Add files using upload-large-folder tool
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- cross_cell_type_generization/L2/Endotheliums/.DS_Store +0 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/.DS_Store +0 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/all_results.json +8 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/checkpoint-912/config.json +25 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/checkpoint-912/trainer_state.json +682 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/config.json +25 -0
- cross_cell_type_generization/L2/Endotheliums/fold2/eval_results.json +8 -0
- cross_cell_type_generization/L2/Endotheliums/fold4/.DS_Store +0 -0
- cross_cell_type_generization/L2/Endotheliums/fold4/all_results.json +8 -0
- cross_cell_type_generization/L2/Endotheliums/fold4/checkpoint-836/config.json +25 -0
- cross_cell_type_generization/L2/Endotheliums/fold4/config.json +25 -0
- cross_cell_type_generization/L2/Endotheliums/fold4/eval_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/all_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/checkpoint-836/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/checkpoint-836/trainer_state.json +624 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold0/eval_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/all_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/checkpoint-760/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/checkpoint-760/trainer_state.json +572 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold1/eval_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/all_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/checkpoint-912/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/checkpoint-912/trainer_state.json +682 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold2/eval_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/all_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/checkpoint-760/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/checkpoint-760/trainer_state.json +572 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold3/eval_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/.DS_Store +0 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/all_results.json +8 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/checkpoint-836/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/checkpoint-836/trainer_state.json +624 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/config.json +25 -0
- cross_cell_type_generization/L2/Epitheliums/fold4/eval_results.json +8 -0
- cross_cell_type_generization/L2/Macrophages/fold0/checkpoint-836/trainer_state.json +624 -0
- cross_cell_type_generization/L2/Neutrophils/.DS_Store +0 -0
- cross_cell_type_generization/L2/Neutrophils/fold0/.DS_Store +0 -0
- cross_cell_type_generization/L2/Neutrophils/fold0/all_results.json +8 -0
- cross_cell_type_generization/L2/Neutrophils/fold0/checkpoint-836/config.json +25 -0
- cross_cell_type_generization/L2/Neutrophils/fold0/checkpoint-836/trainer_state.json +624 -0
- cross_cell_type_generization/L2/Neutrophils/fold0/config.json +25 -0
cross_cell_type_generization/L2/Endotheliums/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Endotheliums/fold2/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Endotheliums/fold2/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8977777777777778,
|
| 3 |
+
"test_loss": 0.2747178077697754,
|
| 4 |
+
"test_macro_f1": 0.8641054594154565,
|
| 5 |
+
"test_runtime": 2.2566,
|
| 6 |
+
"test_samples_per_second": 398.823,
|
| 7 |
+
"test_steps_per_second": 1.773
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold2/checkpoint-912/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold2/checkpoint-912/trainer_state.json
ADDED
|
@@ -0,0 +1,682 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4261420667171478,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Endotheliums/fold2/checkpoint-760",
|
| 4 |
+
"epoch": 12.0,
|
| 5 |
+
"global_step": 912,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6971,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6973,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6956,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6955,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6936,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6925,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6899,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.5309533635988444,
|
| 55 |
+
"eval_loss": 0.6913440823554993,
|
| 56 |
+
"eval_macro_f1": 0.4384314318085172,
|
| 57 |
+
"eval_runtime": 11.986,
|
| 58 |
+
"eval_samples_per_second": 404.305,
|
| 59 |
+
"eval_steps_per_second": 1.585,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6921,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6907,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6922,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6913,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6909,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6908,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6899,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6915,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5371440363186133,
|
| 113 |
+
"eval_loss": 0.6897919774055481,
|
| 114 |
+
"eval_macro_f1": 0.5121441549256929,
|
| 115 |
+
"eval_runtime": 11.826,
|
| 116 |
+
"eval_samples_per_second": 409.777,
|
| 117 |
+
"eval_steps_per_second": 1.607,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6894,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6878,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6904,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.6898,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6891,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6888,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6867,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5489063144861742,
|
| 165 |
+
"eval_loss": 0.6865408420562744,
|
| 166 |
+
"eval_macro_f1": 0.5001134416030564,
|
| 167 |
+
"eval_runtime": 11.6292,
|
| 168 |
+
"eval_samples_per_second": 416.711,
|
| 169 |
+
"eval_steps_per_second": 1.634,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6866,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6857,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6844,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.686,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6826,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6812,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.683,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6796,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5631448617416426,
|
| 223 |
+
"eval_loss": 0.6797442436218262,
|
| 224 |
+
"eval_macro_f1": 0.5293249340542198,
|
| 225 |
+
"eval_runtime": 11.7067,
|
| 226 |
+
"eval_samples_per_second": 413.951,
|
| 227 |
+
"eval_steps_per_second": 1.623,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6752,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6719,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6706,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.6643,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6715,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6626,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6565,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6501,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6114321089558399,
|
| 281 |
+
"eval_loss": 0.6514652371406555,
|
| 282 |
+
"eval_macro_f1": 0.6112632471106266,
|
| 283 |
+
"eval_runtime": 11.451,
|
| 284 |
+
"eval_samples_per_second": 423.194,
|
| 285 |
+
"eval_steps_per_second": 1.659,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6226,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5995,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.574,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5463,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.5097,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.5061,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4848,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.756706562113083,
|
| 333 |
+
"eval_loss": 0.48308002948760986,
|
| 334 |
+
"eval_macro_f1": 0.7565062299709837,
|
| 335 |
+
"eval_runtime": 11.7207,
|
| 336 |
+
"eval_samples_per_second": 413.456,
|
| 337 |
+
"eval_steps_per_second": 1.621,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4526,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4313,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.421,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4192,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4168,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4101,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.4056,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3897,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7820883202641353,
|
| 391 |
+
"eval_loss": 0.4393594563007355,
|
| 392 |
+
"eval_macro_f1": 0.7806234004518389,
|
| 393 |
+
"eval_runtime": 11.5988,
|
| 394 |
+
"eval_samples_per_second": 417.802,
|
| 395 |
+
"eval_steps_per_second": 1.638,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3638,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.3543,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3564,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.358,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3522,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3441,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.356,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7899298390425092,
|
| 443 |
+
"eval_loss": 0.42679837346076965,
|
| 444 |
+
"eval_macro_f1": 0.7898254485802946,
|
| 445 |
+
"eval_runtime": 11.8329,
|
| 446 |
+
"eval_samples_per_second": 409.537,
|
| 447 |
+
"eval_steps_per_second": 1.606,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.335,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.3262,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.299,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.324,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.2908,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.3002,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.3194,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.3309,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7787866281469253,
|
| 501 |
+
"eval_loss": 0.4634370803833008,
|
| 502 |
+
"eval_macro_f1": 0.7780875065122488,
|
| 503 |
+
"eval_runtime": 11.8168,
|
| 504 |
+
"eval_samples_per_second": 410.094,
|
| 505 |
+
"eval_steps_per_second": 1.608,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2966,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2911,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2844,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.281,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2826,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.262,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.2622,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.2888,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.8014857614527445,
|
| 559 |
+
"eval_loss": 0.4261420667171478,
