 313 |
+
"learning_rate": 3.842195540308748e-05,
|
| 314 |
+
"loss": 0.6237,
|
| 315 |
+
"step": 1350
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.2006861063464838,
|
| 319 |
+
"grad_norm": 8.143071174621582,
|
| 320 |
+
"learning_rate": 3.799313893653517e-05,
|
| 321 |
+
"loss": 0.6196,
|
| 322 |
+
"step": 1400
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.2006861063464838,
|
| 326 |
+
"eval_accuracy": 0.8618618618618619,
|
| 327 |
+
"eval_loss": 0.3923815190792084,
|
| 328 |
+
"eval_runtime": 13.8559,
|
| 329 |
+
"eval_samples_per_second": 168.231,
|
| 330 |
+
"eval_steps_per_second": 21.074,
|
| 331 |
+
"step": 1400
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.2435677530017153,
|
| 335 |
+
"grad_norm": 8.181282043457031,
|
| 336 |
+
"learning_rate": 3.756432246998285e-05,
|
| 337 |
+
"loss": 0.5606,
|
| 338 |
+
"step": 1450
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.2864493996569468,
|
| 342 |
+
"grad_norm": 14.405867576599121,
|
| 343 |
+
"learning_rate": 3.713550600343053e-05,
|
| 344 |
+
"loss": 0.6561,
|
| 345 |
+
"step": 1500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.2864493996569468,
|
| 349 |
+
"eval_accuracy": 0.8601458601458601,
|
| 350 |
+
"eval_loss": 0.36444056034088135,
|
| 351 |
+
"eval_runtime": 13.8772,
|
| 352 |
+
"eval_samples_per_second": 167.974,
|
| 353 |
+
"eval_steps_per_second": 21.042,
|
| 354 |
+
"step": 1500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.3293310463121784,
|
| 358 |
+
"grad_norm": 14.855607986450195,
|
| 359 |
+
"learning_rate": 3.670668953687822e-05,
|
| 360 |
+
"loss": 0.5548,
|
| 361 |
+
"step": 1550
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.3722126929674099,
|
| 365 |
+
"grad_norm": 4.9101409912109375,
|
| 366 |
+
"learning_rate": 3.62778730703259e-05,
|
| 367 |
+
"loss": 0.5426,
|
| 368 |
+
"step": 1600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.3722126929674099,
|
| 372 |
+
"eval_accuracy": 0.8738738738738738,
|
| 373 |
+
"eval_loss": 0.3387666344642639,
|
| 374 |
+
"eval_runtime": 14.6787,
|
| 375 |
+
"eval_samples_per_second": 158.801,
|
| 376 |
+
"eval_steps_per_second": 19.893,
|
| 377 |
+
"step": 1600
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.4150943396226414,
|
| 381 |
+
"grad_norm": 3.1763756275177,
|
| 382 |
+
"learning_rate": 3.5849056603773584e-05,
|
| 383 |
+
"loss": 0.5917,
|
| 384 |
+
"step": 1650
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.457975986277873,
|
| 388 |
+
"grad_norm": 2.9457991123199463,
|
| 389 |
+
"learning_rate": 3.5420240137221275e-05,
|
| 390 |
+
"loss": 0.5858,
|
| 391 |
+
"step": 1700
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.457975986277873,
|
| 395 |
+
"eval_accuracy": 0.8652938652938653,
|
| 396 |
+
"eval_loss": 0.3525221347808838,
|
| 397 |
+
"eval_runtime": 14.1719,
|
| 398 |
+
"eval_samples_per_second": 164.48,
|
| 399 |
+
"eval_steps_per_second": 20.604,
|
| 400 |
+
"step": 1700
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.5008576329331045,
|
| 404 |
+
"grad_norm": 7.572704792022705,
|
| 405 |
+
"learning_rate": 3.499142367066895e-05,
|
| 406 |
+
"loss": 0.6098,
|
| 407 |
+
"step": 1750
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.5437392795883362,
|
| 411 |
+
"grad_norm": 1.232725739479065,
|
| 412 |
+
"learning_rate": 3.456260720411664e-05,
|
| 413 |
+
"loss": 0.574,
|
| 414 |
+
"step": 1800
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.5437392795883362,
|
| 418 |
+
"eval_accuracy": 0.8764478764478765,
|
| 419 |
+
"eval_loss": 0.32060739398002625,
|
| 420 |
+
"eval_runtime": 14.3858,
|
| 421 |
+
"eval_samples_per_second": 162.035,
|
| 422 |
+
"eval_steps_per_second": 20.298,
|
| 423 |
+
"step": 1800
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.5866209262435678,
|
| 427 |
+
"grad_norm": 10.918987274169922,
|
| 428 |
+
"learning_rate": 3.413379073756433e-05,
|
| 429 |
+
"loss": 0.5824,
|
| 430 |
+
"step": 1850
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.6295025728987993,
|
| 434 |
+
"grad_norm": 10.205852508544922,
|
| 435 |
+
"learning_rate": 3.3704974271012005e-05,
|
| 436 |
+
"loss": 0.5015,
|
| 437 |
+
"step": 1900
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.6295025728987993,
|
| 441 |
+
"eval_accuracy": 0.882024882024882,
|
| 442 |
+
"eval_loss": 0.32507938146591187,
|
| 443 |
+
"eval_runtime": 14.2196,
|
| 444 |
+
"eval_samples_per_second": 163.929,
|
| 445 |
+
"eval_steps_per_second": 20.535,
