Model save
Browse files- README.md +52 -0
- all_results.json +20 -0
- per_language_test_metrics.csv +16 -0
- test_classification_report.csv +7 -0
- test_confusion_matrix.csv +4 -0
- test_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +776 -0
- validation_results.json +8 -0
README.md
CHANGED
|
@@ -4,6 +4,8 @@ license: apache-2.0
|
|
| 4 |
base_model: bert-base-multilingual-cased
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
model-index:
|
| 8 |
- name: bert-base-multilingual-cased-xnli-nli
|
| 9 |
results: []
|
|
@@ -16,6 +18,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
# bert-base-multilingual-cased-xnli-nli
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on an unknown dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -44,6 +53,49 @@ The following hyperparameters were used during training:
|
|
| 44 |
- num_epochs: 3.0
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
### Framework versions
|
| 48 |
|
| 49 |
- Transformers 5.7.0
|
|
|
|
| 4 |
base_model: bert-base-multilingual-cased
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
model-index:
|
| 10 |
- name: bert-base-multilingual-cased-xnli-nli
|
| 11 |
results: []
|
|
|
|
| 18 |
# bert-base-multilingual-cased-xnli-nli
|
| 19 |
|
| 20 |
This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on an unknown dataset.
|
| 21 |
+
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 0.4953
|
| 23 |
+
- Accuracy: 0.8265
|
| 24 |
+
- Macro F1: 0.8273
|
| 25 |
+
- Weighted F1: 0.8273
|
| 26 |
+
- Macro Precision: 0.8317
|
| 27 |
+
- Macro Recall: 0.8265
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
|
|
| 53 |
- num_epochs: 3.0
|
| 54 |
- mixed_precision_training: Native AMP
|
| 55 |
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 | Weighted F1 | Macro Precision | Macro Recall |
|
| 59 |
+
|:-------------:|:------:|:-----:|:---------------:|:--------:|:--------:|:-----------:|:---------------:|:------------:|
|
| 60 |
+
| 0.9497 | 0.0815 | 1000 | 0.7566 | 0.6876 | 0.6880 | 0.6880 | 0.7059 | 0.6876 |
|
| 61 |
+
| 0.7397 | 0.1630 | 2000 | 0.6681 | 0.7321 | 0.7321 | 0.7321 | 0.7394 | 0.7321 |
|
| 62 |
+
| 0.6674 | 0.2445 | 3000 | 0.6081 | 0.7502 | 0.7514 | 0.7514 | 0.7569 | 0.7502 |
|
| 63 |
+
| 0.6249 | 0.3259 | 4000 | 0.5667 | 0.7647 | 0.7655 | 0.7655 | 0.7688 | 0.7647 |
|
| 64 |
+
| 0.5976 | 0.4074 | 5000 | 0.5627 | 0.7719 | 0.7714 | 0.7714 | 0.7738 | 0.7719 |
|
| 65 |
+
| 0.5774 | 0.4889 | 6000 | 0.5533 | 0.7691 | 0.7705 | 0.7705 | 0.7774 | 0.7691 |
|
| 66 |
+
| 0.5642 | 0.5704 | 7000 | 0.5258 | 0.7984 | 0.7996 | 0.7996 | 0.8086 | 0.7984 |
|
| 67 |
+
| 0.5564 | 0.6519 | 8000 | 0.5392 | 0.7823 | 0.7837 | 0.7837 | 0.7992 | 0.7823 |
|
| 68 |
+
| 0.5467 | 0.7334 | 9000 | 0.4994 | 0.8056 | 0.8057 | 0.8057 | 0.8066 | 0.8056 |
|
| 69 |
+
| 0.5286 | 0.8149 | 10000 | 0.5161 | 0.7960 | 0.7957 | 0.7957 | 0.8010 | 0.7960 |
|
| 70 |
+
| 0.5284 | 0.8963 | 11000 | 0.5209 | 0.7900 | 0.7915 | 0.7915 | 0.8022 | 0.7900 |
|
| 71 |
+
| 0.5125 | 0.9778 | 12000 | 0.5068 | 0.8032 | 0.8044 | 0.8044 | 0.8117 | 0.8032 |
|
| 72 |
+
| 0.4540 | 1.0593 | 13000 | 0.5153 | 0.7980 | 0.7989 | 0.7989 | 0.8076 | 0.7980 |
|
| 73 |
+
| 0.4378 | 1.1408 | 14000 | 0.4989 | 0.8076 | 0.8091 | 0.8091 | 0.8173 | 0.8076 |
|
| 74 |
+
| 0.4366 | 1.2223 | 15000 | 0.5143 | 0.8084 | 0.8083 | 0.8083 | 0.8158 | 0.8084 |
|
| 75 |
+
| 0.4325 | 1.3038 | 16000 | 0.4872 | 0.8056 | 0.8062 | 0.8062 | 0.8125 | 0.8056 |
|
| 76 |
+
| 0.4309 | 1.3853 | 17000 | 0.4673 | 0.8185 | 0.8190 | 0.8190 | 0.8215 | 0.8185 |
|
| 77 |
+
| 0.4326 | 1.4668 | 18000 | 0.4717 | 0.8177 | 0.8186 | 0.8186 | 0.8230 | 0.8177 |
|
| 78 |
+
| 0.4262 | 1.5482 | 19000 | 0.4949 | 0.8153 | 0.8167 | 0.8167 | 0.8267 | 0.8153 |
|
| 79 |
+
| 0.4197 | 1.6297 | 20000 | 0.4759 | 0.8189 | 0.8196 | 0.8196 | 0.8262 | 0.8189 |
|
| 80 |
+
| 0.4139 | 1.7112 | 21000 | 0.4799 | 0.8165 | 0.8172 | 0.8172 | 0.8244 | 0.8165 |
|
| 81 |
+
| 0.4104 | 1.7927 | 22000 | 0.4680 | 0.8281 | 0.8289 | 0.8289 | 0.8330 | 0.8281 |
|
| 82 |
+
| 0.4103 | 1.8742 | 23000 | 0.4697 | 0.8221 | 0.8228 | 0.8228 | 0.8281 | 0.8221 |
|
| 83 |
+
| 0.4132 | 1.9557 | 24000 | 0.4692 | 0.8185 | 0.8194 | 0.8194 | 0.8261 | 0.8185 |
|
| 84 |
+
| 0.3613 | 2.0372 | 25000 | 0.5227 | 0.8237 | 0.8238 | 0.8238 | 0.8265 | 0.8237 |
|
| 85 |
+
| 0.3106 | 2.1186 | 26000 | 0.5053 | 0.8217 | 0.8233 | 0.8233 | 0.8312 | 0.8217 |
|
| 86 |
+
| 0.3107 | 2.2001 | 27000 | 0.4986 | 0.8325 | 0.8331 | 0.8331 | 0.8373 | 0.8325 |
|
| 87 |
+
| 0.3139 | 2.2816 | 28000 | 0.4816 | 0.8333 | 0.8336 | 0.8336 | 0.8357 | 0.8333 |
|
| 88 |
+
| 0.2997 | 2.3631 | 29000 | 0.4880 | 0.8289 | 0.8299 | 0.8299 | 0.8344 | 0.8289 |
|
| 89 |
+
| 0.3104 | 2.4446 | 30000 | 0.5037 | 0.8257 | 0.8267 | 0.8267 | 0.8331 | 0.8257 |
|
| 90 |
+
| 0.3072 | 2.5261 | 31000 | 0.4896 | 0.8273 | 0.8284 | 0.8284 | 0.8337 | 0.8273 |
|
| 91 |
+
| 0.3049 | 2.6076 | 32000 | 0.5036 | 0.8201 | 0.8204 | 0.8204 | 0.8239 | 0.8201 |
|
| 92 |
+
| 0.3067 | 2.6890 | 33000 | 0.4961 | 0.8197 | 0.8204 | 0.8204 | 0.8253 | 0.8197 |
|
| 93 |
+
| 0.3084 | 2.7705 | 34000 | 0.4928 | 0.8237 | 0.8245 | 0.8245 | 0.8298 | 0.8237 |
|
| 94 |
+
| 0.2961 | 2.8520 | 35000 | 0.4940 | 0.8253 | 0.8261 | 0.8261 | 0.8308 | 0.8253 |
|
| 95 |
+
| 0.2939 | 2.9335 | 36000 | 0.4962 | 0.8233 | 0.8243 | 0.8243 | 0.8291 | 0.8233 |