|
| 560 |
+
"eval_macro_f1": 0.8011263890026483,
|
| 561 |
+
"eval_runtime": 11.9419,
|
| 562 |
+
"eval_samples_per_second": 405.8,
|
| 563 |
+
"eval_steps_per_second": 1.591,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2598,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2391,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.2392,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2447,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2499,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.249,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2644,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.811803549319026,
|
| 611 |
+
"eval_loss": 0.4315750300884247,
|
| 612 |
+
"eval_macro_f1": 0.8112433490703433,
|
| 613 |
+
"eval_runtime": 12.076,
|
| 614 |
+
"eval_samples_per_second": 401.293,
|
| 615 |
+
"eval_steps_per_second": 1.573,
|
| 616 |
+
"step": 836
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 11.05,
|
| 620 |
+
"learning_rate": 4.484848484848485e-05,
|
| 621 |
+
"loss": 0.2451,
|
| 622 |
+
"step": 840
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 11.18,
|
| 626 |
+
"learning_rate": 4.46969696969697e-05,
|
| 627 |
+
"loss": 0.2172,
|
| 628 |
+
"step": 850
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 11.32,
|
| 632 |
+
"learning_rate": 4.454545454545455e-05,
|
| 633 |
+
"loss": 0.2152,
|
| 634 |
+
"step": 860
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 11.45,
|
| 638 |
+
"learning_rate": 4.43939393939394e-05,
|
| 639 |
+
"loss": 0.2179,
|
| 640 |
+
"step": 870
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 11.58,
|
| 644 |
+
"learning_rate": 4.4242424242424246e-05,
|
| 645 |
+
"loss": 0.1971,
|
| 646 |
+
"step": 880
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 11.71,
|
| 650 |
+
"learning_rate": 4.409090909090909e-05,
|
| 651 |
+
"loss": 0.2053,
|
| 652 |
+
"step": 890
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 11.84,
|
| 656 |
+
"learning_rate": 4.3939393939393944e-05,
|
| 657 |
+
"loss": 0.2339,
|
| 658 |
+
"step": 900
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 11.97,
|
| 662 |
+
"learning_rate": 4.378787878787879e-05,
|
| 663 |
+
"loss": 0.2393,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 12.0,
|
| 668 |
+
"eval_accuracy": 0.8136607511349566,
|
| 669 |
+
"eval_loss": 0.45733141899108887,
|
| 670 |
+
"eval_macro_f1": 0.8125261148695674,
|
| 671 |
+
"eval_runtime": 12.2485,
|
| 672 |
+
"eval_samples_per_second": 395.64,
|
| 673 |
+
"eval_steps_per_second": 1.551,
|
| 674 |
+
"step": 912
|
| 675 |
+
}
|
| 676 |
+
],
|
| 677 |
+
"max_steps": 3800,
|
| 678 |
+
"num_train_epochs": 50,
|
| 679 |
+
"total_flos": 3204379461058560.0,
|
| 680 |
+
"trial_name": null,
|
| 681 |
+
"trial_params": null
|
| 682 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold2/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold2/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8977777777777778,
|
| 3 |
+
"test_loss": 0.2747178077697754,
|
| 4 |
+
"test_macro_f1": 0.8641054594154565,
|
| 5 |
+
"test_runtime": 2.2566,
|
| 6 |
+
"test_samples_per_second": 398.823,
|
| 7 |
+
"test_steps_per_second": 1.773
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold4/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Endotheliums/fold4/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8866666666666667,
|
| 3 |
+
"test_loss": 0.28596895933151245,
|
| 4 |
+
"test_macro_f1": 0.8465909090909092,
|
| 5 |
+
"test_runtime": 2.26,
|
| 6 |
+
"test_samples_per_second": 398.23,
|
| 7 |
+
"test_steps_per_second": 1.77
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold4/checkpoint-836/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold4/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Endotheliums/fold4/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8866666666666667,
|
| 3 |
+
"test_loss": 0.28596895933151245,
|
| 4 |
+
"test_macro_f1": 0.8465909090909092,
|
| 5 |
+
"test_runtime": 2.26,
|
| 6 |
+
"test_samples_per_second": 398.23,
|
| 7 |
+
"test_steps_per_second": 1.77
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold0/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold0/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7411111111111112,
|
| 3 |
+
"test_loss": 0.5714522004127502,
|
| 4 |
+
"test_macro_f1": 0.7196520316203632,
|
| 5 |
+
"test_runtime": 2.1462,
|
| 6 |
+
"test_samples_per_second": 419.345,
|
| 7 |
+
"test_steps_per_second": 1.864
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold0/checkpoint-836/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold0/checkpoint-836/trainer_state.json
ADDED
|
@@ -0,0 +1,624 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4367808401584625,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Epitheliums/fold0/checkpoint-684",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 836,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6923,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6918,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6929,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6934,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6947,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6927,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6926,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.4943263874561584,
|
| 55 |
+
"eval_loss": 0.693455696105957,
|
| 56 |
+
"eval_macro_f1": 0.4884385137171501,
|
| 57 |
+
"eval_runtime": 11.2302,
|
| 58 |
+
"eval_samples_per_second": 431.603,
|
| 59 |
+
"eval_steps_per_second": 1.692,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6933,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6925,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6938,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6889,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6938,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6899,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6906,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6925,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5207344749329482,
|
| 113 |
+
"eval_loss": 0.6913900375366211,
|
| 114 |
+
"eval_macro_f1": 0.39379505149299676,
|
| 115 |
+
"eval_runtime": 11.9642,
|
| 116 |
+
"eval_samples_per_second": 405.124,
|
| 117 |
+
"eval_steps_per_second": 1.588,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6904,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6911,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6888,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.692,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6904,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6881,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6873,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5409531669073654,
|
| 165 |
+
"eval_loss": 0.6888494491577148,
|
| 166 |
+
"eval_macro_f1": 0.461880985936752,
|
| 167 |
+
"eval_runtime": 11.5789,
|
| 168 |
+
"eval_samples_per_second": 418.608,
|
| 169 |
+
"eval_steps_per_second": 1.641,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6882,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6848,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6853,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.6852,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6852,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6849,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6825,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6813,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5562203424798845,
|
| 223 |
+
"eval_loss": 0.6827021837234497,
|
| 224 |
+
"eval_macro_f1": 0.5513306517049066,
|
| 225 |
+
"eval_runtime": 11.2305,
|
| 226 |
+
"eval_samples_per_second": 431.593,
|
| 227 |
+
"eval_steps_per_second": 1.692,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6798,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6775,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6719,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.665,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6616,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6576,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6465,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6383,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6249226325562204,
|
| 281 |
+
"eval_loss": 0.6389692425727844,
|
| 282 |
+
"eval_macro_f1": 0.6159466513902696,
|
| 283 |
+
"eval_runtime": 12.0575,
|
| 284 |
+
"eval_samples_per_second": 401.989,
|
| 285 |
+
"eval_steps_per_second": 1.576,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6047,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5807,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5336,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5182,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.4865,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.4959,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4698,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7633587786259542,
|
| 333 |
+
"eval_loss": 0.49028173089027405,
|
| 334 |
+
"eval_macro_f1": 0.7568335054238134,
|
| 335 |
+
"eval_runtime": 10.8558,
|
| 336 |
+
"eval_samples_per_second": 446.489,
|
| 337 |
+
"eval_steps_per_second": 1.75,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4584,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4042,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4086,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4152,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4426,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4064,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.387,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3871,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7829585310501341,
|
| 391 |
+
"eval_loss": 0.45330119132995605,
|
| 392 |
+
"eval_macro_f1": 0.7786660383654955,
|
| 393 |
+
"eval_runtime": 11.7864,
|
| 394 |
+
"eval_samples_per_second": 411.236,
|
| 395 |
+
"eval_steps_per_second": 1.612,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3631,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.336,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3518,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.3532,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3533,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3512,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.351,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7914173715700433,
|
| 443 |
+
"eval_loss": 0.4372871518135071,
|
| 444 |
+
"eval_macro_f1": 0.7914172295164006,
|
| 445 |
+
"eval_runtime": 11.8298,
|
| 446 |
+
"eval_samples_per_second": 409.726,
|
| 447 |
+
"eval_steps_per_second": 1.606,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3218,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.323,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.3241,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.2956,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.3094,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.2916,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.2983,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.312,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7963688879719414,
