Model save
Browse files- README.md +21 -14
- all_results.json +17 -17
- model.safetensors +1 -1
- runs/Jun15_00-40-18_92b2e0e6fb20/events.out.tfevents.1749948301.92b2e0e6fb20.2194.9 +3 -0
- runs/Jun15_00-47-20_92b2e0e6fb20/events.out.tfevents.1749948441.92b2e0e6fb20.2194.10 +3 -0
- test_results.json +13 -13
- train_results.json +5 -5
- trainer_state.json +383 -24
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -16,15 +16,15 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [maximuspowers/bert-philosophy-adapted](https://huggingface.co/maximuspowers/bert-philosophy-adapted) on the None dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 0.
|
| 20 |
-
- Exact Match Accuracy: 0.
|
| 21 |
-
- Macro Precision: 0.
|
| 22 |
-
- Macro Recall: 0.
|
| 23 |
-
- Macro F1: 0.
|
| 24 |
-
- Micro Precision: 0.
|
| 25 |
-
- Micro Recall: 0.
|
| 26 |
-
- Micro F1: 0.
|
| 27 |
-
- Hamming Loss: 0.
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
@@ -44,11 +44,11 @@ More information needed
|
|
| 44 |
|
| 45 |
The following hyperparameters were used during training:
|
| 46 |
- learning_rate: 2e-05
|
| 47 |
-
- train_batch_size:
|
| 48 |
-
- eval_batch_size:
|
| 49 |
- seed: 42
|
| 50 |
- gradient_accumulation_steps: 2
|
| 51 |
-
- total_train_batch_size:
|
| 52 |
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
- lr_scheduler_warmup_steps: 100
|
|
@@ -59,8 +59,15 @@ The following hyperparameters were used during training:
|
|
| 59 |
|
| 60 |
| Training Loss | Epoch | Step | Validation Loss | Exact Match Accuracy | Macro Precision | Macro Recall | Macro F1 | Micro Precision | Micro Recall | Micro F1 | Hamming Loss |
|
| 61 |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|:---------------:|:------------:|:--------:|:---------------:|:------------:|:--------:|:------------:|
|
| 62 |
-
|
|
| 63 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
|
| 65 |
|
| 66 |
### Framework versions
|
|
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [maximuspowers/bert-philosophy-adapted](https://huggingface.co/maximuspowers/bert-philosophy-adapted) on the None dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.8156
|
| 20 |
+
- Exact Match Accuracy: 0.275
|
| 21 |
+
- Macro Precision: 0.1574
|
| 22 |
+
- Macro Recall: 0.1134
|
| 23 |
+
- Macro F1: 0.1298
|
| 24 |
+
- Micro Precision: 0.8421
|
| 25 |
+
- Micro Recall: 0.2807
|
| 26 |
+
- Micro F1: 0.4211
|
| 27 |
+
- Hamming Loss: 0.0647
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
|
|
| 44 |
|
| 45 |
The following hyperparameters were used during training:
|
| 46 |
- learning_rate: 2e-05
|
| 47 |
+
- train_batch_size: 8
|
| 48 |
+
- eval_batch_size: 8
|
| 49 |
- seed: 42
|
| 50 |
- gradient_accumulation_steps: 2
|
| 51 |
+
- total_train_batch_size: 16
|
| 52 |
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
- lr_scheduler_warmup_steps: 100
|
|
|
|
| 59 |
|
| 60 |
| Training Loss | Epoch | Step | Validation Loss | Exact Match Accuracy | Macro Precision | Macro Recall | Macro F1 | Micro Precision | Micro Recall | Micro F1 | Hamming Loss |
|
| 61 |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|:---------------:|:------------:|:--------:|:---------------:|:------------:|:--------:|:------------:|
