Spaces:
Sleeping
Sleeping
Upload 4 files
Browse files- config.json +22 -14
- model.safetensors +2 -2
- trainer_state.json +853 -0
config.json
CHANGED
|
@@ -9,25 +9,33 @@
|
|
| 9 |
"hidden_dropout_prob": 0.0,
|
| 10 |
"hidden_size": 768,
|
| 11 |
"id2label": {
|
| 12 |
-
"0": "
|
| 13 |
-
"1": "
|
| 14 |
-
"2": "
|
| 15 |
-
"3": "
|
| 16 |
-
"4": "
|
| 17 |
-
"5": "
|
| 18 |
-
"6": "
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
},
|
| 20 |
"image_size": 224,
|
| 21 |
"initializer_range": 0.02,
|
| 22 |
"intermediate_size": 3072,
|
| 23 |
"label2id": {
|
| 24 |
-
"
|
| 25 |
-
"
|
| 26 |
-
"
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
},
|
| 32 |
"layer_norm_eps": 1e-12,
|
| 33 |
"model_type": "vit",
|
|
|
|
| 9 |
"hidden_dropout_prob": 0.0,
|
| 10 |
"hidden_size": 768,
|
| 11 |
"id2label": {
|
| 12 |
+
"0": "round",
|
| 13 |
+
"1": "vneck",
|
| 14 |
+
"2": "collared",
|
| 15 |
+
"3": "straight",
|
| 16 |
+
"4": "highneck",
|
| 17 |
+
"5": "hoodie",
|
| 18 |
+
"6": "henley",
|
| 19 |
+
"7": "halter",
|
| 20 |
+
"8": "sweetheart",
|
| 21 |
+
"9": "polo",
|
| 22 |
+
"10": "asymmetrical"
|
| 23 |
},
|
| 24 |
"image_size": 224,
|
| 25 |
"initializer_range": 0.02,
|
| 26 |
"intermediate_size": 3072,
|
| 27 |
"label2id": {
|
| 28 |
+
"asymmetrical": 10,
|
| 29 |
+
"collared": 2,
|
| 30 |
+
"halter": 7,
|
| 31 |
+
"henley": 6,
|
| 32 |
+
"highneck": 4,
|
| 33 |
+
"hoodie": 5,
|
| 34 |
+
"polo": 9,
|
| 35 |
+
"round": 0,
|
| 36 |
+
"straight": 3,
|
| 37 |
+
"sweetheart": 8,
|
| 38 |
+
"vneck": 1
|
| 39 |
},
|
| 40 |
"layer_norm_eps": 1e-12,
|
| 41 |
"model_type": "vit",
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24e0f7acb425a3d5310cd89ffea6e9f6179451404295fc887fce8eb9f32e6731
|
| 3 |
+
size 343251660
|
trainer_state.json
ADDED
|
@@ -0,0 +1,853 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7836363636363637,
|
| 3 |
+
"best_model_checkpoint": "/content/drive/MyDrive/autoTaggingProject/ViT/General/Features/NeckLine/Results/model_2024-10-16_test/checkpoint-5192",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 5192,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1059322033898305,
|
| 13 |
+
"grad_norm": 1.528609037399292,
|
| 14 |
+
"learning_rate": 9.964689265536724e-06,
|
| 15 |
+
"loss": 2.3665,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.211864406779661,
|
| 20 |
+
"grad_norm": 1.4840689897537231,
|
| 21 |
+
"learning_rate": 9.929378531073447e-06,
|
| 22 |
+
"loss": 2.2771,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.3177966101694915,
|
| 27 |
+
"grad_norm": 1.6797162294387817,
|
| 28 |
+
"learning_rate": 9.89406779661017e-06,
|
| 29 |
+
"loss": 2.1584,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.423728813559322,
|
| 34 |
+
"grad_norm": 1.5869511365890503,
|
| 35 |
+
"learning_rate": 9.858757062146892e-06,
|
| 36 |
+
"loss": 2.0337,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5296610169491526,
|
| 41 |
+
"grad_norm": 1.625272512435913,
|
| 42 |
+
"learning_rate": 9.823446327683617e-06,
|
| 43 |
+
"loss": 1.945,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.635593220338983,
|
| 48 |
+
"grad_norm": 1.918320655822754,
|
| 49 |
+
"learning_rate": 9.788135593220339e-06,
|
| 50 |
+
"loss": 1.8575,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7415254237288136,
|
| 55 |
+
"grad_norm": 2.73588490486145,
|
| 56 |
+
"learning_rate": 9.752824858757062e-06,
|
| 57 |
+
"loss": 1.795,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.847457627118644,
|
| 62 |
+
"grad_norm": 1.978835105895996,
|
| 63 |
+