|
| 501 |
+
"eval_loss": 0.4367808401584625,
|
| 502 |
+
"eval_macro_f1": 0.7962561816949302,
|
| 503 |
+
"eval_runtime": 11.4634,
|
| 504 |
+
"eval_samples_per_second": 422.825,
|
| 505 |
+
"eval_steps_per_second": 1.657,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2726,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2738,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2666,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2609,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2872,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2827,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.266,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.274,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.7990509593563029,
|
| 559 |
+
"eval_loss": 0.43979689478874207,
|
| 560 |
+
"eval_macro_f1": 0.7988178222756628,
|
| 561 |
+
"eval_runtime": 11.8441,
|
| 562 |
+
"eval_samples_per_second": 409.234,
|
| 563 |
+
"eval_steps_per_second": 1.604,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2323,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2351,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.224,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2463,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2758,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.3099,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2533,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.7854342892510832,
|
| 611 |
+
"eval_loss": 0.4947068691253662,
|
| 612 |
+
"eval_macro_f1": 0.7851853543805845,
|
| 613 |
+
"eval_runtime": 11.9238,
|
| 614 |
+
"eval_samples_per_second": 406.498,
|
| 615 |
+
"eval_steps_per_second": 1.593,
|
| 616 |
+
"step": 836
|
| 617 |
+
}
|
| 618 |
+
],
|
| 619 |
+
"max_steps": 3800,
|
| 620 |
+
"num_train_epochs": 50,
|
| 621 |
+
"total_flos": 2937196312461312.0,
|
| 622 |
+
"trial_name": null,
|
| 623 |
+
"trial_params": null
|
| 624 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold0/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold0/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7411111111111112,
|
| 3 |
+
"test_loss": 0.5714522004127502,
|
| 4 |
+
"test_macro_f1": 0.7196520316203632,
|
| 5 |
+
"test_runtime": 2.1462,
|
| 6 |
+
"test_samples_per_second": 419.345,
|
| 7 |
+
"test_steps_per_second": 1.864
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold1/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold1/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.6855555555555556,
|
| 3 |
+
"test_loss": 0.6318806409835815,
|
| 4 |
+
"test_macro_f1": 0.6547091463439433,
|
| 5 |
+
"test_runtime": 2.2117,
|
| 6 |
+
"test_samples_per_second": 406.921,
|
| 7 |
+
"test_steps_per_second": 1.809
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold1/checkpoint-760/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold1/checkpoint-760/trainer_state.json
ADDED
|
@@ -0,0 +1,572 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4305054545402527,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Epitheliums/fold1/checkpoint-608",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 760,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6965,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6949,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6928,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.692,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6927,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6941,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6915,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.5152703260420965,
|
| 55 |
+
"eval_loss": 0.6928699612617493,
|
| 56 |
+
"eval_macro_f1": 0.39271132206137227,
|
| 57 |
+
"eval_runtime": 11.4979,
|
| 58 |
+
"eval_samples_per_second": 421.469,
|
| 59 |
+
"eval_steps_per_second": 1.652,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6939,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6926,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6922,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6931,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6906,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6918,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6913,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6894,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5181593066446554,
|
| 113 |
+
"eval_loss": 0.6917251944541931,
|
| 114 |
+
"eval_macro_f1": 0.392353250840646,
|
| 115 |
+
"eval_runtime": 12.1114,
|
| 116 |
+
"eval_samples_per_second": 400.118,
|
| 117 |
+
"eval_steps_per_second": 1.569,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6901,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6874,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6898,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.6901,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.69,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6888,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6907,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.536731324803962,
|
| 165 |
+
"eval_loss": 0.6895701885223389,
|
| 166 |
+
"eval_macro_f1": 0.5327198581400359,
|
| 167 |
+
"eval_runtime": 11.3292,
|
| 168 |
+
"eval_samples_per_second": 427.743,
|
| 169 |
+
"eval_steps_per_second": 1.677,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6893,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6851,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6853,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.6861,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6829,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6831,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6826,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6827,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5691291787040859,
|
| 223 |
+
"eval_loss": 0.6838498711585999,
|
| 224 |
+
"eval_macro_f1": 0.5376752831259407,
|
| 225 |
+
"eval_runtime": 11.9548,
|
| 226 |
+
"eval_samples_per_second": 405.361,
|
| 227 |
+
"eval_steps_per_second": 1.589,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6783,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6759,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6725,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.6684,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6616,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6635,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.644,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.633,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.609162195625258,
|
| 281 |
+
"eval_loss": 0.6557519435882568,
|
| 282 |
+
"eval_macro_f1": 0.5843500502222185,
|
| 283 |
+
"eval_runtime": 11.6048,
|
| 284 |
+
"eval_samples_per_second": 417.584,
|
| 285 |
+
"eval_steps_per_second": 1.637,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6185,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.6023,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5674,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5312,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.5049,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.4902,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.473,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7548493602971523,
|
| 333 |
+
"eval_loss": 0.4889010190963745,
|
| 334 |
+
"eval_macro_f1": 0.7548232595790609,
|
| 335 |
+
"eval_runtime": 11.6506,
|
| 336 |
+
"eval_samples_per_second": 415.945,
|
| 337 |
+
"eval_steps_per_second": 1.631,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4607,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4059,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4062,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4139,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4035,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4155,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.4073,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.4093,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7851836566240198,
|
| 391 |
+
"eval_loss": 0.43714773654937744,
|
| 392 |
+
"eval_macro_f1": 0.7847112606262489,
|
| 393 |
+
"eval_runtime": 11.582,
|
| 394 |
+
"eval_samples_per_second": 418.408,
|
| 395 |
+
"eval_steps_per_second": 1.64,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3596,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.3473,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3301,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.337,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3312,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3365,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.3324,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7897234832851836,
|
| 443 |
+
"eval_loss": 0.4305054545402527,
|
| 444 |
+
"eval_macro_f1": 0.7895402204646096,
|
| 445 |
+
"eval_runtime": 11.7911,
|
| 446 |
+
"eval_samples_per_second": 410.989,
|
| 447 |
+
"eval_steps_per_second": 1.611,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3291,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.3035,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.2778,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.2851,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.3087,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.2932,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.2988,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.3107,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7909616178291374,
|
| 501 |
+
"eval_loss": 0.44929826259613037,
|
| 502 |
+
"eval_macro_f1": 0.7908438302377228,
|
| 503 |
+
"eval_runtime": 11.5886,
|
| 504 |
+
"eval_samples_per_second": 418.171,
|
| 505 |
+
"eval_steps_per_second": 1.64,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2866,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2656,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.261,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2708,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2757,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2604,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.2561,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.2514,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.8037556747833264,
|
| 559 |
+
"eval_loss": 0.44588491320610046,
|
| 560 |
+
"eval_macro_f1": 0.8027325242815069,
|
| 561 |
+
"eval_runtime": 11.6074,
|
| 562 |
+
"eval_samples_per_second": 417.491,
|
| 563 |
+
"eval_steps_per_second": 1.637,
|
| 564 |
+
"step": 760
|
| 565 |
+
}
|
| 566 |
+
],
|
| 567 |
+
"max_steps": 3800,
|
| 568 |
+
"num_train_epochs": 50,
|
| 569 |
+
"total_flos": 2670316217548800.0,
|
| 570 |
+
"trial_name": null,
|
| 571 |
+
"trial_params": null
|
| 572 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold1/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold1/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.6855555555555556,
|
| 3 |
+
"test_loss": 0.6318806409835815,
|
| 4 |
+
"test_macro_f1": 0.6547091463439433,
|
| 5 |
+
"test_runtime": 2.2117,
|
| 6 |
+
"test_samples_per_second": 406.921,
|
| 7 |