|
| 96 |
+
| 0.2939 | 3.0 | 36816 | 0.4953 | 0.8265 | 0.8273 | 0.8273 | 0.8317 | 0.8265 |
|
| 97 |
+
|
| 98 |
+
|
| 99 |
### Framework versions
|
| 100 |
|
| 101 |
- Transformers 5.7.0
|
all_results.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"test_accuracy": 0.8255489021956088,
|
| 4 |
+
"test_loss": 0.5061046481132507,
|
| 5 |
+
"test_macro_f1": 0.8259998038703972,
|
| 6 |
+
"test_macro_precision": 0.8314179714743389,
|
| 7 |
+
"test_macro_recall": 0.8255489021956087,
|
| 8 |
+
"test_weighted_f1": 0.8259998038703973,
|
| 9 |
+
"total_flos": 3.0997378586864794e+17,
|
| 10 |
+
"train_loss": 0.44763235650027333,
|
| 11 |
+
"train_runtime": 6591.1585,
|
| 12 |
+
"train_samples_per_second": 178.74,
|
| 13 |
+
"train_steps_per_second": 5.586,
|
| 14 |
+
"validation_accuracy": 0.8265060240963855,
|
| 15 |
+
"validation_loss": 0.4953470230102539,
|
| 16 |
+
"validation_macro_f1": 0.8273315514227123,
|
| 17 |
+
"validation_macro_precision": 0.8317304422729516,
|
| 18 |
+
"validation_macro_recall": 0.8265060240963855,
|
| 19 |
+
"validation_weighted_f1": 0.8273315514227122
|
| 20 |
+
}
|
per_language_test_metrics.csv
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
language,test_ar_loss,test_ar_accuracy,test_ar_macro_f1,test_ar_weighted_f1,test_ar_macro_precision,test_ar_macro_recall,test_bg_loss,test_bg_accuracy,test_bg_macro_f1,test_bg_weighted_f1,test_bg_macro_precision,test_bg_macro_recall,test_de_loss,test_de_accuracy,test_de_macro_f1,test_de_weighted_f1,test_de_macro_precision,test_de_macro_recall,test_el_loss,test_el_accuracy,test_el_macro_f1,test_el_weighted_f1,test_el_macro_precision,test_el_macro_recall,test_en_loss,test_en_accuracy,test_en_macro_f1,test_en_weighted_f1,test_en_macro_precision,test_en_macro_recall,test_es_loss,test_es_accuracy,test_es_macro_f1,test_es_weighted_f1,test_es_macro_precision,test_es_macro_recall,test_fr_loss,test_fr_accuracy,test_fr_macro_f1,test_fr_weighted_f1,test_fr_macro_precision,test_fr_macro_recall,test_hi_loss,test_hi_accuracy,test_hi_macro_f1,test_hi_weighted_f1,test_hi_macro_precision,test_hi_macro_recall,test_ru_loss,test_ru_accuracy,test_ru_macro_f1,test_ru_weighted_f1,test_ru_macro_precision,test_ru_macro_recall,test_sw_loss,test_sw_accuracy,test_sw_macro_f1,test_sw_weighted_f1,test_sw_macro_precision,test_sw_macro_recall,test_th_loss,test_th_accuracy,test_th_macro_f1,test_th_weighted_f1,test_th_macro_precision,test_th_macro_recall,test_tr_loss,test_tr_accuracy,test_tr_macro_f1,test_tr_weighted_f1,test_tr_macro_precision,test_tr_macro_recall,test_ur_loss,test_ur_accuracy,test_ur_macro_f1,test_ur_weighted_f1,test_ur_macro_precision,test_ur_macro_recall,test_vi_loss,test_vi_accuracy,test_vi_macro_f1,test_vi_weighted_f1,test_vi_macro_precision,test_vi_macro_recall,test_zh_loss,test_zh_accuracy,test_zh_macro_f1,test_zh_weighted_f1,test_zh_macro_precision,test_zh_macro_recall
|
| 2 |
+
ar,1.3539726734161377,0.5764471057884232,0.5506247184418194,0.5506247184418194,0.7072062225559423,0.5764471057884232,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 3 |
+
bg,,,,,,,0.9392929673194885,0.6858283433133733,0.6778515004478137,0.6778515004478136,0.7460717750155436,0.6858283433133732,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 4 |
+
de,,,,,,,,,,,,,0.8109791278839111,0.7215568862275449,0.7184948704095709,0.7184948704095709,0.7595501471444294,0.7215568862275449,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 5 |
+
el,,,,,,,,,,,,,,,,,,,1.161865472793579,0.6187624750499002,0.5971455936576054,0.5971455936576056,0.7185712976666127,0.6187624750499002,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 6 |
+
en,,,,,,,,,,,,,,,,,,,,,,,,,0.5061046481132507,0.8255489021956088,0.8259998038703972,0.8259998038703973,0.8314179714743389,0.8255489021956087,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 7 |
+
es,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.7382290363311768,0.7504990019960079,0.7484819940935542,0.7484819940935543,0.782094759794571,0.7504990019960079,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 8 |
+
fr,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.7193288207054138,0.7500998003992015,0.7490057645229787,0.7490057645229788,0.777709092979569,0.7500998003992017,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 9 |
+
hi,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1.37227463722229,0.5766467065868264,0.5436767513713999,0.5436767513713999,0.7065099098757718,0.5766467065868263,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 10 |
+
ru,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.8888742327690125,0.7,0.6931757206025999,0.6931757206025999,0.753540207283359,0.7000000000000001,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 11 |
+
sw,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2.042184591293335,0.46387225548902195,0.39295951629279324,0.39295951629279324,0.6723737864855795,0.46387225548902195,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
|
| 12 |
+
th,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2.0412447452545166,0.46107784431137727,0.3824046168061394,0.38240461680613946,0.7066788551077057,0.4610778443113772,,,,,,,,,,,,,,,,,,,,,,,,
|
| 13 |
+
tr,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1.3107990026474,0.5920159680638722,0.5617609701996313,0.5617609701996314,0.7058253988597007,0.5920159680638722,,,,,,,,,,,,,,,,,,