"learning_rate": 9.717514124293787e-06,
|
| 64 |
+
"loss": 1.7206,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.9533898305084746,
|
| 69 |
+
"grad_norm": 2.827810764312744,
|
| 70 |
+
"learning_rate": 9.682203389830509e-06,
|
| 71 |
+
"loss": 1.643,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"eval_accuracy": 0.5754545454545454,
|
| 77 |
+
"eval_loss": 1.6228725910186768,
|
| 78 |
+
"eval_runtime": 14.283,
|
| 79 |
+
"eval_samples_per_second": 77.014,
|
| 80 |
+
"eval_steps_per_second": 2.45,
|
| 81 |
+
"step": 472
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.0593220338983051,
|
| 85 |
+
"grad_norm": 2.191129207611084,
|
| 86 |
+
"learning_rate": 9.646892655367232e-06,
|
| 87 |
+
"loss": 1.5876,
|
| 88 |
+
"step": 500
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.1652542372881356,
|
| 92 |
+
"grad_norm": 2.16697359085083,
|
| 93 |
+
"learning_rate": 9.611581920903955e-06,
|
| 94 |
+
"loss": 1.5416,
|
| 95 |
+
"step": 550
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.271186440677966,
|
| 99 |
+
"grad_norm": 2.1197969913482666,
|
| 100 |
+
"learning_rate": 9.576271186440679e-06,
|
| 101 |
+
"loss": 1.5015,
|
| 102 |
+
"step": 600
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.3771186440677967,
|
| 106 |
+
"grad_norm": 2.465144634246826,
|
| 107 |
+
"learning_rate": 9.540960451977402e-06,
|
| 108 |
+
"loss": 1.4279,
|
| 109 |
+
"step": 650
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.4830508474576272,
|
| 113 |
+
"grad_norm": 1.8128849267959595,
|
| 114 |
+
"learning_rate": 9.505649717514125e-06,
|
| 115 |
+
"loss": 1.4298,
|
| 116 |
+
"step": 700
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.5889830508474576,
|
| 120 |
+
"grad_norm": 2.8659541606903076,
|
| 121 |
+
"learning_rate": 9.470338983050848e-06,
|
| 122 |
+
"loss": 1.3816,
|
| 123 |
+
"step": 750
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.694915254237288,
|
| 127 |
+
"grad_norm": 2.582030773162842,
|
| 128 |
+
"learning_rate": 9.435028248587572e-06,
|
| 129 |
+
"loss": 1.3254,
|
| 130 |
+
"step": 800
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.8008474576271185,
|
| 134 |
+
"grad_norm": 2.6557815074920654,
|
| 135 |
+
"learning_rate": 9.399717514124295e-06,
|
| 136 |
+
"loss": 1.2937,
|
| 137 |
+
"step": 850
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.9067796610169492,
|
| 141 |
+
"grad_norm": 3.4831998348236084,
|
| 142 |
+
"learning_rate": 9.364406779661017e-06,
|
| 143 |
+
"loss": 1.2511,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.0,
|
| 148 |
+
"eval_accuracy": 0.6927272727272727,
|
| 149 |
+
"eval_loss": 1.3029816150665283,
|
| 150 |
+
"eval_runtime": 14.6712,
|
| 151 |
+
"eval_samples_per_second": 74.977,
|
| 152 |
+
"eval_steps_per_second": 2.386,
|
| 153 |
+
"step": 944
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.01271186440678,
|
| 157 |
+
"grad_norm": 2.9632270336151123,
|
| 158 |
+
"learning_rate": 9.329096045197742e-06,
|
| 159 |
+
"loss": 1.2574,
|
| 160 |
+
"step": 950
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.1186440677966103,
|
| 164 |
+
"grad_norm": 4.297321796417236,
|
| 165 |
+
"learning_rate": 9.293785310734465e-06,
|
| 166 |
+
"loss": 1.213,
|
| 167 |
+
"step": 1000
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.2245762711864407,
|
| 171 |
+
"grad_norm": 5.289111614227295,
|
| 172 |
+
"learning_rate": 9.258474576271187e-06,
|
| 173 |
+
"loss": 1.2127,
|
| 174 |
+
"step": 1050
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.330508474576271,
|
| 178 |
+
"grad_norm": 5.046447277069092,
|
| 179 |
+