+
"test_steps_per_second": 1.809
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold2/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold2/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.76,
|
| 3 |
+
"test_loss": 0.5362856984138489,
|
| 4 |
+
"test_macro_f1": 0.7385114522299162,
|
| 5 |
+
"test_runtime": 2.329,
|
| 6 |
+
"test_samples_per_second": 386.44,
|
| 7 |
+
"test_steps_per_second": 1.718
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold2/checkpoint-912/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold2/checkpoint-912/trainer_state.json
ADDED
|
@@ -0,0 +1,682 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4261420667171478,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Epitheliums/fold2/checkpoint-760",
|
| 4 |
+
"epoch": 12.0,
|
| 5 |
+
"global_step": 912,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6971,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6973,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6956,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6955,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6936,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6925,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6899,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.5309533635988444,
|
| 55 |
+
"eval_loss": 0.6913440823554993,
|
| 56 |
+
"eval_macro_f1": 0.4384314318085172,
|
| 57 |
+
"eval_runtime": 11.6766,
|
| 58 |
+
"eval_samples_per_second": 415.019,
|
| 59 |
+
"eval_steps_per_second": 1.627,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6921,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6907,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6922,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6913,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6909,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6908,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6899,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6915,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5371440363186133,
|
| 113 |
+
"eval_loss": 0.6897919774055481,
|
| 114 |
+
"eval_macro_f1": 0.5121441549256929,
|
| 115 |
+
"eval_runtime": 12.1813,
|
| 116 |
+
"eval_samples_per_second": 397.823,
|
| 117 |
+
"eval_steps_per_second": 1.56,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6894,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6878,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6904,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.6898,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6891,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6888,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6867,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5489063144861742,
|
| 165 |
+
"eval_loss": 0.6865408420562744,
|
| 166 |
+
"eval_macro_f1": 0.5001134416030564,
|
| 167 |
+
"eval_runtime": 11.8908,
|
| 168 |
+
"eval_samples_per_second": 407.543,
|
| 169 |
+
"eval_steps_per_second": 1.598,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6866,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6857,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6844,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.686,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6826,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6812,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.683,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6796,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5631448617416426,
|
| 223 |
+
"eval_loss": 0.6797442436218262,
|
| 224 |
+
"eval_macro_f1": 0.5293249340542198,
|
| 225 |
+
"eval_runtime": 11.9104,
|
| 226 |
+
"eval_samples_per_second": 406.87,
|
| 227 |
+
"eval_steps_per_second": 1.595,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6752,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6719,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6706,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.6643,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6715,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6626,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6565,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6501,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6114321089558399,
|
| 281 |
+
"eval_loss": 0.6514652371406555,
|
| 282 |
+
"eval_macro_f1": 0.6112632471106266,
|
| 283 |
+
"eval_runtime": 11.4625,
|
| 284 |
+
"eval_samples_per_second": 422.771,
|
| 285 |
+
"eval_steps_per_second": 1.658,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6226,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5995,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.574,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5463,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.5097,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.5061,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4848,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.756706562113083,
|
| 333 |
+
"eval_loss": 0.48308002948760986,
|
| 334 |
+
"eval_macro_f1": 0.7565062299709837,
|
| 335 |
+
"eval_runtime": 11.4473,
|
| 336 |
+
"eval_samples_per_second": 423.33,
|
| 337 |
+
"eval_steps_per_second": 1.66,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4526,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4313,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.421,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4192,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4168,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4101,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.4056,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3897,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7820883202641353,
|
| 391 |
+
"eval_loss": 0.4393594563007355,
|
| 392 |
+
"eval_macro_f1": 0.7806234004518389,
|
| 393 |
+
"eval_runtime": 11.9011,
|
| 394 |
+
"eval_samples_per_second": 407.191,
|
| 395 |
+
"eval_steps_per_second": 1.596,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3638,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.3543,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3564,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.358,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3522,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3441,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.356,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7899298390425092,
|
| 443 |
+
"eval_loss": 0.42679837346076965,
|
| 444 |
+
"eval_macro_f1": 0.7898254485802946,
|
| 445 |
+
"eval_runtime": 11.9427,
|
| 446 |
+
"eval_samples_per_second": 405.772,
|
| 447 |
+
"eval_steps_per_second": 1.591,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.335,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.3262,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.299,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.324,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.2908,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.3002,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.3194,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.3309,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7787866281469253,
|
| 501 |
+
"eval_loss": 0.4634370803833008,
|
| 502 |
+
"eval_macro_f1": 0.7780875065122488,
|
| 503 |
+
"eval_runtime": 11.9406,
|
| 504 |
+
"eval_samples_per_second": 405.841,
|
| 505 |
+
"eval_steps_per_second": 1.591,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2966,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2911,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2844,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.281,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2826,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.262,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.2622,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.2888,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.8014857614527445,
|
| 559 |
+
"eval_loss": 0.4261420667171478,
|
| 560 |
+
"eval_macro_f1": 0.8011263890026483,
|
| 561 |
+
"eval_runtime": 11.7787,
|
| 562 |
+
"eval_samples_per_second": 411.419,
|
| 563 |
+
"eval_steps_per_second": 1.613,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2598,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2391,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.2392,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2447,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2499,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.249,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2644,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.811803549319026,
|
| 611 |
+
"eval_loss": 0.4315750300884247,
|
| 612 |
+
"eval_macro_f1": 0.8112433490703433,
|
| 613 |
+
"eval_runtime": 11.9925,
|
| 614 |
+
"eval_samples_per_second": 404.087,
|
| 615 |
+
"eval_steps_per_second": 1.584,
|
| 616 |
+
"step": 836
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 11.05,
|
| 620 |
+
"learning_rate": 4.484848484848485e-05,
|
| 621 |
+
"loss": 0.2451,
|
| 622 |
+
"step": 840
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 11.18,
|
| 626 |
+
"learning_rate": 4.46969696969697e-05,
|
| 627 |
+
"loss": 0.2172,
|
| 628 |
+
"step": 850
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 11.32,
|
| 632 |
+
"learning_rate": 4.454545454545455e-05,
|
| 633 |
+
"loss": 0.2152,
|
| 634 |
+
"step": 860
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 11.45,
|
| 638 |
+
"learning_rate": 4.43939393939394e-05,
|
| 639 |
+
"loss": 0.2179,
|
| 640 |
+
"step": 870
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 11.58,
|
| 644 |
+
"learning_rate": 4.4242424242424246e-05,
|
| 645 |
+
"loss": 0.1971,
|
| 646 |
+
"step": 880
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 11.71,
|
| 650 |
+
"learning_rate": 4.409090909090909e-05,
|
| 651 |
+
"loss": 0.2053,
|
| 652 |
+
"step": 890
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 11.84,
|
| 656 |
+
"learning_rate": 4.3939393939393944e-05,
|
| 657 |
+
"loss": 0.2339,
|
| 658 |
+
"step": 900
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 11.97,
|
| 662 |
+
"learning_rate": 4.378787878787879e-05,
|
| 663 |
+
"loss": 0.2393,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 12.0,
|
| 668 |
+
"eval_accuracy": 0.8136607511349566,
|
| 669 |
+
"eval_loss": 0.45733141899108887,
|
| 670 |
+
"eval_macro_f1": 0.8125261148695674,
|
| 671 |
+
"eval_runtime": 11.5055,
|
| 672 |
+
"eval_samples_per_second": 421.191,
|
| 673 |
+
"eval_steps_per_second": 1.651,
|
| 674 |
+
"step": 912
|
| 675 |
+
}
|
| 676 |
+
],
|
| 677 |
+
"max_steps": 3800,
|
| 678 |
+
"num_train_epochs": 50,
|
| 679 |
+
"total_flos": 3204379461058560.0,
|
| 680 |
+
"trial_name": null,
|
| 681 |
+
"trial_params": null
|
| 682 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold2/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold2/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.76,
|
| 3 |
+
"test_loss": 0.5362856984138489,
|
| 4 |
+
"test_macro_f1": 0.7385114522299162,
|
| 5 |
+
"test_runtime": 2.329,
|
| 6 |
+
"test_samples_per_second": 386.44,
|
| 7 |
+
"test_steps_per_second": 1.718