|
| 62 |
+
| 1.7889 | 5.0 | 100 | 1.0021 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0853 |
|
| 63 |
+
| 1.156 | 10.0 | 200 | 0.8631 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0838 |
|
| 64 |
+
| 0.8775 | 15.0 | 300 | 0.9324 | 0.05 | 0.0588 | 0.0267 | 0.0368 | 1.0 | 0.0877 | 0.1613 | 0.0765 |
|
| 65 |
+
| 0.7747 | 20.0 | 400 | 0.7537 | 0.1 | 0.1092 | 0.0615 | 0.0784 | 0.875 | 0.1228 | 0.2154 | 0.075 |
|
| 66 |
+
| 0.7074 | 25.0 | 500 | 0.8191 | 0.175 | 0.1487 | 0.0845 | 0.1056 | 0.7857 | 0.1930 | 0.3099 | 0.0721 |
|
| 67 |
+
| 0.6281 | 30.0 | 600 | 0.8507 | 0.275 | 0.1574 | 0.1134 | 0.1298 | 0.8421 | 0.2807 | 0.4211 | 0.0647 |
|
| 68 |
+
| 0.5506 | 35.0 | 700 | 0.7439 | 0.25 | 0.1563 | 0.1075 | 0.1256 | 0.8333 | 0.2632 | 0.4 | 0.0662 |
|
| 69 |
+
| 0.5091 | 40.0 | 800 | 0.7972 | 0.275 | 0.1574 | 0.1134 | 0.1298 | 0.8421 | 0.2807 | 0.4211 | 0.0647 |
|
| 70 |
+
| 0.5038 | 45.0 | 900 | 0.8156 | 0.275 | 0.1574 | 0.1134 | 0.1298 | 0.8421 | 0.2807 | 0.4211 | 0.0647 |
|
| 71 |
|
| 72 |
|
| 73 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_exact_match_accuracy": 0.
|
| 4 |
-
"eval_hamming_loss": 0.
|
| 5 |
-
"eval_loss": 0.
|
| 6 |
-
"eval_macro_f1": 0.
|
| 7 |
-
"eval_macro_precision": 0.
|
| 8 |
-
"eval_macro_recall": 0.
|
| 9 |
-
"eval_micro_f1": 0.
|
| 10 |
-
"eval_micro_precision": 0.
|
| 11 |
-
"eval_micro_recall": 0.
|
| 12 |
-
"eval_runtime": 0.
|
| 13 |
-
"eval_samples_per_second":
|
| 14 |
-
"eval_steps_per_second": 13.
|
| 15 |
"total_flos": 0.0,
|
| 16 |
-
"train_loss":
|
| 17 |
-
"train_runtime":
|
| 18 |
-
"train_samples_per_second":
|
| 19 |
-
"train_steps_per_second":
|
| 20 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 45.0,
|
| 3 |
+
"eval_exact_match_accuracy": 0.375,
|
| 4 |
+
"eval_hamming_loss": 0.052941176470588235,
|
| 5 |
+
"eval_loss": 0.5750908255577087,
|
| 6 |
+
"eval_macro_f1": 0.13746934180370715,
|
| 7 |
+
"eval_macro_precision": 0.17058823529411765,
|
| 8 |
+
"eval_macro_recall": 0.12058823529411763,
|
| 9 |
+
"eval_micro_f1": 0.55,
|
| 10 |
+
"eval_micro_precision": 0.9565217391304348,
|
| 11 |
+
"eval_micro_recall": 0.38596491228070173,
|
| 12 |
+
"eval_runtime": 0.2248,
|
| 13 |
+
"eval_samples_per_second": 177.928,
|
| 14 |
+
"eval_steps_per_second": 13.345,
|
| 15 |
"total_flos": 0.0,
|
| 16 |
+
"train_loss": 0.9705644819471572,
|
| 17 |
+
"train_runtime": 232.6541,
|
| 18 |
+
"train_samples_per_second": 67.912,
|
| 19 |
+
"train_steps_per_second": 4.298
|
| 20 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 441154988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:890f9065c802bc97c554e035af9eaa8ef8da20f13c0f284d224585cdb51a36aa
|
| 3 |
size 441154988
|
runs/Jun15_00-40-18_92b2e0e6fb20/events.out.tfevents.1749948301.92b2e0e6fb20.2194.9
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:106c9524987fb915bca106b13f7d5b464b289c45f4975ddfcbcff0c2f11c817f
|
| 3 |
+
size 3837
|
runs/Jun15_00-47-20_92b2e0e6fb20/events.out.tfevents.1749948441.92b2e0e6fb20.2194.10
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b7d3f157a6e34681655d66b8adcdcc4d0909527514c24f24be959d00ecde095
|
| 3 |
+
size 57078
|
test_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_exact_match_accuracy": 0.