"learning_rate": 9.22316384180791e-06,
|
| 180 |
+
"loss": 1.1403,
|
| 181 |
+
"step": 1100
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.4364406779661016,
|
| 185 |
+
"grad_norm": 3.322784662246704,
|
| 186 |
+
"learning_rate": 9.187853107344633e-06,
|
| 187 |
+
"loss": 1.1244,
|
| 188 |
+
"step": 1150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.542372881355932,
|
| 192 |
+
"grad_norm": 2.8034584522247314,
|
| 193 |
+
"learning_rate": 9.152542372881356e-06,
|
| 194 |
+
"loss": 1.1052,
|
| 195 |
+
"step": 1200
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.648305084745763,
|
| 199 |
+
"grad_norm": 2.822523593902588,
|
| 200 |
+
"learning_rate": 9.11723163841808e-06,
|
| 201 |
+
"loss": 1.0929,
|
| 202 |
+
"step": 1250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.7542372881355934,
|
| 206 |
+
"grad_norm": 4.488712787628174,
|
| 207 |
+
"learning_rate": 9.081920903954803e-06,
|
| 208 |
+
"loss": 1.0991,
|
| 209 |
+
"step": 1300
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.860169491525424,
|
| 213 |
+
"grad_norm": 4.986589431762695,
|
| 214 |
+
"learning_rate": 9.046610169491526e-06,
|
| 215 |
+
"loss": 1.0464,
|
| 216 |
+
"step": 1350
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.9661016949152543,
|
| 220 |
+
"grad_norm": 4.071073532104492,
|
| 221 |
+
"learning_rate": 9.01129943502825e-06,
|
| 222 |
+
"loss": 1.0532,
|
| 223 |
+
"step": 1400
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 3.0,
|
| 227 |
+
"eval_accuracy": 0.7209090909090909,
|
| 228 |
+
"eval_loss": 1.0850567817687988,
|
| 229 |
+
"eval_runtime": 17.8307,
|
| 230 |
+
"eval_samples_per_second": 61.691,
|
| 231 |
+
"eval_steps_per_second": 1.963,
|
| 232 |
+
"step": 1416
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.0720338983050848,
|
| 236 |
+
"grad_norm": 4.750606536865234,
|
| 237 |
+
"learning_rate": 8.975988700564973e-06,
|
| 238 |
+
"loss": 1.0453,
|
| 239 |
+
"step": 1450
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.1779661016949152,
|
| 243 |
+
"grad_norm": 3.7604198455810547,
|
| 244 |
+
"learning_rate": 8.940677966101694e-06,
|
| 245 |
+
"loss": 1.0231,
|
| 246 |
+
"step": 1500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.2838983050847457,
|
| 250 |
+
"grad_norm": 5.222371578216553,
|
| 251 |
+
"learning_rate": 8.90536723163842e-06,
|
| 252 |
+
"loss": 0.9812,
|
| 253 |
+
"step": 1550
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.389830508474576,
|
| 257 |
+
"grad_norm": 6.1730570793151855,
|
| 258 |
+
"learning_rate": 8.870056497175143e-06,
|
| 259 |
+
"loss": 0.9771,
|
| 260 |
+
"step": 1600
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.4957627118644066,
|
| 264 |
+
"grad_norm": 5.553199291229248,
|
| 265 |
+
"learning_rate": 8.834745762711864e-06,
|
| 266 |
+
"loss": 0.9682,
|
| 267 |
+
"step": 1650
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.601694915254237,
|
| 271 |
+
"grad_norm": 4.2731451988220215,
|
| 272 |
+
"learning_rate": 8.79943502824859e-06,
|
| 273 |
+
"loss": 1.0006,
|
| 274 |
+
"step": 1700
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.707627118644068,
|
| 278 |
+
"grad_norm": 3.617053985595703,
|
| 279 |
+
"learning_rate": 8.764124293785311e-06,
|
| 280 |
+
"loss": 0.9154,
|
| 281 |
+
"step": 1750
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.8135593220338984,
|
| 285 |
+
"grad_norm": 2.5833144187927246,
|
| 286 |
+
"learning_rate": 8.728813559322034e-06,
|
| 287 |
+
"loss": 0.935,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.919491525423729,
|
| 292 |
+
"grad_norm": 3.1606085300445557,