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold3/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold3/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7144444444444444,
|
| 3 |
+
"test_loss": 0.5890923142433167,
|
| 4 |
+
"test_macro_f1": 0.6757229639453846,
|
| 5 |
+
"test_runtime": 2.1822,
|
| 6 |
+
"test_samples_per_second": 412.429,
|
| 7 |
+
"test_steps_per_second": 1.833
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold3/checkpoint-760/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold3/checkpoint-760/trainer_state.json
ADDED
|
@@ -0,0 +1,572 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.41204845905303955,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Epitheliums/fold3/checkpoint-608",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 760,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.693,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6944,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.693,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6929,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.693,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6938,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.535080478745357,
|
| 55 |
+
"eval_loss": 0.6914414763450623,
|
| 56 |
+
"eval_macro_f1": 0.478596030858335,
|
| 57 |
+
"eval_runtime": 11.7125,
|
| 58 |
+
"eval_samples_per_second": 413.747,
|
| 59 |
+
"eval_steps_per_second": 1.622,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6918,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6935,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6916,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6912,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6914,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.693,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6931,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6924,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5445728435823359,
|
| 113 |
+
"eval_loss": 0.6899434924125671,
|
| 114 |
+
"eval_macro_f1": 0.5103583774116406,
|
| 115 |
+
"eval_runtime": 11.8045,
|
| 116 |
+
"eval_samples_per_second": 410.52,
|
| 117 |
+
"eval_steps_per_second": 1.61,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6907,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6912,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6894,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.6881,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6927,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6881,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.688,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5493190260008254,
|
| 165 |
+
"eval_loss": 0.6871300339698792,
|
| 166 |
+
"eval_macro_f1": 0.48504931597548007,
|
| 167 |
+
"eval_runtime": 11.8167,
|
| 168 |
+
"eval_samples_per_second": 410.097,
|
| 169 |
+
"eval_steps_per_second": 1.608,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.687,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6881,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6841,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.6844,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6826,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6856,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6838,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6808,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5806851011143211,
|
| 223 |
+
"eval_loss": 0.6803114414215088,
|
| 224 |
+
"eval_macro_f1": 0.5563026906415562,
|
| 225 |
+
"eval_runtime": 11.7867,
|
| 226 |
+
"eval_samples_per_second": 411.143,
|
| 227 |
+
"eval_steps_per_second": 1.612,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6766,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6755,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6755,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.6672,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6567,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6572,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6521,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.64,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6217498968221213,
|
| 281 |
+
"eval_loss": 0.6434289216995239,
|
| 282 |
+
"eval_macro_f1": 0.6187340047656543,
|
| 283 |
+
"eval_runtime": 11.6131,
|
| 284 |
+
"eval_samples_per_second": 417.287,
|
| 285 |
+
"eval_steps_per_second": 1.636,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6148,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.592,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5623,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5381,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.5118,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.5057,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4721,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7294676021460998,
|
| 333 |
+
"eval_loss": 0.5136769413948059,
|
| 334 |
+
"eval_macro_f1": 0.7281372017684267,
|
| 335 |
+
"eval_runtime": 11.5552,
|
| 336 |
+
"eval_samples_per_second": 419.377,
|
| 337 |
+
"eval_steps_per_second": 1.644,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4735,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4308,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4256,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4079,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.3995,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4118,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.4,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3843,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7886917044985555,
|
| 391 |
+
"eval_loss": 0.4303540289402008,
|
| 392 |
+
"eval_macro_f1": 0.7881095852975336,
|
| 393 |
+
"eval_runtime": 11.7331,
|
| 394 |
+
"eval_samples_per_second": 413.02,
|
| 395 |
+
"eval_steps_per_second": 1.619,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3576,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.3464,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3473,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.341,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3303,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.33,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.349,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.8049938093272803,
|
| 443 |
+
"eval_loss": 0.41204845905303955,
|
| 444 |
+
"eval_macro_f1": 0.8045103153020186,
|
| 445 |
+
"eval_runtime": 11.3909,
|
| 446 |
+
"eval_samples_per_second": 425.426,
|
| 447 |
+
"eval_steps_per_second": 1.668,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3367,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.2946,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.2885,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.2712,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.2996,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.283,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.2792,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.3107,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.8041683862979777,
|
| 501 |
+
"eval_loss": 0.42064493894577026,
|
| 502 |
+
"eval_macro_f1": 0.8041309451967903,
|
| 503 |
+
"eval_runtime": 11.8705,
|
| 504 |
+
"eval_samples_per_second": 408.238,
|
| 505 |
+
"eval_steps_per_second": 1.601,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.279,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2673,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2442,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2527,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2513,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2798,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.2795,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.2656,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.7938505984316963,
|
| 559 |
+
"eval_loss": 0.4743572771549225,
|
| 560 |
+
"eval_macro_f1": 0.7933136421849105,
|
| 561 |
+
"eval_runtime": 11.9461,
|
| 562 |
+
"eval_samples_per_second": 405.655,
|
| 563 |
+
"eval_steps_per_second": 1.59,
|
| 564 |
+
"step": 760
|
| 565 |
+
}
|
| 566 |
+
],
|
| 567 |
+
"max_steps": 3800,
|
| 568 |
+
"num_train_epochs": 50,
|
| 569 |
+
"total_flos": 2670316217548800.0,
|
| 570 |
+
"trial_name": null,
|
| 571 |
+
"trial_params": null
|
| 572 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold3/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold3/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7144444444444444,
|
| 3 |
+
"test_loss": 0.5890923142433167,
|
| 4 |
+
"test_macro_f1": 0.6757229639453846,
|
| 5 |
+
"test_runtime": 2.1822,
|
| 6 |
+
"test_samples_per_second": 412.429,
|
| 7 |
+
"test_steps_per_second": 1.833
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold4/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Epitheliums/fold4/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7222222222222222,
|
| 3 |
+
"test_loss": 0.6258071660995483,
|
| 4 |
+
"test_macro_f1": 0.6596087164374194,
|
| 5 |
+
"test_runtime": 2.1946,
|
| 6 |
+
"test_samples_per_second": 410.102,
|
| 7 |
+
"test_steps_per_second": 1.823
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold4/checkpoint-836/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold4/checkpoint-836/trainer_state.json
ADDED
|
@@ -0,0 +1,624 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4270257353782654,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Epitheliums/fold4/checkpoint-684",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 836,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6987,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6959,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6954,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6934,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6922,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6934,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6935,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.5305406520841931,
|
| 55 |
+
"eval_loss": 0.6920099258422852,
|
| 56 |
+
"eval_macro_f1": 0.4685555933322892,
|
| 57 |
+
"eval_runtime": 11.7279,
|
| 58 |
+
"eval_samples_per_second": 413.202,
|
| 59 |
+
"eval_steps_per_second": 1.62,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6914,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6916,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6927,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6911,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6894,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6898,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6917,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6922,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5264135369376806,
|
| 113 |
+
"eval_loss": 0.6905979514122009,
|
| 114 |
+
"eval_macro_f1": 0.5149496323187399,
|
| 115 |
+
"eval_runtime": 11.6522,
|
| 116 |
+
"eval_samples_per_second": 415.887,
|
| 117 |
+