|
| 446 |
+
"step": 1900
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.6723842195540308,
|
| 450 |
+
"grad_norm": 7.081895351409912,
|
| 451 |
+
"learning_rate": 3.3276157804459696e-05,
|
| 452 |
+
"loss": 0.4681,
|
| 453 |
+
"step": 1950
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.7152658662092626,
|
| 457 |
+
"grad_norm": 9.101844787597656,
|
| 458 |
+
"learning_rate": 3.284734133790738e-05,
|
| 459 |
+
"loss": 0.493,
|
| 460 |
+
"step": 2000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.7152658662092626,
|
| 464 |
+
"eval_accuracy": 0.8854568854568855,
|
| 465 |
+
"eval_loss": 0.308988094329834,
|
| 466 |
+
"eval_runtime": 14.3569,
|
| 467 |
+
"eval_samples_per_second": 162.36,
|
| 468 |
+
"eval_steps_per_second": 20.339,
|
| 469 |
+
"step": 2000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.758147512864494,
|
| 473 |
+
"grad_norm": 6.105254650115967,
|
| 474 |
+
"learning_rate": 3.241852487135506e-05,
|
| 475 |
+
"loss": 0.533,
|
| 476 |
+
"step": 2050
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.8010291595197256,
|
| 480 |
+
"grad_norm": 14.518338203430176,
|
| 481 |
+
"learning_rate": 3.198970840480275e-05,
|
| 482 |
+
"loss": 0.5381,
|
| 483 |
+
"step": 2100
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.8010291595197256,
|
| 487 |
+
"eval_accuracy": 0.8773058773058773,
|
| 488 |
+
"eval_loss": 0.3106726109981537,
|
| 489 |
+
"eval_runtime": 14.2106,
|
| 490 |
+
"eval_samples_per_second": 164.033,
|
| 491 |
+
"eval_steps_per_second": 20.548,
|
| 492 |
+
"step": 2100
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.8439108061749572,
|
| 496 |
+
"grad_norm": 3.1229450702667236,
|
| 497 |
+
"learning_rate": 3.1560891938250434e-05,
|
| 498 |
+
"loss": 0.5641,
|
| 499 |
+
"step": 2150
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.8867924528301887,
|
| 503 |
+
"grad_norm": 14.663033485412598,
|
| 504 |
+
"learning_rate": 3.113207547169811e-05,
|
| 505 |
+
"loss": 0.4979,
|
| 506 |
+
"step": 2200
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.8867924528301887,
|
| 510 |
+
"eval_accuracy": 0.8901758901758902,
|
| 511 |
+
"eval_loss": 0.2851434051990509,
|
| 512 |
+
"eval_runtime": 14.2521,
|
| 513 |
+
"eval_samples_per_second": 163.554,
|
| 514 |
+
"eval_steps_per_second": 20.488,
|
| 515 |
+
"step": 2200
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.9296740994854202,
|
| 519 |
+
"grad_norm": 3.774958848953247,
|
| 520 |
+
"learning_rate": 3.07032590051458e-05,
|
| 521 |
+
"loss": 0.5117,
|
| 522 |
+
"step": 2250
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.9725557461406518,
|
| 526 |
+
"grad_norm": 25.906217575073242,
|
| 527 |
+
"learning_rate": 3.0274442538593483e-05,
|
| 528 |
+
"loss": 0.5145,
|
| 529 |
+
"step": 2300
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.9725557461406518,
|
| 533 |
+
"eval_accuracy": 0.8927498927498928,
|
| 534 |
+
"eval_loss": 0.2909970283508301,
|
| 535 |
+
"eval_runtime": 14.2775,
|
| 536 |
+
"eval_samples_per_second": 163.264,
|
| 537 |
+
"eval_steps_per_second": 20.452,
|
| 538 |
+
"step": 2300
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.0154373927958833,
|
| 542 |
+
"grad_norm": 8.114184379577637,
|
| 543 |
+
"learning_rate": 2.9845626072041167e-05,
|
| 544 |
+
"loss": 0.4499,
|
| 545 |
+
"step": 2350
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.058319039451115,
|
| 549 |
+
"grad_norm": 9.68442153930664,
|
| 550 |
+
"learning_rate": 2.9416809605488855e-05,
|
| 551 |
+
"loss": 0.4846,
|
| 552 |
+
"step": 2400
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.058319039451115,
|
| 556 |
+
"eval_accuracy": 0.8833118833118833,
|
| 557 |
+
"eval_loss": 0.3212469518184662,
|
| 558 |
+
"eval_runtime": 14.6038,
|
| 559 |
+
"eval_samples_per_second": 159.616,
|
| 560 |
+
"eval_steps_per_second": 19.995,
|
| 561 |
+
"step": 2400
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.1012006861063464,
|
| 565 |
+
"grad_norm": 3.7848308086395264,
|
| 566 |
+
"learning_rate": 2.8987993138936536e-05,
|
| 567 |
+
"loss": 0.4533,
|
| 568 |
+
"step": 2450
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 2.144082332761578,
|
| 572 |
+
"grad_norm": 8.865078926086426,
|
| 573 |
+
"learning_rate": 2.855917667238422e-05,
|
| 574 |
+
"loss": 0.4545,
|
| 575 |
+
"step": 2500
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 2.144082332761578,
|
| 579 |
+
"eval_accuracy": 0.8828828828828829,
|
| 580 |