|
| 293 |
+
"learning_rate": 8.693502824858758e-06,
|
| 294 |
+
"loss": 0.9216,
|
| 295 |
+
"step": 1850
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 4.0,
|
| 299 |
+
"eval_accuracy": 0.74,
|
| 300 |
+
"eval_loss": 0.9795950651168823,
|
| 301 |
+
"eval_runtime": 16.0161,
|
| 302 |
+
"eval_samples_per_second": 68.681,
|
| 303 |
+
"eval_steps_per_second": 2.185,
|
| 304 |
+
"step": 1888
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.02542372881356,
|
| 308 |
+
"grad_norm": 3.959052801132202,
|
| 309 |
+
"learning_rate": 8.65819209039548e-06,
|
| 310 |
+
"loss": 0.9288,
|
| 311 |
+
"step": 1900
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.13135593220339,
|
| 315 |
+
"grad_norm": 3.6853768825531006,
|
| 316 |
+
"learning_rate": 8.622881355932204e-06,
|
| 317 |
+
"loss": 0.879,
|
| 318 |
+
"step": 1950
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.237288135593221,
|
| 322 |
+
"grad_norm": 3.1763620376586914,
|
| 323 |
+
"learning_rate": 8.587570621468927e-06,
|
| 324 |
+
"loss": 0.8912,
|
| 325 |
+
"step": 2000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.343220338983051,
|
| 329 |
+
"grad_norm": 4.019489288330078,
|
| 330 |
+
"learning_rate": 8.55225988700565e-06,
|
| 331 |
+
"loss": 0.9077,
|
| 332 |
+
"step": 2050
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.4491525423728815,
|
| 336 |
+
"grad_norm": 4.087663650512695,
|
| 337 |
+
"learning_rate": 8.516949152542372e-06,
|
| 338 |
+
"loss": 0.8812,
|
| 339 |
+
"step": 2100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.555084745762712,
|
| 343 |
+
"grad_norm": 3.6994051933288574,
|
| 344 |
+
"learning_rate": 8.481638418079097e-06,
|
| 345 |
+
"loss": 0.8344,
|
| 346 |
+
"step": 2150
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.661016949152542,
|
| 350 |
+
"grad_norm": 2.284302234649658,
|
| 351 |
+
"learning_rate": 8.44632768361582e-06,
|
| 352 |
+
"loss": 0.8501,
|
| 353 |
+
"step": 2200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.766949152542373,
|
| 357 |
+
"grad_norm": 4.348343372344971,
|
| 358 |
+
"learning_rate": 8.411016949152542e-06,
|
| 359 |
+
"loss": 0.8712,
|
| 360 |
+
"step": 2250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.872881355932203,
|
| 364 |
+
"grad_norm": 4.996354103088379,
|
| 365 |
+
"learning_rate": 8.375706214689267e-06,
|
| 366 |
+
"loss": 0.8264,
|
| 367 |
+
"step": 2300
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.978813559322034,
|
| 371 |
+
"grad_norm": 3.136770486831665,
|
| 372 |
+
"learning_rate": 8.340395480225989e-06,
|
| 373 |
+
"loss": 0.843,
|
| 374 |
+
"step": 2350
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 5.0,
|
| 378 |
+
"eval_accuracy": 0.7390909090909091,
|
| 379 |
+
"eval_loss": 0.9454855918884277,
|
| 380 |
+
"eval_runtime": 15.0504,
|
| 381 |
+
"eval_samples_per_second": 73.088,
|
| 382 |
+
"eval_steps_per_second": 2.326,
|
| 383 |
+
"step": 2360
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.084745762711864,
|
| 387 |
+
"grad_norm": 6.956203937530518,
|
| 388 |
+
"learning_rate": 8.305084745762712e-06,
|
| 389 |
+
"loss": 0.8094,
|
| 390 |
+
"step": 2400
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.190677966101695,
|
| 394 |
+
"grad_norm": 6.91636323928833,
|
| 395 |
+
"learning_rate": 8.269774011299437e-06,
|
| 396 |
+
"loss": 0.8301,
|
| 397 |
+
"step": 2450
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.296610169491525,
|
| 401 |
+
"grad_norm": 2.561798334121704,
|
| 402 |
+
"learning_rate": 8.234463276836159e-06,
|
| 403 |
+
"loss": 0.8562,
|
| 404 |
+