"eval_steps_per_second": 1.631,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6894,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6902,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6898,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.6885,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6863,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6881,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6863,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5491126702434997,
|
| 165 |
+
"eval_loss": 0.6875786781311035,
|
| 166 |
+
"eval_macro_f1": 0.5392500816434073,
|
| 167 |
+
"eval_runtime": 11.7732,
|
| 168 |
+
"eval_samples_per_second": 411.611,
|
| 169 |
+
"eval_steps_per_second": 1.614,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6889,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6849,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6842,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.685,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6804,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6807,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6804,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6803,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5652084193148988,
|
| 223 |
+
"eval_loss": 0.6798397302627563,
|
| 224 |
+
"eval_macro_f1": 0.5503469219825736,
|
| 225 |
+
"eval_runtime": 11.5469,
|
| 226 |
+
"eval_samples_per_second": 419.68,
|
| 227 |
+
"eval_steps_per_second": 1.645,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6776,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6733,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6733,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.665,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6586,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6521,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6523,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6455,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6209244737928188,
|
| 281 |
+
"eval_loss": 0.6445870995521545,
|
| 282 |
+
"eval_macro_f1": 0.6018380896855913,
|
| 283 |
+
"eval_runtime": 11.7694,
|
| 284 |
+
"eval_samples_per_second": 411.746,
|
| 285 |
+
"eval_steps_per_second": 1.614,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6041,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5997,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5634,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.55,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.5074,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.5078,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.498,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7484523318200578,
|
| 333 |
+
"eval_loss": 0.5012898445129395,
|
| 334 |
+
"eval_macro_f1": 0.7442249898216486,
|
| 335 |
+
"eval_runtime": 11.6672,
|
| 336 |
+
"eval_samples_per_second": 415.353,
|
| 337 |
+
"eval_steps_per_second": 1.628,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4725,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4247,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4162,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4376,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4188,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4139,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.3968,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.4035,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7773421378456459,
|
| 391 |
+
"eval_loss": 0.4468770921230316,
|
| 392 |
+
"eval_macro_f1": 0.7760577346542106,
|
| 393 |
+
"eval_runtime": 11.865,
|
| 394 |
+
"eval_samples_per_second": 408.427,
|
| 395 |
+
"eval_steps_per_second": 1.601,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3567,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.388,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3683,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.3446,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3568,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3595,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.3475,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7868345026826249,
|
| 443 |
+
"eval_loss": 0.4375980496406555,
|
| 444 |
+
"eval_macro_f1": 0.7867887348838893,
|
| 445 |
+
"eval_runtime": 11.4687,
|
| 446 |
+
"eval_samples_per_second": 422.54,
|
| 447 |
+
"eval_steps_per_second": 1.657,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3535,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.2952,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.3154,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.3072,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.3198,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.321,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.3529,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.3068,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.8002476269087908,
|
| 501 |
+
"eval_loss": 0.4270257353782654,
|
| 502 |
+
"eval_macro_f1": 0.798478880984294,
|
| 503 |
+
"eval_runtime": 11.9284,
|
| 504 |
+
"eval_samples_per_second": 406.258,
|
| 505 |
+
"eval_steps_per_second": 1.593,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2903,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2737,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.3023,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2687,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2716,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2902,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.2711,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.2763,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.7971522905489064,
|
| 559 |
+
"eval_loss": 0.4324711263179779,
|
| 560 |
+
"eval_macro_f1": 0.7968233053997001,
|
| 561 |
+
"eval_runtime": 11.3001,
|
| 562 |
+
"eval_samples_per_second": 428.846,
|
| 563 |
+
"eval_steps_per_second": 1.681,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2435,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2436,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.2447,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2504,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2526,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.2603,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2381,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.799009492364837,
|
| 611 |
+
"eval_loss": 0.44539201259613037,
|
| 612 |
+
"eval_macro_f1": 0.7990037065047797,
|
| 613 |
+
"eval_runtime": 11.6885,
|
| 614 |
+
"eval_samples_per_second": 414.595,
|
| 615 |
+
"eval_steps_per_second": 1.626,
|
| 616 |
+
"step": 836
|
| 617 |
+
}
|
| 618 |
+
],
|
| 619 |
+
"max_steps": 3800,
|
| 620 |
+
"num_train_epochs": 50,
|
| 621 |
+
"total_flos": 2937347839303680.0,
|
| 622 |
+
"trial_name": null,
|
| 623 |
+
"trial_params": null
|
| 624 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold4/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Epitheliums/fold4/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.7222222222222222,
|
| 3 |
+
"test_loss": 0.6258071660995483,
|
| 4 |
+
"test_macro_f1": 0.6596087164374194,
|
| 5 |
+
"test_runtime": 2.1946,
|
| 6 |
+
"test_samples_per_second": 410.102,
|
| 7 |
+
"test_steps_per_second": 1.823
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Macrophages/fold0/checkpoint-836/trainer_state.json
ADDED
|
@@ -0,0 +1,624 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4367808401584625,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Macrophages/fold0/checkpoint-684",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 836,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6923,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6918,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6929,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6934,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6947,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6927,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6926,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.4943263874561584,
|
| 55 |
+
"eval_loss": 0.693455696105957,
|
| 56 |
+
"eval_macro_f1": 0.4884385137171501,
|
| 57 |
+
"eval_runtime": 11.6053,
|
| 58 |
+
"eval_samples_per_second": 417.654,
|
| 59 |
+
"eval_steps_per_second": 1.637,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6933,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6925,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6938,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6889,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6938,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6899,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6906,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6925,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5207344749329482,
|
| 113 |
+
"eval_loss": 0.6913900375366211,
|
| 114 |
+
"eval_macro_f1": 0.39379505149299676,
|
| 115 |
+
"eval_runtime": 12.0893,
|
| 116 |
+
"eval_samples_per_second": 400.934,
|
| 117 |
+
"eval_steps_per_second": 1.572,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6904,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6911,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6888,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.692,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6904,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6881,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6873,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5409531669073654,
|
| 165 |
+
"eval_loss": 0.6888494491577148,
|
| 166 |
+
"eval_macro_f1": 0.461880985936752,
|
| 167 |
+
"eval_runtime": 11.9536,
|
| 168 |
+
"eval_samples_per_second": 405.486,
|
| 169 |
+
"eval_steps_per_second": 1.589,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6882,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6848,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6853,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.6852,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6852,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6849,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6825,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6813,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5562203424798845,
|
| 223 |
+
"eval_loss": 0.6827021837234497,
|
| 224 |
+
"eval_macro_f1": 0.5513306517049066,
|
| 225 |
+
"eval_runtime": 11.4766,
|
| 226 |
+
"eval_samples_per_second": 422.337,
|
| 227 |
+
"eval_steps_per_second": 1.656,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6798,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6775,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6719,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.665,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6616,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6576,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6465,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6383,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6249226325562204,
|
| 281 |
+
"eval_loss": 0.6389692425727844,
|
| 282 |
+
"eval_macro_f1": 0.6159466513902696,
|
| 283 |
+
"eval_runtime": 11.6555,
|
| 284 |
+
"eval_samples_per_second": 415.856,
|
| 285 |
+