+
"eval_loss": 0.3031620681285858,
|
| 581 |
+
"eval_runtime": 14.3961,
|
| 582 |
+
"eval_samples_per_second": 161.919,
|
| 583 |
+
"eval_steps_per_second": 20.283,
|
| 584 |
+
"step": 2500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.1869639794168094,
|
| 588 |
+
"grad_norm": 7.019817352294922,
|
| 589 |
+
"learning_rate": 2.8130360205831908e-05,
|
| 590 |
+
"loss": 0.4766,
|
| 591 |
+
"step": 2550
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.2298456260720414,
|
| 595 |
+
"grad_norm": 4.404627799987793,
|
| 596 |
+
"learning_rate": 2.770154373927959e-05,
|
| 597 |
+
"loss": 0.4345,
|
| 598 |
+
"step": 2600
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 2.2298456260720414,
|
| 602 |
+
"eval_accuracy": 0.8923208923208923,
|
| 603 |
+
"eval_loss": 0.2739025354385376,
|
| 604 |
+
"eval_runtime": 14.9267,
|
| 605 |
+
"eval_samples_per_second": 156.163,
|
| 606 |
+
"eval_steps_per_second": 19.562,
|
| 607 |
+
"step": 2600
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 2.2727272727272725,
|
| 611 |
+
"grad_norm": 4.765737056732178,
|
| 612 |
+
"learning_rate": 2.7272727272727273e-05,
|
| 613 |
+
"loss": 0.424,
|
| 614 |
+
"step": 2650
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.3156089193825045,
|
| 618 |
+
"grad_norm": 31.984413146972656,
|
| 619 |
+
"learning_rate": 2.684391080617496e-05,
|
| 620 |
+
"loss": 0.4697,
|
| 621 |
+
"step": 2700
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.3156089193825045,
|
| 625 |
+
"eval_accuracy": 0.8931788931788932,
|
| 626 |
+
"eval_loss": 0.2758677005767822,
|
| 627 |
+
"eval_runtime": 14.2259,
|
| 628 |
+
"eval_samples_per_second": 163.856,
|
| 629 |
+
"eval_steps_per_second": 20.526,
|
| 630 |
+
"step": 2700
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.358490566037736,
|
| 634 |
+
"grad_norm": 5.218204498291016,
|
| 635 |
+
"learning_rate": 2.641509433962264e-05,
|
| 636 |
+
"loss": 0.4927,
|
| 637 |
+
"step": 2750
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 2.4013722126929675,
|
| 641 |
+
"grad_norm": 19.66315269470215,
|
| 642 |
+
"learning_rate": 2.5986277873070326e-05,
|
| 643 |
+
"loss": 0.4313,
|
| 644 |
+
"step": 2800
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.4013722126929675,
|
| 648 |
+
"eval_accuracy": 0.9051909051909052,
|
| 649 |
+
"eval_loss": 0.25858309864997864,
|
| 650 |
+
"eval_runtime": 14.6303,
|
| 651 |
+
"eval_samples_per_second": 159.327,
|
| 652 |
+
"eval_steps_per_second": 19.959,
|
| 653 |
+
"step": 2800
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.444253859348199,
|
| 657 |
+
"grad_norm": 1.5584404468536377,
|
| 658 |
+
"learning_rate": 2.5557461406518014e-05,
|
| 659 |
+
"loss": 0.4136,
|
| 660 |
+
"step": 2850
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.4871355060034306,
|
| 664 |
+
"grad_norm": 6.387534141540527,
|
| 665 |
+
"learning_rate": 2.5128644939965695e-05,
|
| 666 |
+
"loss": 0.4215,
|
| 667 |
+
"step": 2900
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.4871355060034306,
|
| 671 |
+
"eval_accuracy": 0.9004719004719005,
|
| 672 |
+
"eval_loss": 0.2652466893196106,
|
| 673 |
+
"eval_runtime": 14.1925,
|
| 674 |
+
"eval_samples_per_second": 164.242,
|
| 675 |
+
"eval_steps_per_second": 20.574,
|
| 676 |
+
"step": 2900
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.530017152658662,
|
| 680 |
+
"grad_norm": 24.956953048706055,
|
| 681 |
+
"learning_rate": 2.4699828473413382e-05,
|
| 682 |
+
"loss": 0.4174,
|
| 683 |
+
"step": 2950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.5728987993138936,
|
| 687 |
+
"grad_norm": 4.991686820983887,
|
| 688 |
+
"learning_rate": 2.4271012006861067e-05,
|
| 689 |
+
"loss": 0.4504,
|
| 690 |
+
"step": 3000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.5728987993138936,
|
| 694 |
+
"eval_accuracy": 0.8927498927498928,
|
| 695 |
+
"eval_loss": 0.2829649746417999,
|
| 696 |
+
"eval_runtime": 13.721,
|
| 697 |
+
"eval_samples_per_second": 169.885,
|
| 698 |
+
"eval_steps_per_second": 21.281,
|
| 699 |
+
"step": 3000
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.615780445969125,
|
| 703 |
+
"grad_norm": 8.409246444702148,
|
| 704 |
+
"learning_rate": 2.3842195540308747e-05,
|
| 705 |
+
"loss": 0.433,
|
| 706 |
+
"step": 3050
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.6586620926243567,
|
| 710 |
+
"grad_norm": 20.707319259643555,
|
| 711 |
+
"learning_rate": 2.3413379073756435e-05,
|
| 712 |
+