"step": 2500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.4025423728813555,
|
| 408 |
+
"grad_norm": 4.503079891204834,
|
| 409 |
+
"learning_rate": 8.199152542372882e-06,
|
| 410 |
+
"loss": 0.7487,
|
| 411 |
+
"step": 2550
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.508474576271187,
|
| 415 |
+
"grad_norm": 3.560302257537842,
|
| 416 |
+
"learning_rate": 8.163841807909605e-06,
|
| 417 |
+
"loss": 0.8222,
|
| 418 |
+
"step": 2600
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 5.614406779661017,
|
| 422 |
+
"grad_norm": 6.565722465515137,
|
| 423 |
+
"learning_rate": 8.128531073446328e-06,
|
| 424 |
+
"loss": 0.7917,
|
| 425 |
+
"step": 2650
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.720338983050848,
|
| 429 |
+
"grad_norm": 7.790140151977539,
|
| 430 |
+
"learning_rate": 8.093220338983052e-06,
|
| 431 |
+
"loss": 0.7764,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 5.826271186440678,
|
| 436 |
+
"grad_norm": 4.017592430114746,
|
| 437 |
+
"learning_rate": 8.057909604519775e-06,
|
| 438 |
+
"loss": 0.7718,
|
| 439 |
+
"step": 2750
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 5.932203389830509,
|
| 443 |
+
"grad_norm": 6.110499382019043,
|
| 444 |
+
"learning_rate": 8.022598870056498e-06,
|
| 445 |
+
"loss": 0.7445,
|
| 446 |
+
"step": 2800
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 6.0,
|
| 450 |
+
"eval_accuracy": 0.7527272727272727,
|
| 451 |
+
"eval_loss": 0.8885732889175415,
|
| 452 |
+
"eval_runtime": 15.0594,
|
| 453 |
+
"eval_samples_per_second": 73.044,
|
| 454 |
+
"eval_steps_per_second": 2.324,
|
| 455 |
+
"step": 2832
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.038135593220339,
|
| 459 |
+
"grad_norm": 4.600073337554932,
|
| 460 |
+
"learning_rate": 7.987288135593222e-06,
|
| 461 |
+
"loss": 0.7767,
|
| 462 |
+
"step": 2850
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.1440677966101696,
|
| 466 |
+
"grad_norm": 5.356403827667236,
|
| 467 |
+
"learning_rate": 7.951977401129945e-06,
|
| 468 |
+
"loss": 0.708,
|
| 469 |
+
"step": 2900
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.25,
|
| 473 |
+
"grad_norm": 5.1397223472595215,
|
| 474 |
+
"learning_rate": 7.917372881355932e-06,
|
| 475 |
+
"loss": 0.7397,
|
| 476 |
+
"step": 2950
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.3559322033898304,
|
| 480 |
+
"grad_norm": 6.384206771850586,
|
| 481 |
+
"learning_rate": 7.882062146892657e-06,
|
| 482 |
+
"loss": 0.7523,
|
| 483 |
+
"step": 3000
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.461864406779661,
|
| 487 |
+
"grad_norm": 5.545274257659912,
|
| 488 |
+
"learning_rate": 7.846751412429378e-06,
|
| 489 |
+
"loss": 0.7011,
|
| 490 |
+
"step": 3050
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.567796610169491,
|
| 494 |
+
"grad_norm": 9.406649589538574,
|
| 495 |
+
"learning_rate": 7.811440677966102e-06,
|
| 496 |
+
"loss": 0.7975,
|
| 497 |
+
"step": 3100
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 6.673728813559322,
|
| 501 |
+
"grad_norm": 7.81419563293457,
|
| 502 |
+
"learning_rate": 7.776129943502827e-06,
|
| 503 |
+
"loss": 0.7385,
|
| 504 |
+
"step": 3150
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 6.779661016949152,
|
| 508 |
+
"grad_norm": 3.415956974029541,
|
| 509 |
+
"learning_rate": 7.740819209039548e-06,
|
| 510 |
+
"loss": 0.7356,
|
| 511 |
+
"step": 3200
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 6.885593220338983,
|
| 515 |
+
"grad_norm": 5.487062931060791,
|
| 516 |
+
"learning_rate": 7.705508474576271e-06,
|
| 517 |
+
"loss": 0.7121,