|
| 4 |
-
"eval_hamming_loss": 0.
|
| 5 |
-
"eval_loss": 0.
|
| 6 |
-
"eval_macro_f1": 0.
|
| 7 |
-
"eval_macro_precision": 0.
|
| 8 |
-
"eval_macro_recall": 0.
|
| 9 |
-
"eval_micro_f1": 0.
|
| 10 |
-
"eval_micro_precision": 0.
|
| 11 |
-
"eval_micro_recall": 0.
|
| 12 |
-
"eval_runtime": 0.
|
| 13 |
-
"eval_samples_per_second":
|
| 14 |
-
"eval_steps_per_second": 13.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_exact_match_accuracy": 0.375,
|
| 4 |
+
"eval_hamming_loss": 0.052941176470588235,
|
| 5 |
+
"eval_loss": 0.5750908255577087,
|
| 6 |
+
"eval_macro_f1": 0.13746934180370715,
|
| 7 |
+
"eval_macro_precision": 0.17058823529411765,
|
| 8 |
+
"eval_macro_recall": 0.12058823529411763,
|
| 9 |
+
"eval_micro_f1": 0.55,
|
| 10 |
+
"eval_micro_precision": 0.9565217391304348,
|
| 11 |
+
"eval_micro_recall": 0.38596491228070173,
|
| 12 |
+
"eval_runtime": 0.2248,
|
| 13 |
+
"eval_samples_per_second": 177.928,
|
| 14 |
+
"eval_steps_per_second": 13.345
|
| 15 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"total_flos": 0.0,
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 45.0,
|
| 3 |
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.9705644819471572,
|
| 5 |
+
"train_runtime": 232.6541,
|
| 6 |
+
"train_samples_per_second": 67.912,
|
| 7 |
+
"train_steps_per_second": 4.298
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric":
|
| 4 |
-
"best_model_checkpoint":
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 250,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -12,46 +12,405 @@
|
|
| 12 |
{
|
| 13 |
"epoch": 0,
|
| 14 |
"step": 0,
|
| 15 |
-
"train/classification_loss": 0.
|
| 16 |
-
"train/contrastive_loss":
|
| 17 |
-
"train/negative_loss":
|
| 18 |
"train/num_negatives": 190,
|
| 19 |
"train/num_positives": 50,
|
| 20 |
-
"train/positive_loss":
|
| 21 |
-
"train/total_loss":
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0,
|
| 25 |
"step": 0,
|
| 26 |
-
"train/classification_loss": 0.
|
| 27 |
-
"train/contrastive_loss":
|
| 28 |
-
"train/negative_loss":
|
| 29 |
"train/num_negatives": 192,
|
| 30 |
"train/num_positives": 48,
|
| 31 |
-
"train/positive_loss":
|
| 32 |
-
"train/total_loss":
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 5.0,
|
| 36 |
-
"grad_norm":
|
| 37 |
-
"learning_rate": 9.
|
| 38 |
-
"loss":
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 5.0,
|
| 43 |
"step": 50,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"total_flos": 0.0,
|
| 45 |
-
"train_loss": 1.