"loss": 0.4218,
|
| 713 |
+
"step": 3100
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.6586620926243567,
|
| 717 |
+
"eval_accuracy": 0.8936078936078936,
|
| 718 |
+
"eval_loss": 0.2900983989238739,
|
| 719 |
+
"eval_runtime": 13.1413,
|
| 720 |
+
"eval_samples_per_second": 177.38,
|
| 721 |
+
"eval_steps_per_second": 22.22,
|
| 722 |
+
"step": 3100
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 2.7015437392795882,
|
| 726 |
+
"grad_norm": 6.371370792388916,
|
| 727 |
+
"learning_rate": 2.298456260720412e-05,
|
| 728 |
+
"loss": 0.454,
|
| 729 |
+
"step": 3150
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 2.7444253859348198,
|
| 733 |
+
"grad_norm": 2.8135769367218018,
|
| 734 |
+
"learning_rate": 2.25557461406518e-05,
|
| 735 |
+
"loss": 0.4889,
|
| 736 |
+
"step": 3200
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.7444253859348198,
|
| 740 |
+
"eval_accuracy": 0.8910338910338911,
|
| 741 |
+
"eval_loss": 0.2830560803413391,
|
| 742 |
+
"eval_runtime": 13.2735,
|
| 743 |
+
"eval_samples_per_second": 175.613,
|
| 744 |
+
"eval_steps_per_second": 21.999,
|
| 745 |
+
"step": 3200
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 2.7873070325900513,
|
| 749 |
+
"grad_norm": 4.3516058921813965,
|
| 750 |
+
"learning_rate": 2.2126929674099488e-05,
|
| 751 |
+
"loss": 0.5019,
|
| 752 |
+
"step": 3250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.830188679245283,
|
| 756 |
+
"grad_norm": 10.607779502868652,
|
| 757 |
+
"learning_rate": 2.1698113207547172e-05,
|
| 758 |
+
"loss": 0.47,
|
| 759 |
+
"step": 3300
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.830188679245283,
|
| 763 |
+
"eval_accuracy": 0.8927498927498928,
|
| 764 |
+
"eval_loss": 0.2812536060810089,
|
| 765 |
+
"eval_runtime": 13.5257,
|
| 766 |
+
"eval_samples_per_second": 172.338,
|
| 767 |
+
"eval_steps_per_second": 21.589,
|
| 768 |
+
"step": 3300
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.873070325900515,
|
| 772 |
+
"grad_norm": 3.962139844894409,
|
| 773 |
+
"learning_rate": 2.1269296740994853e-05,
|
| 774 |
+
"loss": 0.3699,
|
| 775 |
+
"step": 3350
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 2.915951972555746,
|
| 779 |
+
"grad_norm": 10.8749361038208,
|
| 780 |
+
"learning_rate": 2.084048027444254e-05,
|
| 781 |
+
"loss": 0.4758,
|
| 782 |
+
"step": 3400
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.915951972555746,
|
| 786 |
+
"eval_accuracy": 0.8944658944658944,
|
| 787 |
+
"eval_loss": 0.2788408696651459,
|
| 788 |
+
"eval_runtime": 13.5439,
|
| 789 |
+
"eval_samples_per_second": 172.107,
|
| 790 |
+
"eval_steps_per_second": 21.56,
|
| 791 |
+
"step": 3400
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.958833619210978,
|
| 795 |
+
"grad_norm": 20.483320236206055,
|
| 796 |
+
"learning_rate": 2.0411663807890225e-05,
|
| 797 |
+
"loss": 0.4581,
|
| 798 |
+
"step": 3450
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 3.0017152658662094,
|
| 802 |
+
"grad_norm": 12.251073837280273,
|
| 803 |
+
"learning_rate": 1.998284734133791e-05,
|
| 804 |
+
"loss": 0.4482,
|
| 805 |
+
"step": 3500
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 3.0017152658662094,
|
| 809 |
+
"eval_accuracy": 0.8953238953238953,
|
| 810 |
+
"eval_loss": 0.26482564210891724,
|
| 811 |
+
"eval_runtime": 13.5212,
|
| 812 |
+
"eval_samples_per_second": 172.396,
|
| 813 |
+
"eval_steps_per_second": 21.596,
|
| 814 |
+
"step": 3500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.044596912521441,
|
| 818 |
+
"grad_norm": 6.609012603759766,
|
| 819 |
+
"learning_rate": 1.9554030874785594e-05,
|
| 820 |
+
"loss": 0.4557,
|
| 821 |
+
"step": 3550
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.0874785591766725,
|
| 825 |
+
"grad_norm": 31.05156898498535,
|
| 826 |
+
"learning_rate": 1.9125214408233278e-05,
|
| 827 |
+
"loss": 0.3735,
|
| 828 |
+
"step": 3600
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.0874785591766725,
|
| 832 |
+
"eval_accuracy": 0.8987558987558988,
|
| 833 |
+
"eval_loss": 0.25668570399284363,
|
| 834 |
+
"eval_runtime": 13.5267,
|
| 835 |
+
"eval_samples_per_second": 172.325,
|
| 836 |
+
"eval_steps_per_second": 21.587,
|
| 837 |
+
"step": 3600
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 3.130360205831904,
|
| 841 |
+
"grad_norm": 33.095829010009766,
|
| 842 |
+
"learning_rate": 1.8696397941680962e-05,
|
| 843 |
+
"loss": 0.3992,
|
| 844 |
+
"step": 3650