|
| 14 |
+
ur,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1.8040847778320312,0.5017964071856288,0.44211555989060064,0.4421155598906006,0.7111868024641205,0.5017964071856288,,,,,,,,,,,,
|
| 15 |
+
vi,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.9244682788848877,0.691816367265469,0.6842715299504544,0.6842715299504544,0.7490229022570477,0.691816367265469,,,,,,
|
| 16 |
+
zh,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.9174747467041016,0.6960079840319361,0.6854962930367434,0.6854962930367434,0.7517611094915589,0.6960079840319362
|
test_classification_report.csv
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
,precision,recall,f1-score,support
|
| 2 |
+
entailment,0.89937106918239,0.7706586826347306,0.8300548210254757,1670.0
|
| 3 |
+
neutral,0.7611619150080688,0.8473053892215568,0.8019268914706715,1670.0
|
| 4 |
+
contradiction,0.8337209302325581,0.858682634730539,0.8460176991150442,1670.0
|
| 5 |
+
accuracy,0.8255489021956088,0.8255489021956088,0.8255489021956088,0.8255489021956088
|
| 6 |
+
macro avg,0.8314179714743389,0.8255489021956087,0.8259998038703972,5010.0
|
| 7 |
+
weighted avg,0.831417971474339,0.8255489021956088,0.8259998038703973,5010.0
|
test_confusion_matrix.csv
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
,entailment,neutral,contradiction
|
| 2 |
+
entailment,1287,255,128
|
| 3 |
+
neutral,97,1415,158
|
| 4 |
+
contradiction,47,189,1434
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"test_accuracy": 0.8255489021956088,
|
| 3 |
+
"test_loss": 0.5061046481132507,
|
| 4 |
+
"test_macro_f1": 0.8259998038703972,
|
| 5 |
+
"test_macro_precision": 0.8314179714743389,
|
| 6 |
+
"test_macro_recall": 0.8255489021956087,
|
| 7 |
+
"test_weighted_f1": 0.8259998038703973
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 3.0997378586864794e+17,
|
| 4 |
+
"train_loss": 0.44763235650027333,
|
| 5 |
+
"train_runtime": 6591.1585,
|
| 6 |
+
"train_samples_per_second": 178.74,
|
| 7 |
+
"train_steps_per_second": 5.586
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,776 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 28000,
|
| 3 |
+
"best_metric": 0.8336231795481547,
|
| 4 |
+
"best_model_checkpoint": "xnli-bert-base-multilingual-cased/checkpoint-20000",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 1000,
|
| 7 |
+
"global_step": 36816,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08148631029986962,
|
| 14 |
+
"grad_norm": 8.549784660339355,
|
| 15 |
+
"learning_rate": 9.035762788592124e-06,
|
| 16 |
+
"loss": 0.9496568603515625,
|
| 17 |
+
"step": 1000
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.08148631029986962,
|
| 21 |
+
"eval_accuracy": 0.6875502008032128,
|
| 22 |
+
"eval_loss": 0.7566303014755249,
|
| 23 |
+
"eval_macro_f1": 0.6879840073314515,
|
| 24 |
+
"eval_macro_precision": 0.7059287785401316,
|
| 25 |
+
"eval_macro_recall": 0.6875502008032129,
|
| 26 |
+
"eval_runtime": 5.7297,
|
| 27 |
+
"eval_samples_per_second": 434.576,
|
| 28 |
+
"eval_steps_per_second": 6.807,
|
| 29 |
+
"eval_weighted_f1": 0.6879840073314514,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.16297262059973924,
|
| 34 |
+
"grad_norm": 9.595952987670898,
|
| 35 |
+
"learning_rate": 1.80805794477139e-05,
|
| 36 |
+
"loss": 0.7396701049804687,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16297262059973924,
|
| 41 |
+
"eval_accuracy": 0.7321285140562249,
|
| 42 |
+
"eval_loss": 0.6680552363395691,
|
| 43 |
+
"eval_macro_f1": 0.7321148909338402,
|
| 44 |
+
"eval_macro_precision": 0.7394279488762515,
|
| 45 |
+
"eval_macro_recall": 0.7321285140562249,
|
| 46 |
+
"eval_runtime": 5.7558,
|
| 47 |
+
"eval_samples_per_second": 432.604,
|
| 48 |
+
"eval_steps_per_second": 6.776,
|
| 49 |
+
"eval_weighted_f1": 0.7321148909338402,
|
| 50 |
+
"step": 2000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.24445893089960888,
|
| 54 |
+
"grad_norm": 8.339354515075684,
|
| 55 |
+
"learning_rate": 1.954460080330569e-05,
|
| 56 |
+
"loss": 0.6673572998046875,
|
| 57 |
+
"step": 3000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.24445893089960888,
|
| 61 |
+
"eval_accuracy": 0.7502008032128514,
|
| 62 |
+
"eval_loss": 0.6080823540687561,
|
| 63 |
+
"eval_macro_f1": 0.7513857720729318,
|
| 64 |
+
"eval_macro_precision": 0.7568885322566881,
|
| 65 |
+
"eval_macro_recall": 0.7502008032128513,
|
| 66 |
+
"eval_runtime": 5.732,
|
| 67 |
+
"eval_samples_per_second": 434.405,
|
| 68 |
+
"eval_steps_per_second": 6.804,
|
| 69 |
+
"eval_weighted_f1": 0.7513857720729317,
|
| 70 |
+
"step": 3000
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.3259452411994785,
|
| 74 |
+
"grad_norm": 11.520033836364746,
|
| 75 |
+
"learning_rate": 1.896726095876557e-05,
|
| 76 |
+
"loss": 0.624869873046875,
|
| 77 |
+
"step": 4000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.3259452411994785,
|
| 81 |
+
"eval_accuracy": 0.7646586345381526,
|
| 82 |
+
"eval_loss": 0.5667121410369873,
|
| 83 |
+
"eval_macro_f1": 0.7655319847054555,
|
| 84 |
+
"eval_macro_precision": 0.7687715805467343,
|
| 85 |
+
"eval_macro_recall": 0.7646586345381526,
|
| 86 |
+
"eval_runtime": 5.7557,
|
| 87 |
+
"eval_samples_per_second": 432.614,
|
| 88 |
+
"eval_steps_per_second": 6.776,