|
| 46 |
-
"train_runtime":
|
| 47 |
-
"train_samples_per_second":
|
| 48 |
-
"train_steps_per_second":
|
| 49 |
}
|
| 50 |
],
|
| 51 |
"logging_steps": 50,
|
| 52 |
-
"max_steps":
|
| 53 |
"num_input_tokens_seen": 0,
|
| 54 |
-
"num_train_epochs":
|
| 55 |
"save_steps": 500,
|
| 56 |
"stateful_callbacks": {
|
| 57 |
"EarlyStoppingCallback": {
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 500,
|
| 3 |
+
"best_metric": 0.3380281690140845,
|
| 4 |
+
"best_model_checkpoint": "./bert-philosophy-classifier/checkpoint-500",
|
| 5 |
+
"epoch": 50.0,
|
| 6 |
"eval_steps": 250,
|
| 7 |
+
"global_step": 500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 12 |
{
|
| 13 |
"epoch": 0,
|
| 14 |
"step": 0,
|
| 15 |
+
"train/classification_loss": 0.679158627986908,
|
| 16 |
+
"train/contrastive_loss": 9.516982078552246,
|
| 17 |
+
"train/negative_loss": 9.516908645629883,
|
| 18 |
"train/num_negatives": 190,
|
| 19 |
"train/num_positives": 50,
|
| 20 |
+
"train/positive_loss": 7.310241926461458e-05,
|
| 21 |
+
"train/total_loss": 2.582555055618286
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0,
|
| 25 |
"step": 0,
|
| 26 |
+
"train/classification_loss": 0.6693864464759827,
|
| 27 |
+
"train/contrastive_loss": 9.331222534179688,
|
| 28 |
+
"train/negative_loss": 9.331130981445312,
|
| 29 |
"train/num_negatives": 192,
|
| 30 |
"train/num_positives": 48,
|
| 31 |
+
"train/positive_loss": 9.195055463351309e-05,
|
| 32 |
+
"train/total_loss": 2.535630941390991
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 5.0,
|
| 36 |
+
"grad_norm": 17.358003616333008,
|
| 37 |
+
"learning_rate": 9.600000000000001e-06,
|
| 38 |
+
"loss": 3.6257,
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 5.0,
|
| 43 |
"step": 50,
|
| 44 |
+
"train/classification_loss": 0.6362661719322205,
|
| 45 |
+
"train/contrastive_loss": 1.4868279695510864,
|
| 46 |
+
"train/negative_loss": 1.3825407028198242,
|
| 47 |
+
"train/num_negatives": 170,
|
| 48 |
+
"train/num_positives": 66,
|
| 49 |
+
"train/positive_loss": 0.10428724437952042,
|
| 50 |
+
"train/total_loss": 0.9336317777633667
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 5.0,
|
| 54 |
+
"step": 50,
|
| 55 |
+
"train/classification_loss": 0.6401901245117188,
|
| 56 |
+
"train/contrastive_loss": 1.6219159364700317,
|
| 57 |
+
"train/negative_loss": 1.440779447555542,
|
| 58 |
+
"train/num_negatives": 186,
|
| 59 |
+
"train/num_positives": 54,
|
| 60 |
+
"train/positive_loss": 0.18113651871681213,
|
| 61 |
+
"train/total_loss": 0.964573323726654
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 10.0,
|
| 65 |
+
"grad_norm": 20.2652645111084,
|
| 66 |
+
"learning_rate": 1.9600000000000002e-05,
|
| 67 |
+
"loss": 1.6163,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 10.0,
|
| 72 |
+
"step": 100,
|
| 73 |
+
"train/classification_loss": 0.44592994451522827,
|
| 74 |
+
"train/contrastive_loss": 0.9996287822723389,
|
| 75 |
+
"train/negative_loss": 0.8505972027778625,
|
| 76 |
+
"train/num_negatives": 190,
|
| 77 |
+
"train/num_positives": 50,
|
| 78 |
+
"train/positive_loss": 0.1490315943956375,
|
| 79 |
+
"train/total_loss": 0.645855724811554
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 10.0,
|
| 83 |
+