|
| 518 |
+
"step": 3250
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 6.991525423728813,
|
| 522 |
+
"grad_norm": 5.682718276977539,
|
| 523 |
+
"learning_rate": 7.670197740112995e-06,
|
| 524 |
+
"loss": 0.7191,
|
| 525 |
+
"step": 3300
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 7.0,
|
| 529 |
+
"eval_accuracy": 0.7545454545454545,
|
| 530 |
+
"eval_loss": 0.8240677118301392,
|
| 531 |
+
"eval_runtime": 14.7678,
|
| 532 |
+
"eval_samples_per_second": 74.486,
|
| 533 |
+
"eval_steps_per_second": 2.37,
|
| 534 |
+
"step": 3304
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.0974576271186445,
|
| 538 |
+
"grad_norm": 8.319087028503418,
|
| 539 |
+
"learning_rate": 7.634887005649718e-06,
|
| 540 |
+
"loss": 0.7096,
|
| 541 |
+
"step": 3350
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.203389830508475,
|
| 545 |
+
"grad_norm": 5.857816696166992,
|
| 546 |
+
"learning_rate": 7.599576271186442e-06,
|
| 547 |
+
"loss": 0.7006,
|
| 548 |
+
"step": 3400
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.309322033898305,
|
| 552 |
+
"grad_norm": 4.400519371032715,
|
| 553 |
+
"learning_rate": 7.564265536723165e-06,
|
| 554 |
+
"loss": 0.706,
|
| 555 |
+
"step": 3450
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.415254237288136,
|
| 559 |
+
"grad_norm": 4.573615550994873,
|
| 560 |
+
"learning_rate": 7.528954802259888e-06,
|
| 561 |
+
"loss": 0.6543,
|
| 562 |
+
"step": 3500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.521186440677966,
|
| 566 |
+
"grad_norm": 7.545746803283691,
|
| 567 |
+
"learning_rate": 7.49364406779661e-06,
|
| 568 |
+
"loss": 0.6498,
|
| 569 |
+
"step": 3550
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.627118644067797,
|
| 573 |
+
"grad_norm": 6.38883638381958,
|
| 574 |
+
"learning_rate": 7.4583333333333345e-06,
|
| 575 |
+
"loss": 0.6849,
|
| 576 |
+
"step": 3600
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 7.733050847457627,
|
| 580 |
+
"grad_norm": 4.496486186981201,
|
| 581 |
+
"learning_rate": 7.423022598870057e-06,
|
| 582 |
+
"loss": 0.6741,
|
| 583 |
+
"step": 3650
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 7.838983050847458,
|
| 587 |
+
"grad_norm": 5.2381792068481445,
|
| 588 |
+
"learning_rate": 7.38771186440678e-06,
|
| 589 |
+
"loss": 0.6441,
|
| 590 |
+
"step": 3700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 7.944915254237288,
|
| 594 |
+
"grad_norm": 6.047347068786621,
|
| 595 |
+
"learning_rate": 7.3524011299435035e-06,
|
| 596 |
+
"loss": 0.7089,
|
| 597 |
+
"step": 3750
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 8.0,
|
| 601 |
+
"eval_accuracy": 0.7718181818181818,
|
| 602 |
+
"eval_loss": 0.845079779624939,
|
| 603 |
+
"eval_runtime": 14.9091,
|
| 604 |
+
"eval_samples_per_second": 73.78,
|
| 605 |
+
"eval_steps_per_second": 2.348,
|
| 606 |
+
"step": 3776
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 8.05084745762712,
|
| 610 |
+
"grad_norm": 5.383782386779785,
|
| 611 |
+
"learning_rate": 7.317090395480226e-06,
|
| 612 |
+
"loss": 0.6375,
|
| 613 |
+
"step": 3800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 8.15677966101695,
|
| 617 |
+
"grad_norm": 7.663337707519531,
|
| 618 |
+
"learning_rate": 7.28177966101695e-06,
|
| 619 |
+
"loss": 0.6101,
|
| 620 |
+
"step": 3850
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 8.26271186440678,
|
| 624 |
+
"grad_norm": 4.593461036682129,
|
| 625 |
+
"learning_rate": 7.2464689265536725e-06,
|
| 626 |
+
"loss": 0.6356,
|
| 627 |
+
"step": 3900
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 8.36864406779661,