|
| 89 |
+
"eval_weighted_f1": 0.7655319847054555,
|
| 90 |
+
"step": 4000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.4074315514993481,
|
| 94 |
+
"grad_norm": 9.68370246887207,
|
| 95 |
+
"learning_rate": 1.8389343196463144e-05,
|
| 96 |
+
"loss": 0.5975911254882813,
|
| 97 |
+
"step": 5000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.4074315514993481,
|
| 101 |
+
"eval_accuracy": 0.7718875502008032,
|
| 102 |
+
"eval_loss": 0.5626751780509949,
|
| 103 |
+
"eval_macro_f1": 0.7713603225457878,
|
| 104 |
+
"eval_macro_precision": 0.7738434184862757,
|
| 105 |
+
"eval_macro_recall": 0.7718875502008031,
|
| 106 |
+
"eval_runtime": 5.7021,
|
| 107 |
+
"eval_samples_per_second": 436.685,
|
| 108 |
+
"eval_steps_per_second": 6.84,
|
| 109 |
+
"eval_weighted_f1": 0.7713603225457878,
|
| 110 |
+
"step": 5000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.48891786179921776,
|
| 114 |
+
"grad_norm": 7.512914657592773,
|
| 115 |
+
"learning_rate": 1.7811425434160722e-05,
|
| 116 |
+
"loss": 0.5774126586914062,
|
| 117 |
+
"step": 6000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.48891786179921776,
|
| 121 |
+
"eval_accuracy": 0.7690763052208835,
|
| 122 |
+
"eval_loss": 0.5533095002174377,
|
| 123 |
+
"eval_macro_f1": 0.7705184971308583,
|
| 124 |
+
"eval_macro_precision": 0.7774179830770156,
|
| 125 |
+
"eval_macro_recall": 0.7690763052208837,
|
| 126 |
+
"eval_runtime": 5.7085,
|
| 127 |
+
"eval_samples_per_second": 436.195,
|
| 128 |
+
"eval_steps_per_second": 6.832,
|
| 129 |
+
"eval_weighted_f1": 0.7705184971308584,
|
| 130 |
+
"step": 6000
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.5704041720990873,
|
| 134 |
+
"grad_norm": 6.213924884796143,
|
| 135 |
+
"learning_rate": 1.7233507671858297e-05,
|
| 136 |
+
"loss": 0.564161865234375,
|
| 137 |
+
"step": 7000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.5704041720990873,
|
| 141 |
+
"eval_accuracy": 0.7983935742971887,
|
| 142 |
+
"eval_loss": 0.5257512331008911,
|
| 143 |
+
"eval_macro_f1": 0.7996476536549393,
|
| 144 |
+
"eval_macro_precision": 0.8086006123421692,
|
| 145 |
+
"eval_macro_recall": 0.7983935742971887,
|
| 146 |
+
"eval_runtime": 5.7171,
|
| 147 |
+
"eval_samples_per_second": 435.538,
|
| 148 |
+
"eval_steps_per_second": 6.822,
|
| 149 |
+
"eval_weighted_f1": 0.7996476536549393,
|
| 150 |
+
"step": 7000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.651890482398957,
|
| 154 |
+
"grad_norm": 6.368387222290039,
|
| 155 |
+
"learning_rate": 1.6656167827318175e-05,
|
| 156 |
+
"loss": 0.5564470825195312,
|
| 157 |
+
"step": 8000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.651890482398957,
|
| 161 |
+
"eval_accuracy": 0.7823293172690763,
|
| 162 |
+
"eval_loss": 0.5391626954078674,
|
| 163 |
+
"eval_macro_f1": 0.7837277723275831,
|
| 164 |
+
"eval_macro_precision": 0.7991643111185872,
|
| 165 |
+
"eval_macro_recall": 0.7823293172690763,
|
| 166 |
+
"eval_runtime": 5.9629,
|
| 167 |
+
"eval_samples_per_second": 417.584,
|
| 168 |
+
"eval_steps_per_second": 6.54,
|
| 169 |
+
"eval_weighted_f1": 0.7837277723275831,
|
| 170 |
+
"step": 8000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7333767926988266,
|
| 174 |
+
"grad_norm": 5.787830829620361,
|
| 175 |
+
"learning_rate": 1.607882798277805e-05,
|
| 176 |
+
"loss": 0.5466936645507813,
|
| 177 |
+
"step": 9000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7333767926988266,
|
| 181 |
+
"eval_accuracy": 0.8056224899598393,
|
| 182 |
+
"eval_loss": 0.4993748366832733,
|
| 183 |
+
"eval_macro_f1": 0.8056989803398834,
|
| 184 |
+
"eval_macro_precision": 0.8065890808004245,
|
| 185 |
+
"eval_macro_recall": 0.8056224899598394,
|
| 186 |
+
"eval_runtime": 5.7439,
|
| 187 |
+
"eval_samples_per_second": 433.502,
|
| 188 |
+
"eval_steps_per_second": 6.79,
|
| 189 |
+
"eval_weighted_f1": 0.8056989803398834,
|
| 190 |
+
"step": 9000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.8148631029986962,
|
| 194 |
+
"grad_norm": 9.061445236206055,
|
| 195 |
+
"learning_rate": 1.5500910220475625e-05,
|
| 196 |
+
"loss": 0.5286115112304688,
|
| 197 |
+
"step": 10000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.8148631029986962,
|
| 201 |
+
"eval_accuracy": 0.7959839357429719,
|
| 202 |
+
"eval_loss": 0.5160648226737976,
|
| 203 |
+
"eval_macro_f1": 0.7957087786334758,
|
| 204 |
+
"eval_macro_precision": 0.8009795557642422,
|
| 205 |
+
"eval_macro_recall": 0.795983935742972,
|
| 206 |
+
"eval_runtime": 5.7002,
|
| 207 |
+
"eval_samples_per_second": 436.827,
|
| 208 |
+
"eval_steps_per_second": 6.842,
|
| 209 |
+
"eval_weighted_f1": 0.7957087786334757,
|
| 210 |
+
"step": 10000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.8963494132985659,
|
| 214 |
+
"grad_norm": 7.46348762512207,
|
| 215 |
+
"learning_rate": 1.4922992458173202e-05,
|
| 216 |
+
"loss": 0.5284017944335937,
|
| 217 |
+
"step": 11000
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.8963494132985659,
|
| 221 |
+
"eval_accuracy": 0.7899598393574297,
|
| 222 |
+
"eval_loss": 0.5209371447563171,
|
| 223 |
+
"eval_macro_f1": 0.7915388898553019,
|
| 224 |
+
"eval_macro_precision": 0.8022448341508318,
|
| 225 |
+
"eval_macro_recall": 0.7899598393574297,
|
| 226 |
+
"eval_runtime": 5.737,
|
| 227 |
+
"eval_samples_per_second": 434.021,
|