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 3.1732418524871355,
|
| 848 |
+
"grad_norm": 10.586121559143066,
|
| 849 |
+
"learning_rate": 1.8267581475128647e-05,
|
| 850 |
+
"loss": 0.4338,
|
| 851 |
+
"step": 3700
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.1732418524871355,
|
| 855 |
+
"eval_accuracy": 0.9086229086229086,
|
| 856 |
+
"eval_loss": 0.2435009777545929,
|
| 857 |
+
"eval_runtime": 13.3274,
|
| 858 |
+
"eval_samples_per_second": 174.903,
|
| 859 |
+
"eval_steps_per_second": 21.91,
|
| 860 |
+
"step": 3700
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 3.216123499142367,
|
| 864 |
+
"grad_norm": 2.1484551429748535,
|
| 865 |
+
"learning_rate": 1.783876500857633e-05,
|
| 866 |
+
"loss": 0.4019,
|
| 867 |
+
"step": 3750
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 3.2590051457975986,
|
| 871 |
+
"grad_norm": 1.3401719331741333,
|
| 872 |
+
"learning_rate": 1.7409948542024015e-05,
|
| 873 |
+
"loss": 0.3423,
|
| 874 |
+
"step": 3800
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 3.2590051457975986,
|
| 878 |
+
"eval_accuracy": 0.9021879021879022,
|
| 879 |
+
"eval_loss": 0.25972625613212585,
|
| 880 |
+
"eval_runtime": 14.1124,
|
| 881 |
+
"eval_samples_per_second": 165.174,
|
| 882 |
+
"eval_steps_per_second": 20.691,
|
| 883 |
+
"step": 3800
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.30188679245283,
|
| 887 |
+
"grad_norm": 2.6398582458496094,
|
| 888 |
+
"learning_rate": 1.69811320754717e-05,
|
| 889 |
+
"loss": 0.3813,
|
| 890 |
+
"step": 3850
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.3447684391080617,
|
| 894 |
+
"grad_norm": 8.979487419128418,
|
| 895 |
+
"learning_rate": 1.6552315608919384e-05,
|
| 896 |
+
"loss": 0.3689,
|
| 897 |
+
"step": 3900
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.3447684391080617,
|
| 901 |
+
"eval_accuracy": 0.9077649077649078,
|
| 902 |
+
"eval_loss": 0.261087566614151,
|
| 903 |
+
"eval_runtime": 13.587,
|
| 904 |
+
"eval_samples_per_second": 171.561,
|
| 905 |
+
"eval_steps_per_second": 21.491,
|
| 906 |
+
"step": 3900
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 3.387650085763293,
|
| 910 |
+
"grad_norm": 4.436942100524902,
|
| 911 |
+
"learning_rate": 1.6123499142367068e-05,
|
| 912 |
+
"loss": 0.4099,
|
| 913 |
+
"step": 3950
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 3.4305317324185247,
|
| 917 |
+
"grad_norm": 4.749749660491943,
|
| 918 |
+
"learning_rate": 1.5694682675814752e-05,
|
| 919 |
+
"loss": 0.347,
|
| 920 |
+
"step": 4000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 3.4305317324185247,
|
| 924 |
+
"eval_accuracy": 0.9103389103389103,
|
| 925 |
+
"eval_loss": 0.24434834718704224,
|
| 926 |
+
"eval_runtime": 13.493,
|
| 927 |
+
"eval_samples_per_second": 172.757,
|
| 928 |
+
"eval_steps_per_second": 21.641,
|
| 929 |
+
"step": 4000
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.4734133790737562,
|
| 933 |
+
"grad_norm": 62.90760040283203,
|
| 934 |
+
"learning_rate": 1.5265866209262437e-05,
|
| 935 |
+
"loss": 0.3681,
|
| 936 |
+
"step": 4050
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.516295025728988,
|
| 940 |
+
"grad_norm": 29.556215286254883,
|
| 941 |
+
"learning_rate": 1.4837049742710121e-05,
|
| 942 |
+
"loss": 0.2854,
|
| 943 |
+
"step": 4100
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 3.516295025728988,
|
| 947 |
+
"eval_accuracy": 0.9051909051909052,
|
| 948 |
+
"eval_loss": 0.25928205251693726,
|
| 949 |
+
"eval_runtime": 13.3683,
|
| 950 |
+
"eval_samples_per_second": 174.368,
|
| 951 |
+
"eval_steps_per_second": 21.843,
|
| 952 |
+
"step": 4100
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 3.5591766723842193,
|
| 956 |
+
"grad_norm": 7.2456207275390625,
|
| 957 |
+
"learning_rate": 1.4408233276157807e-05,
|
| 958 |
+
"loss": 0.4479,
|
| 959 |
+
"step": 4150
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 3.6020583190394513,
|
| 963 |
+
"grad_norm": 9.010571479797363,
|
| 964 |
+
"learning_rate": 1.397941680960549e-05,
|
| 965 |
+
"loss": 0.3513,
|
| 966 |
+
"step": 4200
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 3.6020583190394513,
|
| 970 |
+
"eval_accuracy": 0.9043329043329044,
|
| 971 |
+
"eval_loss": 0.2614918053150177,
|
| 972 |
+
"eval_runtime": 13.6533,
|
| 973 |
+
"eval_samples_per_second": 170.728,
|
| 974 |
+
"eval_steps_per_second": 21.387,
|
| 975 |
+
"step": 4200
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.644939965694683,