"step": 100,
|
| 84 |
+
"train/classification_loss": 0.4469062089920044,
|
| 85 |
+
"train/contrastive_loss": 1.1934728622436523,
|
| 86 |
+
"train/negative_loss": 0.9309344291687012,
|
| 87 |
+
"train/num_negatives": 206,
|
| 88 |
+
"train/num_positives": 30,
|
| 89 |
+
"train/positive_loss": 0.26253849267959595,
|
| 90 |
+
"train/total_loss": 0.685600757598877
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 15.0,
|
| 94 |
+
"grad_norm": 8.280580520629883,
|
| 95 |
+
"learning_rate": 1.76e-05,
|
| 96 |
+
"loss": 1.1607,
|
| 97 |
+
"step": 150
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 15.0,
|
| 101 |
+
"step": 150,
|
| 102 |
+
"train/classification_loss": 0.32877856492996216,
|
| 103 |
+
"train/contrastive_loss": 1.1336479187011719,
|
| 104 |
+
"train/negative_loss": 0.9981224536895752,
|
| 105 |
+
"train/num_negatives": 170,
|
| 106 |
+
"train/num_positives": 66,
|
| 107 |
+
"train/positive_loss": 0.1355254054069519,
|
| 108 |
+
"train/total_loss": 0.5555081367492676
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 15.0,
|
| 112 |
+
"step": 150,
|
| 113 |
+
"train/classification_loss": 0.3285317122936249,
|
| 114 |
+
"train/contrastive_loss": 0.8869010210037231,
|
| 115 |
+
"train/negative_loss": 0.591576099395752,
|
| 116 |
+
"train/num_negatives": 202,
|
| 117 |
+
"train/num_positives": 38,
|
| 118 |
+
"train/positive_loss": 0.2953248918056488,
|
| 119 |
+
"train/total_loss": 0.5059119462966919
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 20.0,
|
| 123 |
+
"grad_norm": 7.707197189331055,
|
| 124 |
+
"learning_rate": 1.5100000000000001e-05,
|
| 125 |
+
"loss": 0.9196,
|
| 126 |
+
"step": 200
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 20.0,
|
| 130 |
+
"step": 200,
|
| 131 |
+
"train/classification_loss": 0.293140172958374,
|
| 132 |
+
"train/contrastive_loss": 0.7223706245422363,
|
| 133 |
+
"train/negative_loss": 0.5778605937957764,
|
| 134 |
+
"train/num_negatives": 202,
|
| 135 |
+
"train/num_positives": 30,
|
| 136 |
+
"train/positive_loss": 0.14451001584529877,
|
| 137 |
+
"train/total_loss": 0.4376143217086792
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 20.0,
|
| 141 |
+
"step": 200,
|
| 142 |
+
"train/classification_loss": 0.2644300162792206,
|
| 143 |
+
"train/contrastive_loss": 0.4585617780685425,
|
| 144 |
+
"train/negative_loss": 0.39372602105140686,
|
| 145 |
+
"train/num_negatives": 184,
|
| 146 |
+
"train/num_positives": 56,
|
| 147 |
+
"train/positive_loss": 0.06483575701713562,
|
| 148 |
+
"train/total_loss": 0.3561423718929291
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 25.0,
|
| 152 |
+
"grad_norm": 6.953479766845703,
|
| 153 |
+
"learning_rate": 1.2600000000000001e-05,
|
| 154 |
+
"loss": 0.811,
|
| 155 |
+
"step": 250
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 25.0,
|
| 159 |
+
"step": 250,
|
| 160 |
+
"train/classification_loss": 0.2595597505569458,
|
| 161 |
+
"train/contrastive_loss": 2.3272764682769775,
|
| 162 |
+
"train/negative_loss": 1.8330672979354858,
|
| 163 |
+
"train/num_negatives": 196,
|
| 164 |
+
"train/num_positives": 44,
|
| 165 |
+
"train/positive_loss": 0.4942092299461365,
|
| 166 |
+
"train/total_loss": 0.7250150442123413
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 25.0,