"eval_steps_per_second": 1.63,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6047,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5807,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5336,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5182,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.4865,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.4959,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4698,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7633587786259542,
|
| 333 |
+
"eval_loss": 0.49028173089027405,
|
| 334 |
+
"eval_macro_f1": 0.7568335054238134,
|
| 335 |
+
"eval_runtime": 11.4376,
|
| 336 |
+
"eval_samples_per_second": 423.777,
|
| 337 |
+
"eval_steps_per_second": 1.661,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4584,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4042,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4086,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4152,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4426,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4064,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.387,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3871,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7829585310501341,
|
| 391 |
+
"eval_loss": 0.45330119132995605,
|
| 392 |
+
"eval_macro_f1": 0.7786660383654955,
|
| 393 |
+
"eval_runtime": 11.7816,
|
| 394 |
+
"eval_samples_per_second": 411.404,
|
| 395 |
+
"eval_steps_per_second": 1.613,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3631,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.336,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3518,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.3532,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3533,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3512,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.351,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7914173715700433,
|
| 443 |
+
"eval_loss": 0.4372871518135071,
|
| 444 |
+
"eval_macro_f1": 0.7914172295164006,
|
| 445 |
+
"eval_runtime": 11.5409,
|
| 446 |
+
"eval_samples_per_second": 419.986,
|
| 447 |
+
"eval_steps_per_second": 1.646,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3218,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.323,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.3241,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.2956,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.3094,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.2916,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.2983,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.312,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7963688879719414,
|
| 501 |
+
"eval_loss": 0.4367808401584625,
|
| 502 |
+
"eval_macro_f1": 0.7962561816949302,
|
| 503 |
+
"eval_runtime": 12.0535,
|
| 504 |
+
"eval_samples_per_second": 402.125,
|
| 505 |
+
"eval_steps_per_second": 1.576,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2726,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2738,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2666,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2609,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2872,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2827,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.266,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.274,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.7990509593563029,
|
| 559 |
+
"eval_loss": 0.43979689478874207,
|
| 560 |
+
"eval_macro_f1": 0.7988178222756628,
|
| 561 |
+
"eval_runtime": 11.8554,
|
| 562 |
+
"eval_samples_per_second": 408.843,
|
| 563 |
+
"eval_steps_per_second": 1.603,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2323,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2351,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.224,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2463,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2758,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.3099,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2533,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.7854342892510832,
|
| 611 |
+
"eval_loss": 0.4947068691253662,
|
| 612 |
+
"eval_macro_f1": 0.7851853543805845,
|
| 613 |
+
"eval_runtime": 11.7671,
|
| 614 |
+
"eval_samples_per_second": 411.909,
|
| 615 |
+
"eval_steps_per_second": 1.615,
|
| 616 |
+
"step": 836
|
| 617 |
+
}
|
| 618 |
+
],
|
| 619 |
+
"max_steps": 3800,
|
| 620 |
+
"num_train_epochs": 50,
|
| 621 |
+
"total_flos": 2937196312461312.0,
|
| 622 |
+
"trial_name": null,
|
| 623 |
+
"trial_params": null
|
| 624 |
+
}
|
cross_cell_type_generization/L2/Neutrophils/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Neutrophils/fold0/.DS_Store
ADDED
|
Binary file (6.15 kB). View file
|
|
|
cross_cell_type_generization/L2/Neutrophils/fold0/all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8855555555555555,
|
| 3 |
+
"test_loss": 0.2811035215854645,
|
| 4 |
+
"test_macro_f1": 0.8028125272542017,
|
| 5 |
+
"test_runtime": 2.1514,
|
| 6 |
+
"test_samples_per_second": 418.342,
|
| 7 |
+
"test_steps_per_second": 1.859
|
| 8 |
+
}
|
cross_cell_type_generization/L2/Neutrophils/fold0/checkpoint-836/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
cross_cell_type_generization/L2/Neutrophils/fold0/checkpoint-836/trainer_state.json
ADDED
|
@@ -0,0 +1,624 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.4367808401584625,
|
| 3 |
+
"best_model_checkpoint": "/vsphhome/fengguoqing/Geneformer/models/data_diversity/L2/Neutrophils/fold0/checkpoint-684",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 836,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.13,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 0.6923,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.26,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 0.6918,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.39,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 0.6929,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.53,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 0.6934,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.66,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 0.6947,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.79,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.6927,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.92,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.6926,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.0,
|
| 54 |
+
"eval_accuracy": 0.4943263874561584,
|
| 55 |
+
"eval_loss": 0.693455696105957,
|
| 56 |
+
"eval_macro_f1": 0.4884385137171501,
|
| 57 |
+
"eval_runtime": 11.6645,
|
| 58 |
+
"eval_samples_per_second": 415.534,
|
| 59 |
+
"eval_steps_per_second": 1.629,
|
| 60 |
+
"step": 76
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.05,
|
| 64 |
+
"learning_rate": 8.000000000000001e-06,
|
| 65 |
+
"loss": 0.6933,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.18,
|
| 70 |
+
"learning_rate": 9e-06,
|
| 71 |
+
"loss": 0.6925,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.32,
|
| 76 |
+
"learning_rate": 1e-05,
|
| 77 |
+
"loss": 0.6938,
|
| 78 |
+
"step": 100
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.45,
|
| 82 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 83 |
+
"loss": 0.6889,
|
| 84 |
+
"step": 110
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.58,
|
| 88 |
+
"learning_rate": 1.2e-05,
|
| 89 |
+
"loss": 0.6938,
|
| 90 |
+
"step": 120
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.71,
|
| 94 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 95 |
+
"loss": 0.6899,
|
| 96 |
+
"step": 130
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.84,
|
| 100 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 101 |
+
"loss": 0.6906,
|
| 102 |
+
"step": 140
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.97,
|
| 106 |
+
"learning_rate": 1.5e-05,
|
| 107 |
+
"loss": 0.6925,
|
| 108 |
+
"step": 150
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5207344749329482,
|
| 113 |
+
"eval_loss": 0.6913900375366211,
|
| 114 |
+
"eval_macro_f1": 0.39379505149299676,
|
| 115 |
+
"eval_runtime": 11.7399,
|
| 116 |
+
"eval_samples_per_second": 412.865,
|
| 117 |
+
"eval_steps_per_second": 1.618,
|
| 118 |
+
"step": 152
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.11,
|
| 122 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 123 |
+
"loss": 0.6904,
|
| 124 |
+
"step": 160
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.24,
|
| 128 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 129 |
+
"loss": 0.6911,
|
| 130 |
+
"step": 170
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.37,
|
| 134 |
+
"learning_rate": 1.8e-05,
|
| 135 |
+
"loss": 0.6888,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.5,
|
| 140 |
+
"learning_rate": 1.9e-05,
|
| 141 |
+
"loss": 0.692,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.63,
|
| 146 |
+
"learning_rate": 2e-05,
|
| 147 |
+
"loss": 0.6904,
|
| 148 |
+
"step": 200
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.76,
|
| 152 |
+
"learning_rate": 2.1e-05,
|
| 153 |
+
"loss": 0.6881,
|
| 154 |
+
"step": 210
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.89,
|
| 158 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 159 |
+
"loss": 0.6873,
|
| 160 |
+
"step": 220
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 3.0,
|
| 164 |
+
"eval_accuracy": 0.5409531669073654,
|
| 165 |
+
"eval_loss": 0.6888494491577148,
|
| 166 |
+
"eval_macro_f1": 0.461880985936752,
|
| 167 |
+
"eval_runtime": 11.9909,
|
| 168 |
+
"eval_samples_per_second": 404.224,
|
| 169 |
+
"eval_steps_per_second": 1.585,
|
| 170 |
+
"step": 228
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.03,
|
| 174 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 175 |
+
"loss": 0.6882,
|
| 176 |
+
"step": 230
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.16,
|
| 180 |
+
"learning_rate": 2.4e-05,
|
| 181 |
+
"loss": 0.6848,
|
| 182 |
+
"step": 240
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.29,
|
| 186 |
+
"learning_rate": 2.5e-05,
|
| 187 |
+
"loss": 0.6853,
|
| 188 |
+
"step": 250
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.42,
|
| 192 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 193 |
+
"loss": 0.6852,
|
| 194 |
+
"step": 260
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.55,
|
| 198 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 199 |
+
"loss": 0.6852,
|
| 200 |
+
"step": 270
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 3.68,
|
| 204 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 205 |
+
"loss": 0.6849,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.82,
|
| 210 |
+
"learning_rate": 2.9e-05,
|
| 211 |
+
"loss": 0.6825,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.95,
|
| 216 |
+
"learning_rate": 3e-05,
|
| 217 |
+
"loss": 0.6813,
|
| 218 |
+
"step": 300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.0,
|
| 222 |
+
"eval_accuracy": 0.5562203424798845,
|
| 223 |
+
"eval_loss": 0.6827021837234497,
|
| 224 |
+
"eval_macro_f1": 0.5513306517049066,
|
| 225 |
+
"eval_runtime": 11.4479,
|
| 226 |
+
"eval_samples_per_second": 423.395,
|
| 227 |
+