|
| 979 |
+
"grad_norm": 8.282753944396973,
|
| 980 |
+
"learning_rate": 1.3550600343053174e-05,
|
| 981 |
+
"loss": 0.3045,
|
| 982 |
+
"step": 4250
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.6878216123499143,
|
| 986 |
+
"grad_norm": 8.212636947631836,
|
| 987 |
+
"learning_rate": 1.312178387650086e-05,
|
| 988 |
+
"loss": 0.4081,
|
| 989 |
+
"step": 4300
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.6878216123499143,
|
| 993 |
+
"eval_accuracy": 0.9073359073359073,
|
| 994 |
+
"eval_loss": 0.25328412652015686,
|
| 995 |
+
"eval_runtime": 13.7186,
|
| 996 |
+
"eval_samples_per_second": 169.915,
|
| 997 |
+
"eval_steps_per_second": 21.285,
|
| 998 |
+
"step": 4300
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 3.730703259005146,
|
| 1002 |
+
"grad_norm": 1.7500250339508057,
|
| 1003 |
+
"learning_rate": 1.2692967409948542e-05,
|
| 1004 |
+
"loss": 0.3552,
|
| 1005 |
+
"step": 4350
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 3.7735849056603774,
|
| 1009 |
+
"grad_norm": 38.88422393798828,
|
| 1010 |
+
"learning_rate": 1.2264150943396227e-05,
|
| 1011 |
+
"loss": 0.3346,
|
| 1012 |
+
"step": 4400
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 3.7735849056603774,
|
| 1016 |
+
"eval_accuracy": 0.9021879021879022,
|
| 1017 |
+
"eval_loss": 0.27931204438209534,
|
| 1018 |
+
"eval_runtime": 15.0366,
|
| 1019 |
+
"eval_samples_per_second": 155.022,
|
| 1020 |
+
"eval_steps_per_second": 19.419,
|
| 1021 |
+
"step": 4400
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 3.816466552315609,
|
| 1025 |
+
"grad_norm": 18.03175163269043,
|
| 1026 |
+
"learning_rate": 1.1835334476843911e-05,
|
| 1027 |
+
"loss": 0.3502,
|
| 1028 |
+
"step": 4450
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 3.8593481989708405,
|
| 1032 |
+
"grad_norm": 6.491177082061768,
|
| 1033 |
+
"learning_rate": 1.1406518010291597e-05,
|
| 1034 |
+
"loss": 0.3495,
|
| 1035 |
+
"step": 4500
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 3.8593481989708405,
|
| 1039 |
+
"eval_accuracy": 0.9056199056199056,
|
| 1040 |
+
"eval_loss": 0.2461409866809845,
|
| 1041 |
+
"eval_runtime": 14.6288,
|
| 1042 |
+
"eval_samples_per_second": 159.344,
|
| 1043 |
+
"eval_steps_per_second": 19.961,
|
| 1044 |
+
"step": 4500
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.902229845626072,
|
| 1048 |
+
"grad_norm": 22.98536491394043,
|
| 1049 |
+
"learning_rate": 1.097770154373928e-05,
|
| 1050 |
+
"loss": 0.3662,
|
| 1051 |
+
"step": 4550
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 3.9451114922813035,
|
| 1055 |
+
"grad_norm": 24.458087921142578,
|
| 1056 |
+
"learning_rate": 1.0548885077186965e-05,
|
| 1057 |
+
"loss": 0.3979,
|
| 1058 |
+
"step": 4600
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 3.9451114922813035,
|
| 1062 |
+
"eval_accuracy": 0.9103389103389103,
|
| 1063 |
+
"eval_loss": 0.24052156507968903,
|
| 1064 |
+
"eval_runtime": 14.1187,
|
| 1065 |
+
"eval_samples_per_second": 165.1,
|
| 1066 |
+
"eval_steps_per_second": 20.682,
|
| 1067 |
+
"step": 4600
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.987993138936535,
|
| 1071 |
+
"grad_norm": 22.207815170288086,
|
| 1072 |
+
"learning_rate": 1.012006861063465e-05,
|
| 1073 |
+
"loss": 0.379,
|
| 1074 |
+
"step": 4650
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.030874785591767,
|
| 1078 |
+
"grad_norm": 19.702865600585938,
|
| 1079 |
+
"learning_rate": 9.691252144082332e-06,
|
| 1080 |
+
"loss": 0.3441,
|
| 1081 |
+
"step": 4700
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.030874785591767,
|
| 1085 |
+
"eval_accuracy": 0.9017589017589017,
|
| 1086 |
+
"eval_loss": 0.27815285325050354,
|
| 1087 |
+
"eval_runtime": 14.2114,
|
| 1088 |
+
"eval_samples_per_second": 164.024,
|
| 1089 |
+
"eval_steps_per_second": 20.547,
|
| 1090 |
+
"step": 4700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.073756432246999,
|
| 1094 |
+
"grad_norm": 9.863823890686035,
|
| 1095 |
+
"learning_rate": 9.262435677530018e-06,
|
| 1096 |
+
"loss": 0.4128,
|
| 1097 |
+
"step": 4750
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.11663807890223,
|
| 1101 |
+
"grad_norm": 13.863909721374512,
|
| 1102 |
+
"learning_rate": 8.833619210977703e-06,
|
| 1103 |
+
"loss": 0.3357,
|
| 1104 |
+
"step": 4800
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 4.11663807890223,
|
| 1108 |
+
"eval_accuracy": 0.9094809094809094,
|
| 1109 |
+