|
| 170 |
+
"step": 250,
|
| 171 |
+
"train/classification_loss": 0.2660799026489258,
|
| 172 |
+
"train/contrastive_loss": 3.3698394298553467,
|
| 173 |
+
"train/negative_loss": 1.8154842853546143,
|
| 174 |
+
"train/num_negatives": 210,
|
| 175 |
+
"train/num_positives": 30,
|
| 176 |
+
"train/positive_loss": 1.5543551445007324,
|
| 177 |
+
"train/total_loss": 0.9400478005409241
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 25.0,
|
| 181 |
+
"step": 250,
|
| 182 |
+
"train/classification_loss": 0.2840481698513031,
|
| 183 |
+
"train/contrastive_loss": 1.1826257705688477,
|
| 184 |
+
"train/negative_loss": 1.1373339891433716,
|
| 185 |
+
"train/num_negatives": 46,
|
| 186 |
+
"train/num_positives": 8,
|
| 187 |
+
"train/positive_loss": 0.045291826128959656,
|
| 188 |
+
"train/total_loss": 0.5205733180046082
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 25.0,
|
| 192 |
+
"eval_exact_match_accuracy": 0.1,
|
| 193 |
+
"eval_hamming_loss": 0.075,
|
| 194 |
+
"eval_loss": 0.7701398134231567,
|
| 195 |
+
"eval_macro_f1": 0.0784313725490196,
|
| 196 |
+
"eval_macro_precision": 0.1092436974789916,
|
| 197 |
+
"eval_macro_recall": 0.06149732620320855,
|
| 198 |
+
"eval_micro_f1": 0.2153846153846154,
|
| 199 |
+
"eval_micro_precision": 0.875,
|
| 200 |
+
"eval_micro_recall": 0.12280701754385964,
|
| 201 |
+
"eval_runtime": 0.219,
|
| 202 |
+
"eval_samples_per_second": 182.685,
|
| 203 |
+
"eval_steps_per_second": 13.701,
|
| 204 |
+
"step": 250
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 25.0,
|
| 208 |
+
"step": 250,
|
| 209 |
+
"train/classification_loss": 0.25078481435775757,
|
| 210 |
+
"train/contrastive_loss": 0.9467111825942993,
|
| 211 |
+
"train/negative_loss": 0.8433182835578918,
|
| 212 |
+
"train/num_negatives": 198,
|
| 213 |
+
"train/num_positives": 40,
|
| 214 |
+
"train/positive_loss": 0.10339287668466568,
|
| 215 |
+
"train/total_loss": 0.44012707471847534
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 25.0,
|
| 219 |
+
"step": 250,
|
| 220 |
+
"train/classification_loss": 0.23322956264019012,
|
| 221 |
+
"train/contrastive_loss": 0.4987642168998718,
|
| 222 |
+
"train/negative_loss": 0.48307880759239197,
|
| 223 |
+
"train/num_negatives": 172,
|
| 224 |
+
"train/num_positives": 68,
|
| 225 |
+
"train/positive_loss": 0.015685414895415306,
|
| 226 |
+
"train/total_loss": 0.3329824209213257
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 30.0,
|
| 230 |
+
"grad_norm": 11.7496976852417,
|
| 231 |
+
"learning_rate": 1.0100000000000002e-05,
|
| 232 |
+
"loss": 0.7395,
|
| 233 |
+
"step": 300
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 30.0,
|
| 237 |
+
"step": 300,
|
| 238 |
+
"train/classification_loss": 0.22414086759090424,
|
| 239 |
+
"train/contrastive_loss": 0.9544009566307068,
|
| 240 |
+
"train/negative_loss": 0.6044885516166687,
|
| 241 |
+
"train/num_negatives": 186,
|
| 242 |
+
"train/num_positives": 54,
|
| 243 |
+
"train/positive_loss": 0.3499124050140381,
|
| 244 |
+
"train/total_loss": 0.41502106189727783
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 30.0,
|
| 248 |
+
"step": 300,
|
| 249 |
+
"train/classification_loss": 0.21396367251873016,
|
| 250 |
+