| 228 |
+
"eval_steps_per_second": 6.798,
|
| 229 |
+
"eval_weighted_f1": 0.7915388898553017,
|
| 230 |
+
"step": 11000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.9778357235984355,
|
| 234 |
+
"grad_norm": 7.875688552856445,
|
| 235 |
+
"learning_rate": 1.434507469587078e-05,
|
| 236 |
+
"loss": 0.512461181640625,
|
| 237 |
+
"step": 12000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.9778357235984355,
|
| 241 |
+
"eval_accuracy": 0.8032128514056225,
|
| 242 |
+
"eval_loss": 0.5068432688713074,
|
| 243 |
+
"eval_macro_f1": 0.8043573883527108,
|
| 244 |
+
"eval_macro_precision": 0.81166966102744,
|
| 245 |
+
"eval_macro_recall": 0.8032128514056226,
|
| 246 |
+
"eval_runtime": 5.7117,
|
| 247 |
+
"eval_samples_per_second": 435.951,
|
| 248 |
+
"eval_steps_per_second": 6.828,
|
| 249 |
+
"eval_weighted_f1": 0.8043573883527108,
|
| 250 |
+
"step": 12000
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.0593220338983051,
|
| 254 |
+
"grad_norm": 9.916313171386719,
|
| 255 |
+
"learning_rate": 1.3767734851330657e-05,
|
| 256 |
+
"loss": 0.4540028076171875,
|
| 257 |
+
"step": 13000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.0593220338983051,
|
| 261 |
+
"eval_accuracy": 0.797991967871486,
|
| 262 |
+
"eval_loss": 0.5152743458747864,
|
| 263 |
+
"eval_macro_f1": 0.7989356423881748,
|
| 264 |
+
"eval_macro_precision": 0.8076428289748036,
|
| 265 |
+
"eval_macro_recall": 0.7979919678714861,
|
| 266 |
+
"eval_runtime": 5.7807,
|
| 267 |
+
"eval_samples_per_second": 430.743,
|
| 268 |
+
"eval_steps_per_second": 6.747,
|
| 269 |
+
"eval_weighted_f1": 0.7989356423881748,
|
| 270 |
+
"step": 13000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.1408083441981747,
|
| 274 |
+
"grad_norm": 8.481383323669434,
|
| 275 |
+
"learning_rate": 1.3189817089028232e-05,
|
| 276 |
+
"loss": 0.4377615051269531,
|
| 277 |
+
"step": 14000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.1408083441981747,
|
| 281 |
+
"eval_accuracy": 0.8076305220883534,
|
| 282 |
+
"eval_loss": 0.4988712668418884,
|
| 283 |
+
"eval_macro_f1": 0.8091233320564388,
|
| 284 |
+
"eval_macro_precision": 0.8173258294801867,
|
| 285 |
+
"eval_macro_recall": 0.8076305220883535,
|
| 286 |
+
"eval_runtime": 5.6253,
|
| 287 |
+
"eval_samples_per_second": 442.645,
|
| 288 |
+
"eval_steps_per_second": 6.933,
|
| 289 |
+
"eval_weighted_f1": 0.8091233320564387,
|
| 290 |
+
"step": 14000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.2222946544980444,
|
| 294 |
+
"grad_norm": 8.306678771972656,
|
| 295 |
+
"learning_rate": 1.261247724448811e-05,
|
| 296 |
+
"loss": 0.43664401245117185,
|
| 297 |
+
"step": 15000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.2222946544980444,
|
| 301 |
+
"eval_accuracy": 0.808433734939759,
|
| 302 |
+
"eval_loss": 0.5143300890922546,
|
| 303 |
+
"eval_macro_f1": 0.8083432960372489,
|
| 304 |
+
"eval_macro_precision": 0.8157697998180963,
|
| 305 |
+
"eval_macro_recall": 0.8084337349397591,
|
| 306 |
+
"eval_runtime": 5.7952,
|
| 307 |
+
"eval_samples_per_second": 429.668,
|
| 308 |
+
"eval_steps_per_second": 6.73,
|
| 309 |
+
"eval_weighted_f1": 0.8083432960372489,
|
| 310 |
+
"step": 15000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.303780964797914,
|
| 314 |
+
"grad_norm": 9.033668518066406,
|
| 315 |
+
"learning_rate": 1.2034559482185687e-05,
|
| 316 |
+
"loss": 0.43245242309570314,
|
| 317 |
+
"step": 16000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.303780964797914,
|
| 321 |
+
"eval_accuracy": 0.8056224899598393,
|
| 322 |
+
"eval_loss": 0.4871886968612671,
|
| 323 |
+
"eval_macro_f1": 0.8061545718943539,
|
| 324 |
+
"eval_macro_precision": 0.8125407452656841,
|
| 325 |
+
"eval_macro_recall": 0.8056224899598394,
|
| 326 |
+
"eval_runtime": 5.7227,
|
| 327 |
+
"eval_samples_per_second": 435.108,
|
| 328 |
+
"eval_steps_per_second": 6.815,
|
| 329 |
+
"eval_weighted_f1": 0.8061545718943538,
|
| 330 |
+
"step": 16000
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.3852672750977835,
|
| 334 |
+
"grad_norm": 9.388423919677734,
|
| 335 |
+
"learning_rate": 1.1457219637645565e-05,
|
| 336 |
+
"loss": 0.430885009765625,
|
| 337 |
+
"step": 17000
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 1.3852672750977835,
|
| 341 |
+
"eval_accuracy": 0.8184738955823293,
|
| 342 |
+
"eval_loss": 0.4673291742801666,
|
| 343 |
+
"eval_macro_f1": 0.8189626473149584,
|
| 344 |
+
"eval_macro_precision": 0.8214953129431827,
|
| 345 |
+
"eval_macro_recall": 0.8184738955823293,
|
| 346 |
+
"eval_runtime": 5.722,
|
| 347 |
+
"eval_samples_per_second": 435.163,
|
| 348 |
+
"eval_steps_per_second": 6.816,
|
| 349 |
+
"eval_weighted_f1": 0.8189626473149585,
|
| 350 |
+
"step": 17000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.4667535853976532,
|
| 354 |
+
"grad_norm": 6.731133460998535,
|
| 355 |
+
"learning_rate": 1.087930187534314e-05,
|
| 356 |
+
"loss": 0.4325551452636719,
|
| 357 |
+
"step": 18000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.4667535853976532,
|
| 361 |
+
"eval_accuracy": 0.8176706827309237,
|
| 362 |
+
"eval_loss": 0.4717461168766022,
|
| 363 |
+
"eval_macro_f1": 0.8185687309421157,
|
| 364 |
+
"eval_macro_precision": 0.8229753636899145,
|
| 365 |
+
"eval_macro_recall": 0.8176706827309236,
|
| 366 |
+
"eval_runtime": 5.7426,