"eval_loss": 0.26747584342956543,
|
| 1110 |
+
"eval_runtime": 13.9108,
|
| 1111 |
+
"eval_samples_per_second": 167.568,
|
| 1112 |
+
"eval_steps_per_second": 20.991,
|
| 1113 |
+
"step": 4800
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 4.159519725557462,
|
| 1117 |
+
"grad_norm": 45.47288131713867,
|
| 1118 |
+
"learning_rate": 8.404802744425385e-06,
|
| 1119 |
+
"loss": 0.3301,
|
| 1120 |
+
"step": 4850
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 4.202401372212693,
|
| 1124 |
+
"grad_norm": 4.5591864585876465,
|
| 1125 |
+
"learning_rate": 7.975986277873071e-06,
|
| 1126 |
+
"loss": 0.2867,
|
| 1127 |
+
"step": 4900
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 4.202401372212693,
|
| 1131 |
+
"eval_accuracy": 0.9060489060489061,
|
| 1132 |
+
"eval_loss": 0.27105966210365295,
|
| 1133 |
+
"eval_runtime": 15.175,
|
| 1134 |
+
"eval_samples_per_second": 153.608,
|
| 1135 |
+
"eval_steps_per_second": 19.242,
|
| 1136 |
+
"step": 4900
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.245283018867925,
|
| 1140 |
+
"grad_norm": 0.21429812908172607,
|
| 1141 |
+
"learning_rate": 7.547169811320755e-06,
|
| 1142 |
+
"loss": 0.3518,
|
| 1143 |
+
"step": 4950
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.288164665523156,
|
| 1147 |
+
"grad_norm": 4.873584747314453,
|
| 1148 |
+
"learning_rate": 7.11835334476844e-06,
|
| 1149 |
+
"loss": 0.3084,
|
| 1150 |
+
"step": 5000
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.288164665523156,
|
| 1154 |
+
"eval_accuracy": 0.9116259116259117,
|
| 1155 |
+
"eval_loss": 0.24336792528629303,
|
| 1156 |
+
"eval_runtime": 13.447,
|
| 1157 |
+
"eval_samples_per_second": 173.347,
|
| 1158 |
+
"eval_steps_per_second": 21.715,
|
| 1159 |
+
"step": 5000
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 4.331046312178388,
|
| 1163 |
+
"grad_norm": 3.682224988937378,
|
| 1164 |
+
"learning_rate": 6.689536878216124e-06,
|
| 1165 |
+
"loss": 0.3335,
|
| 1166 |
+
"step": 5050
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 4.373927958833619,
|
| 1170 |
+
"grad_norm": 8.78370475769043,
|
| 1171 |
+
"learning_rate": 6.2607204116638075e-06,
|
| 1172 |
+
"loss": 0.4865,
|
| 1173 |
+
"step": 5100
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 4.373927958833619,
|
| 1177 |
+
"eval_accuracy": 0.9154869154869155,
|
| 1178 |
+
"eval_loss": 0.2427978515625,
|
| 1179 |
+
"eval_runtime": 13.4143,
|
| 1180 |
+
"eval_samples_per_second": 173.77,
|
| 1181 |
+
"eval_steps_per_second": 21.768,
|
| 1182 |
+
"step": 5100
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 4.416809605488851,
|
| 1186 |
+
"grad_norm": 5.311465263366699,
|
| 1187 |
+
"learning_rate": 5.831903945111493e-06,
|
| 1188 |
+
"loss": 0.3465,
|
| 1189 |
+
"step": 5150
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 4.459691252144083,
|
| 1193 |
+
"grad_norm": 9.669562339782715,
|
| 1194 |
+
"learning_rate": 5.403087478559177e-06,
|
| 1195 |
+
"loss": 0.3557,
|
| 1196 |
+
"step": 5200
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.459691252144083,
|
| 1200 |
+
"eval_accuracy": 0.9141999141999142,
|
| 1201 |
+
"eval_loss": 0.2416524589061737,
|
| 1202 |
+
"eval_runtime": 13.3034,
|
| 1203 |
+
"eval_samples_per_second": 175.218,
|
| 1204 |
+
"eval_steps_per_second": 21.949,
|
| 1205 |
+
"step": 5200
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 4.502572898799314,
|
| 1209 |
+
"grad_norm": 8.7540283203125,
|
| 1210 |
+
"learning_rate": 4.974271012006861e-06,
|
| 1211 |
+
"loss": 0.3254,
|
| 1212 |
+
"step": 5250
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 4.545454545454545,
|
| 1216 |
+
"grad_norm": 0.46742549538612366,
|
| 1217 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 1218 |
+
"loss": 0.3384,
|
| 1219 |
+
"step": 5300
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 4.545454545454545,
|
| 1223 |
+
"eval_accuracy": 0.9133419133419134,
|
| 1224 |
+
"eval_loss": 0.24515940248966217,
|
| 1225 |
+
"eval_runtime": 13.6233,
|
| 1226 |
+
"eval_samples_per_second": 171.103,
|
| 1227 |
+
"eval_steps_per_second": 21.434,
|
| 1228 |
+
"step": 5300
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 4.588336192109777,
|
| 1232 |
+
"grad_norm": 23.387781143188477,
|
| 1233 |
+
"learning_rate": 4.116638078902231e-06,
|
| 1234 |
+
"loss": 0.3508,
|
| 1235 |
+
"step": 5350
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 4.631217838765009,
|
| 1239 |
+
"grad_norm": 3.3974390029907227,