|
| 631 |
+
"grad_norm": 7.878734111785889,
|
| 632 |
+
"learning_rate": 7.211158192090396e-06,
|
| 633 |
+
"loss": 0.6412,
|
| 634 |
+
"step": 3950
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 8.474576271186441,
|
| 638 |
+
"grad_norm": 3.501059055328369,
|
| 639 |
+
"learning_rate": 7.17584745762712e-06,
|
| 640 |
+
"loss": 0.6296,
|
| 641 |
+
"step": 4000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.580508474576272,
|
| 645 |
+
"grad_norm": 3.8199708461761475,
|
| 646 |
+
"learning_rate": 7.140536723163842e-06,
|
| 647 |
+
"loss": 0.654,
|
| 648 |
+
"step": 4050
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 8.686440677966102,
|
| 652 |
+
"grad_norm": 6.42057466506958,
|
| 653 |
+
"learning_rate": 7.105225988700566e-06,
|
| 654 |
+
"loss": 0.6378,
|
| 655 |
+
"step": 4100
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 8.792372881355933,
|
| 659 |
+
"grad_norm": 6.310295104980469,
|
| 660 |
+
"learning_rate": 7.069915254237288e-06,
|
| 661 |
+
"loss": 0.6235,
|
| 662 |
+
"step": 4150
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 8.898305084745763,
|
| 666 |
+
"grad_norm": 4.627810478210449,
|
| 667 |
+
"learning_rate": 7.034604519774012e-06,
|
| 668 |
+
"loss": 0.6631,
|
| 669 |
+
"step": 4200
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 9.0,
|
| 673 |
+
"eval_accuracy": 0.77,
|
| 674 |
+
"eval_loss": 0.8626542687416077,
|
| 675 |
+
"eval_runtime": 15.548,
|
| 676 |
+
"eval_samples_per_second": 70.749,
|
| 677 |
+
"eval_steps_per_second": 2.251,
|
| 678 |
+
"step": 4248
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 9.004237288135593,
|
| 682 |
+
"grad_norm": 5.442898273468018,
|
| 683 |
+
"learning_rate": 6.999293785310735e-06,
|
| 684 |
+
"loss": 0.6358,
|
| 685 |
+
"step": 4250
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 9.110169491525424,
|
| 689 |
+
"grad_norm": 5.042696475982666,
|
| 690 |
+
"learning_rate": 6.963983050847458e-06,
|
| 691 |
+
"loss": 0.6183,
|
| 692 |
+
"step": 4300
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 9.216101694915254,
|
| 696 |
+
"grad_norm": 5.006898403167725,
|
| 697 |
+
"learning_rate": 6.928672316384182e-06,
|
| 698 |
+
"loss": 0.6438,
|
| 699 |
+
"step": 4350
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 9.322033898305085,
|
| 703 |
+
"grad_norm": 6.093140125274658,
|
| 704 |
+
"learning_rate": 6.893361581920905e-06,
|
| 705 |
+
"loss": 0.5826,
|
| 706 |
+
"step": 4400
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 9.427966101694915,
|
| 710 |
+
"grad_norm": 4.637847900390625,
|
| 711 |
+
"learning_rate": 6.858050847457628e-06,
|
| 712 |
+
"loss": 0.5559,
|
| 713 |
+
"step": 4450
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 9.533898305084746,
|
| 717 |
+
"grad_norm": 2.860111951828003,
|
| 718 |
+
"learning_rate": 6.82274011299435e-06,
|
| 719 |
+
"loss": 0.5577,
|
| 720 |
+
"step": 4500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 9.639830508474576,
|
| 724 |
+
"grad_norm": 10.876856803894043,
|
| 725 |
+
"learning_rate": 6.7874293785310745e-06,
|
| 726 |
+
"loss": 0.6233,
|
| 727 |
+
"step": 4550
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 9.745762711864407,
|
| 731 |
+
"grad_norm": 5.635727882385254,
|
| 732 |
+
"learning_rate": 6.752118644067798e-06,
|
| 733 |
+
"loss": 0.5703,
|
| 734 |
+
"step": 4600
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 9.851694915254237,
|
| 738 |
+
"grad_norm": 6.9388532638549805,
|
| 739 |
+
"learning_rate": 6.71680790960452e-06,
|
| 740 |
+
"loss": 0.6323,
|
| 741 |
+
"step": 4650