"eval_steps_per_second": 1.66,
|
| 228 |
+
"step": 304
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 4.08,
|
| 232 |
+
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 0.6798,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.21,
|
| 238 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 239 |
+
"loss": 0.6775,
|
| 240 |
+
"step": 320
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.34,
|
| 244 |
+
"learning_rate": 3.3e-05,
|
| 245 |
+
"loss": 0.6719,
|
| 246 |
+
"step": 330
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.47,
|
| 250 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 251 |
+
"loss": 0.665,
|
| 252 |
+
"step": 340
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.61,
|
| 256 |
+
"learning_rate": 3.5e-05,
|
| 257 |
+
"loss": 0.6616,
|
| 258 |
+
"step": 350
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.74,
|
| 262 |
+
"learning_rate": 3.6e-05,
|
| 263 |
+
"loss": 0.6576,
|
| 264 |
+
"step": 360
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.87,
|
| 268 |
+
"learning_rate": 3.7e-05,
|
| 269 |
+
"loss": 0.6465,
|
| 270 |
+
"step": 370
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 5.0,
|
| 274 |
+
"learning_rate": 3.8e-05,
|
| 275 |
+
"loss": 0.6383,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 5.0,
|
| 280 |
+
"eval_accuracy": 0.6249226325562204,
|
| 281 |
+
"eval_loss": 0.6389692425727844,
|
| 282 |
+
"eval_macro_f1": 0.6159466513902696,
|
| 283 |
+
"eval_runtime": 12.1096,
|
| 284 |
+
"eval_samples_per_second": 400.262,
|
| 285 |
+
"eval_steps_per_second": 1.569,
|
| 286 |
+
"step": 380
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.13,
|
| 290 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 291 |
+
"loss": 0.6047,
|
| 292 |
+
"step": 390
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.26,
|
| 296 |
+
"learning_rate": 4e-05,
|
| 297 |
+
"loss": 0.5807,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 5.39,
|
| 302 |
+
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 0.5336,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 5.53,
|
| 308 |
+
"learning_rate": 4.2e-05,
|
| 309 |
+
"loss": 0.5182,
|
| 310 |
+
"step": 420
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.66,
|
| 314 |
+
"learning_rate": 4.3e-05,
|
| 315 |
+
"loss": 0.4865,
|
| 316 |
+
"step": 430
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.79,
|
| 320 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 321 |
+
"loss": 0.4959,
|
| 322 |
+
"step": 440
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.92,
|
| 326 |
+
"learning_rate": 4.5e-05,
|
| 327 |
+
"loss": 0.4698,
|
| 328 |
+
"step": 450
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 6.0,
|
| 332 |
+
"eval_accuracy": 0.7633587786259542,
|
| 333 |
+
"eval_loss": 0.49028173089027405,
|
| 334 |
+
"eval_macro_f1": 0.7568335054238134,
|
| 335 |
+
"eval_runtime": 11.5665,
|
| 336 |
+
"eval_samples_per_second": 419.056,
|
| 337 |
+
"eval_steps_per_second": 1.643,
|
| 338 |
+
"step": 456
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.05,
|
| 342 |
+
"learning_rate": 4.600000000000001e-05,
|
| 343 |
+
"loss": 0.4584,
|
| 344 |
+
"step": 460
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.18,
|
| 348 |
+
"learning_rate": 4.7e-05,
|
| 349 |
+
"loss": 0.4042,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.32,
|
| 354 |
+
"learning_rate": 4.8e-05,
|
| 355 |
+
"loss": 0.4086,
|
| 356 |
+
"step": 480
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 6.45,
|
| 360 |
+
"learning_rate": 4.9e-05,
|
| 361 |
+
"loss": 0.4152,
|
| 362 |
+
"step": 490
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 6.58,
|
| 366 |
+
"learning_rate": 5e-05,
|
| 367 |
+
"loss": 0.4426,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.71,
|
| 372 |
+
"learning_rate": 4.984848484848485e-05,
|
| 373 |
+
"loss": 0.4064,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.84,
|
| 378 |
+
"learning_rate": 4.9696969696969694e-05,
|
| 379 |
+
"loss": 0.387,
|
| 380 |
+
"step": 520
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.97,
|
| 384 |
+
"learning_rate": 4.9545454545454553e-05,
|
| 385 |
+
"loss": 0.3871,
|
| 386 |
+
"step": 530
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"eval_accuracy": 0.7829585310501341,
|
| 391 |
+
"eval_loss": 0.45330119132995605,
|
| 392 |
+
"eval_macro_f1": 0.7786660383654955,
|
| 393 |
+
"eval_runtime": 11.8487,
|
| 394 |
+
"eval_samples_per_second": 409.075,
|
| 395 |
+
"eval_steps_per_second": 1.604,
|
| 396 |
+
"step": 532
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 7.11,
|
| 400 |
+
"learning_rate": 4.93939393939394e-05,
|
| 401 |
+
"loss": 0.3631,
|
| 402 |
+
"step": 540
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 7.24,
|
| 406 |
+
"learning_rate": 4.9242424242424245e-05,
|
| 407 |
+
"loss": 0.336,
|
| 408 |
+
"step": 550
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.37,
|
| 412 |
+
"learning_rate": 4.909090909090909e-05,
|
| 413 |
+
"loss": 0.3518,
|
| 414 |
+
"step": 560
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 7.5,
|
| 418 |
+
"learning_rate": 4.8939393939393944e-05,
|
| 419 |
+
"loss": 0.3532,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 7.63,
|
| 424 |
+
"learning_rate": 4.878787878787879e-05,
|
| 425 |
+
"loss": 0.3533,
|
| 426 |
+
"step": 580
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.76,
|
| 430 |
+
"learning_rate": 4.863636363636364e-05,
|
| 431 |
+
"loss": 0.3512,
|
| 432 |
+
"step": 590
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 7.89,
|
| 436 |
+
"learning_rate": 4.848484848484849e-05,
|
| 437 |
+
"loss": 0.351,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.0,
|
| 442 |
+
"eval_accuracy": 0.7914173715700433,
|
| 443 |
+
"eval_loss": 0.4372871518135071,
|
| 444 |
+
"eval_macro_f1": 0.7914172295164006,
|
| 445 |
+
"eval_runtime": 11.6679,
|
| 446 |
+
"eval_samples_per_second": 415.414,
|
| 447 |
+
"eval_steps_per_second": 1.628,
|
| 448 |
+
"step": 608
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.03,
|
| 452 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 453 |
+
"loss": 0.3218,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 8.16,
|
| 458 |
+
"learning_rate": 4.8181818181818186e-05,
|
| 459 |
+
"loss": 0.323,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.29,
|
| 464 |
+
"learning_rate": 4.803030303030303e-05,
|
| 465 |
+
"loss": 0.3241,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 8.42,
|
| 470 |
+
"learning_rate": 4.787878787878788e-05,
|
| 471 |
+
"loss": 0.2956,
|
| 472 |
+
"step": 640
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.55,
|
| 476 |
+
"learning_rate": 4.772727272727273e-05,
|
| 477 |
+
"loss": 0.3094,
|
| 478 |
+
"step": 650
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.68,
|
| 482 |
+
"learning_rate": 4.7575757575757576e-05,
|
| 483 |
+
"loss": 0.2916,
|
| 484 |
+
"step": 660
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 8.82,
|
| 488 |
+
"learning_rate": 4.742424242424243e-05,
|
| 489 |
+
"loss": 0.2983,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.95,
|
| 494 |
+
"learning_rate": 4.7272727272727275e-05,
|
| 495 |
+
"loss": 0.312,
|
| 496 |
+
"step": 680
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 9.0,
|
| 500 |
+
"eval_accuracy": 0.7963688879719414,
|
| 501 |
+
"eval_loss": 0.4367808401584625,
|
| 502 |
+
"eval_macro_f1": 0.7962561816949302,
|
| 503 |
+
"eval_runtime": 11.5439,
|
| 504 |
+
"eval_samples_per_second": 419.877,
|
| 505 |
+
"eval_steps_per_second": 1.646,
|
| 506 |
+
"step": 684
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.08,
|
| 510 |
+
"learning_rate": 4.712121212121212e-05,
|
| 511 |
+
"loss": 0.2726,
|
| 512 |
+
"step": 690
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.21,
|
| 516 |
+
"learning_rate": 4.696969696969697e-05,
|
| 517 |
+
"loss": 0.2738,
|
| 518 |
+
"step": 700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.34,
|
| 522 |
+
"learning_rate": 4.681818181818182e-05,
|
| 523 |
+
"loss": 0.2666,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 9.47,
|
| 528 |
+
"learning_rate": 4.666666666666667e-05,
|
| 529 |
+
"loss": 0.2609,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.61,
|
| 534 |
+
"learning_rate": 4.651515151515152e-05,
|
| 535 |
+
"loss": 0.2872,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.74,
|
| 540 |
+
"learning_rate": 4.636363636363636e-05,
|
| 541 |
+
"loss": 0.2827,
|
| 542 |
+
"step": 740
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 9.87,
|
| 546 |
+
"learning_rate": 4.621212121212121e-05,
|
| 547 |
+
"loss": 0.266,
|
| 548 |
+
"step": 750
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.0,
|
| 552 |
+
"learning_rate": 4.606060606060607e-05,
|
| 553 |
+
"loss": 0.274,
|
| 554 |
+
"step": 760
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.0,
|
| 558 |
+
"eval_accuracy": 0.7990509593563029,
|
| 559 |
+
"eval_loss": 0.43979689478874207,
|
| 560 |
+
"eval_macro_f1": 0.7988178222756628,
|
| 561 |
+
"eval_runtime": 11.7674,
|
| 562 |
+
"eval_samples_per_second": 411.901,
|
| 563 |
+
"eval_steps_per_second": 1.615,
|
| 564 |
+
"step": 760
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.13,
|
| 568 |
+
"learning_rate": 4.5909090909090914e-05,
|
| 569 |
+
"loss": 0.2323,
|
| 570 |
+
"step": 770
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 10.26,
|
| 574 |
+
"learning_rate": 4.575757575757576e-05,
|
| 575 |
+
"loss": 0.2351,
|
| 576 |
+
"step": 780
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.39,
|
| 580 |
+
"learning_rate": 4.5606060606060606e-05,
|
| 581 |
+
"loss": 0.224,
|
| 582 |
+
"step": 790
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.53,
|
| 586 |
+
"learning_rate": 4.545454545454546e-05,
|
| 587 |
+
"loss": 0.2463,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 10.66,
|
| 592 |
+
"learning_rate": 4.5303030303030304e-05,
|
| 593 |
+
"loss": 0.2758,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 10.79,
|
| 598 |
+
"learning_rate": 4.515151515151516e-05,
|
| 599 |
+
"loss": 0.3099,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.92,
|
| 604 |
+
"learning_rate": 4.5e-05,
|
| 605 |
+
"loss": 0.2533,
|
| 606 |
+
"step": 830
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 11.0,
|
| 610 |
+
"eval_accuracy": 0.7854342892510832,
|
| 611 |
+
"eval_loss": 0.4947068691253662,
|
| 612 |
+
"eval_macro_f1": 0.7851853543805845,
|
| 613 |
+
"eval_runtime": 11.8229,
|
| 614 |
+
"eval_samples_per_second": 409.968,
|
| 615 |
+
"eval_steps_per_second": 1.607,
|
| 616 |
+
"step": 836
|
| 617 |
+
}
|
| 618 |
+
],
|
| 619 |
+
"max_steps": 3800,
|
| 620 |
+
"num_train_epochs": 50,
|
| 621 |
+
"total_flos": 2937196312461312.0,
|
| 622 |
+
"trial_name": null,
|
| 623 |
+
"trial_params": null
|
| 624 |
+
}
|
cross_cell_type_generization/L2/Neutrophils/fold0/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.02,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.02,
|
| 9 |
+
"hidden_size": 256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 512,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 4,
|
| 16 |
+
"num_hidden_layers": 2,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.28.0",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|