|
| 1240 |
+
"learning_rate": 3.687821612349914e-06,
|
| 1241 |
+
"loss": 0.3147,
|
| 1242 |
+
"step": 5400
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 4.631217838765009,
|
| 1246 |
+
"eval_accuracy": 0.9141999141999142,
|
| 1247 |
+
"eval_loss": 0.24695377051830292,
|
| 1248 |
+
"eval_runtime": 13.4806,
|
| 1249 |
+
"eval_samples_per_second": 172.915,
|
| 1250 |
+
"eval_steps_per_second": 21.661,
|
| 1251 |
+
"step": 5400
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 4.67409948542024,
|
| 1255 |
+
"grad_norm": 26.67882537841797,
|
| 1256 |
+
"learning_rate": 3.259005145797599e-06,
|
| 1257 |
+
"loss": 0.3364,
|
| 1258 |
+
"step": 5450
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 4.716981132075472,
|
| 1262 |
+
"grad_norm": 0.8814979791641235,
|
| 1263 |
+
"learning_rate": 2.830188679245283e-06,
|
| 1264 |
+
"loss": 0.2858,
|
| 1265 |
+
"step": 5500
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 4.716981132075472,
|
| 1269 |
+
"eval_accuracy": 0.9103389103389103,
|
| 1270 |
+
"eval_loss": 0.24419519305229187,
|
| 1271 |
+
"eval_runtime": 13.9344,
|
| 1272 |
+
"eval_samples_per_second": 167.284,
|
| 1273 |
+
"eval_steps_per_second": 20.955,
|
| 1274 |
+
"step": 5500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 4.759862778730703,
|
| 1278 |
+
"grad_norm": 13.250667572021484,
|
| 1279 |
+
"learning_rate": 2.4013722126929674e-06,
|
| 1280 |
+
"loss": 0.3674,
|
| 1281 |
+
"step": 5550
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 4.802744425385935,
|
| 1285 |
+
"grad_norm": 11.3861722946167,
|
| 1286 |
+
"learning_rate": 1.972555746140652e-06,
|
| 1287 |
+
"loss": 0.3664,
|
| 1288 |
+
"step": 5600
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 4.802744425385935,
|
| 1292 |
+
"eval_accuracy": 0.9064779064779065,
|
| 1293 |
+
"eval_loss": 0.2746657431125641,
|
| 1294 |
+
"eval_runtime": 13.6304,
|
| 1295 |
+
"eval_samples_per_second": 171.015,
|
| 1296 |
+
"eval_steps_per_second": 21.423,
|
| 1297 |
+
"step": 5600
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.845626072041166,
|
| 1301 |
+
"grad_norm": 3.5501041412353516,
|
| 1302 |
+
"learning_rate": 1.5437392795883364e-06,
|
| 1303 |
+
"loss": 0.3152,
|
| 1304 |
+
"step": 5650
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.888507718696398,
|
| 1308 |
+
"grad_norm": 24.59325408935547,
|
| 1309 |
+
"learning_rate": 1.1149228130360207e-06,
|
| 1310 |
+
"loss": 0.3549,
|
| 1311 |
+
"step": 5700
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.888507718696398,
|
| 1315 |
+
"eval_accuracy": 0.9099099099099099,
|
| 1316 |
+
"eval_loss": 0.26241084933280945,
|
| 1317 |
+
"eval_runtime": 15.5182,
|
| 1318 |
+
"eval_samples_per_second": 150.211,
|
| 1319 |
+
"eval_steps_per_second": 18.817,
|
| 1320 |
+
"step": 5700
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 4.931389365351629,
|
| 1324 |
+
"grad_norm": 32.23326110839844,
|
| 1325 |
+
"learning_rate": 6.86106346483705e-07,
|
| 1326 |
+
"loss": 0.3796,
|
| 1327 |
+
"step": 5750
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 4.974271012006861,
|
| 1331 |
+
"grad_norm": 30.20063591003418,
|
| 1332 |
+
"learning_rate": 2.572898799313894e-07,
|
| 1333 |
+
"loss": 0.3656,
|
| 1334 |
+
"step": 5800
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 4.974271012006861,
|
| 1338 |
+
"eval_accuracy": 0.9047619047619048,
|
| 1339 |
+
"eval_loss": 0.27714791893959045,
|
| 1340 |
+
"eval_runtime": 14.0154,
|
| 1341 |
+
"eval_samples_per_second": 166.317,
|
| 1342 |
+
"eval_steps_per_second": 20.834,
|
| 1343 |
+
"step": 5800
|
| 1344 |
+
}
|
| 1345 |
+
],
|
| 1346 |
+
"logging_steps": 50,
|
| 1347 |
+
"max_steps": 5830,
|
| 1348 |
+
"num_input_tokens_seen": 0,
|
| 1349 |
+
"num_train_epochs": 5,
|
| 1350 |
+
"save_steps": 100,
|
| 1351 |
+
"stateful_callbacks": {
|
| 1352 |
+
"TrainerControl": {
|
| 1353 |
+
"args": {
|
| 1354 |
+
"should_epoch_stop": false,
|
| 1355 |
+
"should_evaluate": false,
|
| 1356 |
+
"should_log": false,
|
| 1357 |
+
"should_save": true,
|
| 1358 |
+
"should_training_stop": true
|
| 1359 |
+
},
|
| 1360 |
+
"attributes": {}
|
| 1361 |
+
}
|
| 1362 |
+
},
|
| 1363 |
+
"total_flos": 2.7169922825846784e+17,
|
| 1364 |
+
"train_batch_size": 8,
|
| 1365 |
+
"trial_name": null,
|
| 1366 |
+
"trial_params": null
|
| 1367 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c96a7c79e124bc4e6eec8eaf3b1366f4645272db90ffcc952f5d9b302bbfdbd
|
| 3 |
+
size 5368
|