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 9.957627118644067,
|
| 745 |
+
"grad_norm": 7.485644340515137,
|
| 746 |
+
"learning_rate": 6.6814971751412435e-06,
|
| 747 |
+
"loss": 0.6021,
|
| 748 |
+
"step": 4700
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 10.0,
|
| 752 |
+
"eval_accuracy": 0.7772727272727272,
|
| 753 |
+
"eval_loss": 0.8030957579612732,
|
| 754 |
+
"eval_runtime": 14.6023,
|
| 755 |
+
"eval_samples_per_second": 75.331,
|
| 756 |
+
"eval_steps_per_second": 2.397,
|
| 757 |
+
"step": 4720
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 10.063559322033898,
|
| 761 |
+
"grad_norm": 13.398885726928711,
|
| 762 |
+
"learning_rate": 6.646186440677966e-06,
|
| 763 |
+
"loss": 0.5698,
|
| 764 |
+
"step": 4750
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 10.169491525423728,
|
| 768 |
+
"grad_norm": 7.821059226989746,
|
| 769 |
+
"learning_rate": 6.61087570621469e-06,
|
| 770 |
+
"loss": 0.5364,
|
| 771 |
+
"step": 4800
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 10.275423728813559,
|
| 775 |
+
"grad_norm": 7.535600662231445,
|
| 776 |
+
"learning_rate": 6.576271186440678e-06,
|
| 777 |
+
"loss": 0.5959,
|
| 778 |
+
"step": 4850
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 10.38135593220339,
|
| 782 |
+
"grad_norm": 8.605109214782715,
|
| 783 |
+
"learning_rate": 6.540960451977402e-06,
|
| 784 |
+
"loss": 0.5837,
|
| 785 |
+
"step": 4900
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 10.48728813559322,
|
| 789 |
+
"grad_norm": 5.240401268005371,
|
| 790 |
+
"learning_rate": 6.505649717514125e-06,
|
| 791 |
+
"loss": 0.5796,
|
| 792 |
+
"step": 4950
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 10.59322033898305,
|
| 796 |
+
"grad_norm": 6.025055408477783,
|
| 797 |
+
"learning_rate": 6.4703389830508476e-06,
|
| 798 |
+
"loss": 0.5749,
|
| 799 |
+
"step": 5000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 10.69915254237288,
|
| 803 |
+
"grad_norm": 6.605931282043457,
|
| 804 |
+
"learning_rate": 6.435028248587572e-06,
|
| 805 |
+
"loss": 0.5185,
|
| 806 |
+
"step": 5050
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 10.805084745762711,
|
| 810 |
+
"grad_norm": 3.71102237701416,
|
| 811 |
+
"learning_rate": 6.399717514124294e-06,
|
| 812 |
+
"loss": 0.5289,
|
| 813 |
+
"step": 5100
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 10.911016949152543,
|
| 817 |
+
"grad_norm": 4.973482608795166,
|
| 818 |
+
"learning_rate": 6.3644067796610174e-06,
|
| 819 |
+
"loss": 0.5691,
|
| 820 |
+
"step": 5150
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 11.0,
|
| 824 |
+
"eval_accuracy": 0.7836363636363637,
|
| 825 |
+
"eval_loss": 0.8530685305595398,
|
| 826 |
+
"eval_runtime": 14.8369,
|
| 827 |
+
"eval_samples_per_second": 74.14,
|
| 828 |
+
"eval_steps_per_second": 2.359,
|
| 829 |
+
"step": 5192
|
| 830 |
+
}
|
| 831 |
+
],
|
| 832 |
+
"logging_steps": 50,
|
| 833 |
+
"max_steps": 14160,
|
| 834 |
+
"num_input_tokens_seen": 0,
|
| 835 |
+
"num_train_epochs": 30,
|
| 836 |
+
"save_steps": 500,
|
| 837 |
+
"stateful_callbacks": {
|
| 838 |
+
"TrainerControl": {
|
| 839 |
+
"args": {
|
| 840 |
+
"should_epoch_stop": false,
|
| 841 |
+
"should_evaluate": false,
|
| 842 |
+
"should_log": false,
|
| 843 |
+
"should_save": true,
|
| 844 |
+
"should_training_stop": false
|
| 845 |
+
},
|
| 846 |
+
"attributes": {}
|
| 847 |
+
}
|
| 848 |
+
},
|
| 849 |
+
"total_flos": 1.286478541133356e+19,
|
| 850 |
+
"train_batch_size": 32,
|
| 851 |
+
"trial_name": null,
|
| 852 |
+
"trial_params": null
|
| 853 |
+
}
|