|
| 367 |
+
"eval_samples_per_second": 433.6,
|
| 368 |
+
"eval_steps_per_second": 6.791,
|
| 369 |
+
"eval_weighted_f1": 0.8185687309421157,
|
| 370 |
+
"step": 18000
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 1.548239895697523,
|
| 374 |
+
"grad_norm": 6.908151149749756,
|
| 375 |
+
"learning_rate": 1.0301962030803017e-05,
|
| 376 |
+
"loss": 0.4262227783203125,
|
| 377 |
+
"step": 19000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.548239895697523,
|
| 381 |
+
"eval_accuracy": 0.8152610441767069,
|
| 382 |
+
"eval_loss": 0.4948924481868744,
|
| 383 |
+
"eval_macro_f1": 0.8166867448251058,
|
| 384 |
+
"eval_macro_precision": 0.8266924018205198,
|
| 385 |
+
"eval_macro_recall": 0.8152610441767069,
|
| 386 |
+
"eval_runtime": 5.7543,
|
| 387 |
+
"eval_samples_per_second": 432.72,
|
| 388 |
+
"eval_steps_per_second": 6.778,
|
| 389 |
+
"eval_weighted_f1": 0.8166867448251058,
|
| 390 |
+
"step": 19000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.6297262059973925,
|
| 394 |
+
"grad_norm": 6.7513227462768555,
|
| 395 |
+
"learning_rate": 9.724622186262896e-06,
|
| 396 |
+
"loss": 0.41966336059570314,
|
| 397 |
+
"step": 20000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.6297262059973925,
|
| 401 |
+
"eval_accuracy": 0.8188755020080322,
|
| 402 |
+
"eval_loss": 0.47589999437332153,
|
| 403 |
+
"eval_macro_f1": 0.8196458915047441,
|
| 404 |
+
"eval_macro_precision": 0.8261850771204332,
|
| 405 |
+
"eval_macro_recall": 0.8188755020080322,
|
| 406 |
+
"eval_runtime": 5.7291,
|
| 407 |
+
"eval_samples_per_second": 434.624,
|
| 408 |
+
"eval_steps_per_second": 6.807,
|
| 409 |
+
"eval_weighted_f1": 0.8196458915047441,
|
| 410 |
+
"step": 20000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.711212516297262,
|
| 414 |
+
"grad_norm": 10.855615615844727,
|
| 415 |
+
"learning_rate": 9.14670442396047e-06,
|
| 416 |
+
"loss": 0.4138599853515625,
|
| 417 |
+
"step": 21000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.711212516297262,
|
| 421 |
+
"eval_accuracy": 0.8164658634538152,
|
| 422 |
+
"eval_loss": 0.47990933060646057,
|
| 423 |
+
"eval_macro_f1": 0.8171903180778263,
|
| 424 |
+
"eval_macro_precision": 0.8243709799384801,
|
| 425 |
+
"eval_macro_recall": 0.8164658634538152,
|
| 426 |
+
"eval_runtime": 5.7933,
|
| 427 |
+
"eval_samples_per_second": 429.804,
|
| 428 |
+
"eval_steps_per_second": 6.732,
|
| 429 |
+
"eval_weighted_f1": 0.8171903180778263,
|
| 430 |
+
"step": 21000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.7926988265971318,
|
| 434 |
+
"grad_norm": 9.089241027832031,
|
| 435 |
+
"learning_rate": 8.568786661658047e-06,
|
| 436 |
+
"loss": 0.4103529052734375,
|
| 437 |
+
"step": 22000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.7926988265971318,
|
| 441 |
+
"eval_accuracy": 0.8281124497991967,
|
| 442 |
+
"eval_loss": 0.4679791331291199,
|
| 443 |
+
"eval_macro_f1": 0.8289008383222325,
|
| 444 |
+
"eval_macro_precision": 0.8329625436009871,
|
| 445 |
+
"eval_macro_recall": 0.8281124497991968,
|
| 446 |
+
"eval_runtime": 5.7356,
|
| 447 |
+
"eval_samples_per_second": 434.13,
|
| 448 |
+
"eval_steps_per_second": 6.8,
|
| 449 |
+
"eval_weighted_f1": 0.8289008383222325,
|
| 450 |
+
"step": 22000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.8741851368970013,
|
| 454 |
+
"grad_norm": 9.242090225219727,
|
| 455 |
+
"learning_rate": 7.990868899355623e-06,
|
| 456 |
+
"loss": 0.410257568359375,
|
| 457 |
+
"step": 23000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.8741851368970013,
|
| 461 |
+
"eval_accuracy": 0.8220883534136546,
|
| 462 |
+
"eval_loss": 0.4697434902191162,
|
| 463 |
+
"eval_macro_f1": 0.8227827215470326,
|
| 464 |
+
"eval_macro_precision": 0.8280806806550514,
|
| 465 |
+
"eval_macro_recall": 0.8220883534136546,
|
| 466 |
+
"eval_runtime": 5.7629,
|
| 467 |
+
"eval_samples_per_second": 432.073,
|
| 468 |
+
"eval_steps_per_second": 6.767,
|
| 469 |
+
"eval_weighted_f1": 0.8227827215470326,
|
| 470 |
+
"step": 23000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.9556714471968708,
|
| 474 |
+
"grad_norm": 5.53393030166626,
|
| 475 |
+
"learning_rate": 7.4135290548155004e-06,
|
| 476 |
+
"loss": 0.4131949157714844,
|
| 477 |
+
"step": 24000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.9556714471968708,
|
| 481 |
+
"eval_accuracy": 0.8184738955823293,
|
| 482 |
+
"eval_loss": 0.4692005217075348,
|
| 483 |
+
"eval_macro_f1": 0.8193875507628111,
|
| 484 |
+
"eval_macro_precision": 0.8260978099246933,
|
| 485 |
+
"eval_macro_recall": 0.8184738955823293,
|
| 486 |
+
"eval_runtime": 5.7422,
|
| 487 |
+
"eval_samples_per_second": 433.634,
|
| 488 |
+
"eval_steps_per_second": 6.792,
|
| 489 |
+
"eval_weighted_f1": 0.8193875507628111,
|
| 490 |
+
"step": 24000
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 2.0371577574967406,
|
| 494 |
+
"grad_norm": 12.3930025100708,
|
| 495 |
+
"learning_rate": 6.835611292513076e-06,
|
| 496 |
+
"loss": 0.36134652709960935,
|
| 497 |
+
"step": 25000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.0371577574967406,
|
| 501 |
+
"eval_accuracy": 0.8236947791164658,
|
| 502 |
+
"eval_loss": 0.5227094292640686,
|
| 503 |
+
"eval_macro_f1": 0.823792592459842,
|
| 504 |
+
"eval_macro_precision": 0.826498500307936,
|
| 505 |
+
"eval_macro_recall": 0.8236947791164658,