"train/contrastive_loss": 0.4959838390350342,
|
| 251 |
+
"train/negative_loss": 0.4717627763748169,
|
| 252 |
+
"train/num_negatives": 198,
|
| 253 |
+
"train/num_positives": 42,
|
| 254 |
+
"train/positive_loss": 0.02422107383608818,
|
| 255 |
+
"train/total_loss": 0.3131604492664337
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 35.0,
|
| 259 |
+
"grad_norm": 5.532268047332764,
|
| 260 |
+
"learning_rate": 7.600000000000001e-06,
|
| 261 |
+
"loss": 0.6737,
|
| 262 |
+
"step": 350
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 35.0,
|
| 266 |
+
"step": 350,
|
| 267 |
+
"train/classification_loss": 0.21886315941810608,
|
| 268 |
+
"train/contrastive_loss": 0.5652549266815186,
|
| 269 |
+
"train/negative_loss": 0.4682881832122803,
|
| 270 |
+
"train/num_negatives": 172,
|
| 271 |
+
"train/num_positives": 68,
|
| 272 |
+
"train/positive_loss": 0.09696672856807709,
|
| 273 |
+
"train/total_loss": 0.33191415667533875
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 35.0,
|
| 277 |
+
"step": 350,
|
| 278 |
+
"train/classification_loss": 0.1887310892343521,
|
| 279 |
+
"train/contrastive_loss": 0.18129800260066986,
|
| 280 |
+
"train/negative_loss": 0.17543496191501617,
|
| 281 |
+
"train/num_negatives": 152,
|
| 282 |
+
"train/num_positives": 88,
|
| 283 |
+
"train/positive_loss": 0.005863038823008537,
|
| 284 |
+
"train/total_loss": 0.22499069571495056
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 40.0,
|
| 288 |
+
"grad_norm": 5.668190002441406,
|
| 289 |
+
"learning_rate": 5.1e-06,
|
| 290 |
+
"loss": 0.6269,
|
| 291 |
+
"step": 400
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 40.0,
|
| 295 |
+
"step": 400,
|
| 296 |
+
"train/classification_loss": 0.18238115310668945,
|
| 297 |
+
"train/contrastive_loss": 0.33620232343673706,
|
| 298 |
+
"train/negative_loss": 0.2550373077392578,
|
| 299 |
+
"train/num_negatives": 156,
|
| 300 |
+
"train/num_positives": 84,
|
| 301 |
+
"train/positive_loss": 0.08116500079631805,
|
| 302 |
+
"train/total_loss": 0.24962162971496582
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 40.0,
|
| 306 |
+
"step": 400,
|
| 307 |
+
"train/classification_loss": 0.22312195599079132,
|
| 308 |
+
"train/contrastive_loss": 0.6893786191940308,
|
| 309 |
+
"train/negative_loss": 0.6744635105133057,
|
| 310 |
+
"train/num_negatives": 206,
|
| 311 |
+
"train/num_positives": 34,
|
| 312 |
+
"train/positive_loss": 0.014915116131305695,
|
| 313 |
+
"train/total_loss": 0.36099767684936523
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 45.0,
|
| 317 |
+
"grad_norm": 4.168755054473877,
|
| 318 |
+
"learning_rate": 2.6e-06,
|
| 319 |
+
"loss": 0.6025,
|
| 320 |
+
"step": 450
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 45.0,
|
| 324 |
+
"step": 450,
|
| 325 |
+
"train/classification_loss": 0.1997791826725006,
|
| 326 |
+
"train/contrastive_loss": 0.606022834777832,
|
| 327 |
+
"train/negative_loss": 0.5423316955566406,
|
| 328 |
+
"train/num_negatives": 216,
|
| 329 |
+
"train/num_positives": 24,
|
| 330 |
+
"train/positive_loss": 0.0636911541223526,
|
| 331 |
+
"train/total_loss": 0.32098376750946045
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 45.0,
|
| 335 |
+
"step": 450,
|
| 336 |
+
"train/classification_loss": 0.21638567745685577,