|
| 506 |
+
"eval_runtime": 5.7365,
|
| 507 |
+
"eval_samples_per_second": 434.06,
|
| 508 |
+
"eval_steps_per_second": 6.799,
|
| 509 |
+
"eval_weighted_f1": 0.823792592459842,
|
| 510 |
+
"step": 25000
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.1186440677966103,
|
| 514 |
+
"grad_norm": 10.391782760620117,
|
| 515 |
+
"learning_rate": 6.258271447972955e-06,
|
| 516 |
+
"loss": 0.3106440734863281,
|
| 517 |
+
"step": 26000
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.1186440677966103,
|
| 521 |
+
"eval_accuracy": 0.8216867469879519,
|
| 522 |
+
"eval_loss": 0.5052626729011536,
|
| 523 |
+
"eval_macro_f1": 0.8233042973483187,
|
| 524 |
+
"eval_macro_precision": 0.8312149477097931,
|
| 525 |
+
"eval_macro_recall": 0.8216867469879517,
|
| 526 |
+
"eval_runtime": 5.7601,
|
| 527 |
+
"eval_samples_per_second": 432.284,
|
| 528 |
+
"eval_steps_per_second": 6.771,
|
| 529 |
+
"eval_weighted_f1": 0.8233042973483186,
|
| 530 |
+
"step": 26000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.2001303780964796,
|
| 534 |
+
"grad_norm": 8.926087379455566,
|
| 535 |
+
"learning_rate": 5.68035368567053e-06,
|
| 536 |
+
"loss": 0.3107466735839844,
|
| 537 |
+
"step": 27000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.2001303780964796,
|
| 541 |
+
"eval_accuracy": 0.8325301204819278,
|
| 542 |
+
"eval_loss": 0.4985923171043396,
|
| 543 |
+
"eval_macro_f1": 0.8331416775647565,
|
| 544 |
+
"eval_macro_precision": 0.8372987568543514,
|
| 545 |
+
"eval_macro_recall": 0.8325301204819278,
|
| 546 |
+
"eval_runtime": 5.6613,
|
| 547 |
+
"eval_samples_per_second": 439.828,
|
| 548 |
+
"eval_steps_per_second": 6.889,
|
| 549 |
+
"eval_weighted_f1": 0.8331416775647563,
|
| 550 |
+
"step": 27000
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.2816166883963493,
|
| 554 |
+
"grad_norm": 10.237663269042969,
|
| 555 |
+
"learning_rate": 5.103013841130407e-06,
|
| 556 |
+
"loss": 0.3139285888671875,
|
| 557 |
+
"step": 28000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.2816166883963493,
|
| 561 |
+
"eval_accuracy": 0.8333333333333334,
|
| 562 |
+
"eval_loss": 0.4816053807735443,
|
| 563 |
+
"eval_macro_f1": 0.8336231795481547,
|
| 564 |
+
"eval_macro_precision": 0.8357287673721228,
|
| 565 |
+
"eval_macro_recall": 0.8333333333333334,
|
| 566 |
+
"eval_runtime": 5.7536,
|
| 567 |
+
"eval_samples_per_second": 432.773,
|
| 568 |
+
"eval_steps_per_second": 6.778,
|
| 569 |
+
"eval_weighted_f1": 0.8336231795481548,
|
| 570 |
+
"step": 28000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.363102998696219,
|
| 574 |
+
"grad_norm": 9.762932777404785,
|
| 575 |
+
"learning_rate": 4.525096078827983e-06,
|
| 576 |
+
"loss": 0.29973651123046874,
|
| 577 |
+
"step": 29000
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 2.363102998696219,
|
| 581 |
+
"eval_accuracy": 0.8289156626506025,
|
| 582 |
+
"eval_loss": 0.4879550039768219,
|
| 583 |
+
"eval_macro_f1": 0.8299340181736099,
|
| 584 |
+
"eval_macro_precision": 0.8343532590605207,
|
| 585 |
+
"eval_macro_recall": 0.8289156626506023,
|
| 586 |
+
"eval_runtime": 5.7575,
|
| 587 |
+
"eval_samples_per_second": 432.477,
|
| 588 |
+
"eval_steps_per_second": 6.774,
|
| 589 |
+
"eval_weighted_f1": 0.8299340181736099,
|
| 590 |
+
"step": 29000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.444589308996089,
|
| 594 |
+
"grad_norm": 8.687711715698242,
|
| 595 |
+
"learning_rate": 3.947756234287861e-06,
|
| 596 |
+
"loss": 0.31043545532226563,
|
| 597 |
+
"step": 30000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.444589308996089,
|
| 601 |
+
"eval_accuracy": 0.8257028112449799,
|
| 602 |
+
"eval_loss": 0.5036651492118835,
|
| 603 |
+
"eval_macro_f1": 0.8267076999131673,
|
| 604 |
+
"eval_macro_precision": 0.8331173229269838,
|
| 605 |
+
"eval_macro_recall": 0.8257028112449799,
|
| 606 |
+
"eval_runtime": 5.6326,
|
| 607 |
+
"eval_samples_per_second": 442.071,
|
| 608 |
+
"eval_steps_per_second": 6.924,
|
| 609 |
+
"eval_weighted_f1": 0.8267076999131673,
|
| 610 |
+
"step": 30000
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 2.526075619295958,
|
| 614 |
+
"grad_norm": 10.50027084350586,
|
| 615 |
+
"learning_rate": 3.3698384719854367e-06,
|
| 616 |
+
"loss": 0.3071921081542969,
|
| 617 |
+
"step": 31000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 2.526075619295958,
|
| 621 |
+
"eval_accuracy": 0.8273092369477911,
|
| 622 |
+
"eval_loss": 0.489608496427536,
|
| 623 |
+
"eval_macro_f1": 0.8284013022990927,
|
| 624 |
+
"eval_macro_precision": 0.8337436376169335,
|
| 625 |
+
"eval_macro_recall": 0.8273092369477911,
|
| 626 |
+
"eval_runtime": 5.8635,
|
| 627 |
+
"eval_samples_per_second": 424.662,
|
| 628 |
+
"eval_steps_per_second": 6.651,
|
| 629 |
+
"eval_weighted_f1": 0.8284013022990928,
|
| 630 |
+
"step": 31000
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.607561929595828,
|
| 634 |
+
"grad_norm": 21.264141082763672,
|
| 635 |
+
"learning_rate": 2.7919207096830126e-06,
|
| 636 |
+
"loss": 0.30490707397460937,
|
| 637 |
+
"step": 32000
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 2.607561929595828,
|
| 641 |
+
"eval_accuracy": 0.8200803212851405,
|
| 642 |
+
"eval_loss": 0.5036051273345947,
|
| 643 |
+
"eval_macro_f1": 0.820417379908795,
|
| 644 |
+
"eval_macro_precision": 0.8239042794142288,
|
| 645 |
+
"eval_macro_recall": 0.8200803212851406,
|
| 646 |
+
"eval_runtime": 5.8272,
|
| 647 |
+
"eval_samples_per_second": 427.303,
|
| 648 |
+
"eval_steps_per_second": 6.693,
|
| 649 |
+
"eval_weighted_f1": 0.820417379908795,
|
| 650 |
+
"step": 32000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.6890482398956976,
|
| 654 |
+
"grad_norm": 10.950037002563477,
|
| 655 |
+
"learning_rate": 2.214002947380588e-06,
|
| 656 |
+
"loss": 0.30665084838867185,
|
| 657 |
+
"step": 33000
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.6890482398956976,
|
| 661 |
+
"eval_accuracy": 0.8196787148594378,
|
| 662 |
+
"eval_loss": 0.4961128830909729,
|
| 663 |
+
"eval_macro_f1": 0.8204229849224803,
|
| 664 |
+
"eval_macro_precision": 0.8252666860958886,
|
| 665 |
+
"eval_macro_recall": 0.8196787148594377,
|
| 666 |
+
"eval_runtime": 5.7283,
|
| 667 |
+
"eval_samples_per_second": 434.685,
|
| 668 |
+
"eval_steps_per_second": 6.808,
|
| 669 |
+
"eval_weighted_f1": 0.8204229849224803,
|
| 670 |
+
"step": 33000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.770534550195567,
|
| 674 |
+
"grad_norm": 8.680035591125488,
|
| 675 |
+
"learning_rate": 1.6366631028404658e-06,
|
| 676 |
+
"loss": 0.3084286804199219,
|
| 677 |
+
"step": 34000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.770534550195567,
|
| 681 |
+
"eval_accuracy": 0.8236947791164658,
|
| 682 |
+
"eval_loss": 0.49275636672973633,
|
| 683 |
+
"eval_macro_f1": 0.8244750977167675,
|
| 684 |
+
"eval_macro_precision": 0.8298012603410335,
|
| 685 |
+
"eval_macro_recall": 0.8236947791164658,
|
| 686 |
+
"eval_runtime": 5.9171,
|
| 687 |
+
"eval_samples_per_second": 420.814,
|
| 688 |
+
"eval_steps_per_second": 6.591,
|
| 689 |
+
"eval_weighted_f1": 0.8244750977167675,
|
| 690 |
+
"step": 34000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.8520208604954367,
|
| 694 |
+
"grad_norm": 4.615978717803955,
|
| 695 |
+
"learning_rate": 1.0587453405380415e-06,
|
| 696 |
+
"loss": 0.29611859130859375,
|
| 697 |
+
"step": 35000
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.8520208604954367,
|
| 701 |
+
"eval_accuracy": 0.8253012048192772,
|
| 702 |
+
"eval_loss": 0.4939655661582947,
|
| 703 |
+
"eval_macro_f1": 0.8261424158881535,
|
| 704 |
+
"eval_macro_precision": 0.830750130118512,
|
| 705 |
+
"eval_macro_recall": 0.8253012048192772,
|
| 706 |
+
"eval_runtime": 5.7867,
|
| 707 |
+
"eval_samples_per_second": 430.299,
|
| 708 |
+
"eval_steps_per_second": 6.74,
|
| 709 |
+
"eval_weighted_f1": 0.8261424158881534,
|
| 710 |
+
"step": 35000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.9335071707953064,
|
| 714 |
+
"grad_norm": 7.382562160491943,
|
| 715 |
+
"learning_rate": 4.814054959979196e-07,
|
| 716 |
+
"loss": 0.29393319702148435,
|
| 717 |
+
"step": 36000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 2.9335071707953064,
|
| 721 |
+
"eval_accuracy": 0.8232931726907631,
|
| 722 |
+
"eval_loss": 0.4962431788444519,
|
| 723 |
+
"eval_macro_f1": 0.8243093400266494,
|
| 724 |
+
"eval_macro_precision": 0.8291060023138069,
|
| 725 |
+
"eval_macro_recall": 0.8232931726907631,
|
| 726 |
+
"eval_runtime": 5.9218,
|
| 727 |
+
"eval_samples_per_second": 420.479,
|
| 728 |
+
"eval_steps_per_second": 6.586,
|
| 729 |
+
"eval_weighted_f1": 0.8243093400266494,
|
| 730 |
+
"step": 36000
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.0,
|
| 734 |
+
"eval_accuracy": 0.8265060240963855,
|
| 735 |
+
"eval_loss": 0.4953470230102539,
|
| 736 |
+
"eval_macro_f1": 0.8273315514227123,
|
| 737 |
+
"eval_macro_precision": 0.8317304422729516,
|
| 738 |
+
"eval_macro_recall": 0.8265060240963855,
|
| 739 |
+
"eval_runtime": 5.7004,
|
| 740 |
+
"eval_samples_per_second": 436.811,
|
| 741 |
+
"eval_steps_per_second": 6.842,
|
| 742 |
+
"eval_weighted_f1": 0.8273315514227122,
|
| 743 |
+
"step": 36816
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.0,
|
| 747 |
+
"step": 36816,
|
| 748 |
+
"total_flos": 3.0997378586864794e+17,
|
| 749 |
+
"train_loss": 0.44763235650027333,
|
| 750 |
+
"train_runtime": 6591.1585,
|
| 751 |
+
"train_samples_per_second": 178.74,
|
| 752 |
+
"train_steps_per_second": 5.586
|
| 753 |
+
}
|
| 754 |
+
],
|
| 755 |
+
"logging_steps": 1000,
|
| 756 |
+
"max_steps": 36816,
|
| 757 |
+
"num_input_tokens_seen": 0,
|
| 758 |
+
"num_train_epochs": 3,
|
| 759 |
+
"save_steps": 10000,
|
| 760 |
+
"stateful_callbacks": {
|
| 761 |
+
"TrainerControl": {
|
| 762 |
+
"args": {
|
| 763 |
+
"should_epoch_stop": false,
|
| 764 |
+
"should_evaluate": false,
|
| 765 |
+
"should_log": false,
|
| 766 |
+
"should_save": true,
|
| 767 |
+
"should_training_stop": true
|
| 768 |
+
},
|
| 769 |
+
"attributes": {}
|
| 770 |
+
}
|
| 771 |
+
},
|
| 772 |
+
"total_flos": 3.0997378586864794e+17,
|
| 773 |
+
"train_batch_size": 32,
|
| 774 |
+
"trial_name": null,
|
| 775 |
+
"trial_params": null
|
| 776 |
+
}
|
validation_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"validation_accuracy": 0.8265060240963855,
|
| 3 |
+
"validation_loss": 0.4953470230102539,
|
| 4 |
+
"validation_macro_f1": 0.8273315514227123,
|
| 5 |
+
"validation_macro_precision": 0.8317304422729516,
|
| 6 |
+
"validation_macro_recall": 0.8265060240963855,
|
| 7 |
+
"validation_weighted_f1": 0.8273315514227122
|
| 8 |
+
}
|