|
| 337 |
+
"train/contrastive_loss": 0.37323933839797974,
|
| 338 |
+
"train/negative_loss": 0.3397449851036072,
|
| 339 |
+
"train/num_negatives": 164,
|
| 340 |
+
"train/num_positives": 62,
|
| 341 |
+
"train/positive_loss": 0.03349434956908226,
|
| 342 |
+
"train/total_loss": 0.291033536195755
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 50.0,
|
| 346 |
+
"grad_norm": 5.069293022155762,
|
| 347 |
+
"learning_rate": 1.0000000000000001e-07,
|
| 348 |
+
"loss": 0.58,
|
| 349 |
+
"step": 500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 50.0,
|
| 353 |
+
"step": 500,
|
| 354 |
+
"train/classification_loss": 0.22550146281719208,
|
| 355 |
+
"train/contrastive_loss": 2.474167823791504,
|
| 356 |
+
"train/negative_loss": 1.7999987602233887,
|
| 357 |
+
"train/num_negatives": 196,
|
| 358 |
+
"train/num_positives": 44,
|
| 359 |
+
"train/positive_loss": 0.6741690635681152,
|
| 360 |
+
"train/total_loss": 0.7203350067138672
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 50.0,
|
| 364 |
+
"step": 500,
|
| 365 |
+
"train/classification_loss": 0.23388545215129852,
|
| 366 |
+
"train/contrastive_loss": 3.272613763809204,
|
| 367 |
+
"train/negative_loss": 1.7668838500976562,
|
| 368 |
+
"train/num_negatives": 210,
|
| 369 |
+
"train/num_positives": 30,
|
| 370 |
+
"train/positive_loss": 1.5057299137115479,
|
| 371 |
+
"train/total_loss": 0.8884082436561584
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 50.0,
|
| 375 |
+
"step": 500,
|
| 376 |
+
"train/classification_loss": 0.2511661648750305,
|
| 377 |
+
"train/contrastive_loss": 0.6579197645187378,
|
| 378 |
+
"train/negative_loss": 0.537192702293396,
|
| 379 |
+
"train/num_negatives": 46,
|
| 380 |
+
"train/num_positives": 8,
|
| 381 |
+
"train/positive_loss": 0.12072707712650299,
|
| 382 |
+
"train/total_loss": 0.38275012373924255
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 50.0,
|
| 386 |
+
"eval_exact_match_accuracy": 0.2,
|
| 387 |
+
"eval_hamming_loss": 0.06911764705882353,
|
| 388 |
+
"eval_loss": 0.7200472950935364,
|
| 389 |
+
"eval_macro_f1": 0.11519607843137254,
|
| 390 |
+
"eval_macro_precision": 0.15826330532212887,
|
| 391 |
+
"eval_macro_recall": 0.09090909090909091,
|
| 392 |
+
"eval_micro_f1": 0.3380281690140845,
|
| 393 |
+
"eval_micro_precision": 0.8571428571428571,
|
| 394 |
+
"eval_micro_recall": 0.21052631578947367,
|
| 395 |
+
"eval_runtime": 0.219,
|
| 396 |
+
"eval_samples_per_second": 182.641,
|
| 397 |
+
"eval_steps_per_second": 13.698,
|
| 398 |
+
"step": 500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 50.0,
|
| 402 |
+
"step": 500,
|
| 403 |
"total_flos": 0.0,
|
| 404 |
+
"train_loss": 1.1355848159790038,
|
| 405 |
+
"train_runtime": 246.5817,
|
| 406 |
+
"train_samples_per_second": 64.076,
|
| 407 |
+
"train_steps_per_second": 2.028
|
| 408 |
}
|
| 409 |
],
|
| 410 |
"logging_steps": 50,
|
| 411 |
+
"max_steps": 500,
|
| 412 |
"num_input_tokens_seen": 0,
|
| 413 |
+
"num_train_epochs": 50,
|
| 414 |
"save_steps": 500,
|
| 415 |
"stateful_callbacks": {
|
| 416 |
"EarlyStoppingCallback": {
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:66e01343304a8027b49b07fccbfd92f2c7fc70a061de471b2412977d28ec9eac
|
| 3 |
size 5368
|