End of training
Browse files- README.md +9 -9
- all_results.json +11 -11
- runs/Jun15_00-47-20_92b2e0e6fb20/events.out.tfevents.1749948709.92b2e0e6fb20.2194.11 +3 -0
- test_results.json +12 -12
- trainer_state.json +1059 -286
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 |
|
|
|
|
| 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: 0.5291
|
| 20 |
+
- Exact Match Accuracy: 0.4
|
| 21 |
+
- Macro Precision: 0.1658
|
| 22 |
+
- Macro Recall: 0.1265
|
| 23 |
+
- Macro F1: 0.1410
|
| 24 |
+
- Micro Precision: 0.92
|
| 25 |
+
- Micro Recall: 0.4035
|
| 26 |
+
- Micro F1: 0.5610
|
| 27 |
+
- Hamming Loss: 0.0529
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
all_results.json
CHANGED
|
@@ -1,17 +1,17 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 45.0,
|
| 3 |
-
"eval_exact_match_accuracy": 0.
|
| 4 |
"eval_hamming_loss": 0.052941176470588235,
|
| 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":
|
| 15 |
"total_flos": 0.0,
|
| 16 |
"train_loss": 0.9705644819471572,
|
| 17 |
"train_runtime": 232.6541,
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 45.0,
|
| 3 |
+
"eval_exact_match_accuracy": 0.4,
|
| 4 |
"eval_hamming_loss": 0.052941176470588235,
|
| 5 |
+
"eval_loss": 0.5290737152099609,
|
| 6 |
+
"eval_macro_f1": 0.14097904608067482,
|
| 7 |
+
"eval_macro_precision": 0.1657754010695187,
|
| 8 |
+
"eval_macro_recall": 0.1264705882352941,
|
| 9 |
+
"eval_micro_f1": 0.5609756097560976,
|
| 10 |
+
"eval_micro_precision": 0.92,
|
| 11 |
+
"eval_micro_recall": 0.40350877192982454,
|
| 12 |
+
"eval_runtime": 0.2121,
|
| 13 |
+
"eval_samples_per_second": 188.615,
|
| 14 |
+
"eval_steps_per_second": 23.577,
|
| 15 |
"total_flos": 0.0,
|
| 16 |
"train_loss": 0.9705644819471572,
|
| 17 |
"train_runtime": 232.6541,
|
runs/Jun15_00-47-20_92b2e0e6fb20/events.out.tfevents.1749948709.92b2e0e6fb20.2194.11
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3a7d599c58099528756c657a42cc2454d32e18e282d12c575b8ebcb6fe242d8
|
| 3 |
+
size 5853
|
test_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_exact_match_accuracy": 0.
|
| 4 |
"eval_hamming_loss": 0.052941176470588235,
|
| 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":
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 45.0,
|
| 3 |
+
"eval_exact_match_accuracy": 0.4,
|
| 4 |
"eval_hamming_loss": 0.052941176470588235,
|
| 5 |
+
"eval_loss": 0.5290737152099609,
|
| 6 |
+
"eval_macro_f1": 0.14097904608067482,
|
| 7 |
+
"eval_macro_precision": 0.1657754010695187,
|
| 8 |
+
"eval_macro_recall": 0.1264705882352941,
|
| 9 |
+
"eval_micro_f1": 0.5609756097560976,
|
| 10 |
+
"eval_micro_precision": 0.92,
|
| 11 |
+
"eval_micro_recall": 0.40350877192982454,
|
| 12 |
+
"eval_runtime": 0.2121,
|
| 13 |
+
"eval_samples_per_second": 188.615,
|
| 14 |
+
"eval_steps_per_second": 23.577
|
| 15 |
}
|
trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint":
|
| 5 |
-
"epoch":
|
| 6 |
-
"eval_steps":
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -12,406 +12,1179 @@
|
|
| 12 |
{
|
| 13 |
"epoch": 0,
|
| 14 |
"step": 0,
|
| 15 |
-
"train/classification_loss": 0.
|
| 16 |
-
"train/contrastive_loss": 9.
|
| 17 |
-
"train/negative_loss": 9.
|
| 18 |
-
"train/num_negatives":
|
| 19 |
-
"train/num_positives":
|
| 20 |
-
"train/positive_loss": 7.
|
| 21 |
-
"train/total_loss": 2.
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0,
|
| 25 |
"step": 0,
|
| 26 |
-
"train/classification_loss": 0.
|
| 27 |
-
"train/contrastive_loss": 9.
|
| 28 |
-
"train/negative_loss": 9.
|
| 29 |
-
"train/num_negatives":
|
| 30 |
-
"train/num_positives":
|
| 31 |
-
"train/positive_loss":
|
| 32 |
-
"train/total_loss": 2.
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 5
|
| 36 |
-
"grad_norm":
|
| 37 |
"learning_rate": 9.600000000000001e-06,
|
| 38 |
-
"loss": 3.
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 5
|
| 43 |
"step": 50,
|
| 44 |
-
"train/classification_loss": 0.
|
| 45 |
-
"train/contrastive_loss": 1.
|
| 46 |
-
"train/negative_loss": 1.
|
| 47 |
-
"train/num_negatives":
|
| 48 |
-
"train/num_positives":
|
| 49 |
-
"train/positive_loss": 0.
|
| 50 |
-
"train/total_loss":
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 5
|
| 54 |
"step": 50,
|
| 55 |
-
"train/classification_loss": 0.
|
| 56 |
-
"train/contrastive_loss": 1.
|
| 57 |
-
"train/negative_loss": 1.
|
| 58 |
-
"train/num_negatives":
|
| 59 |
-
"train/num_positives":
|
| 60 |
-
"train/positive_loss": 0.
|
| 61 |
-
"train/total_loss": 0.
|
| 62 |
},
|
| 63 |
{
|
| 64 |
-
"epoch":
|
| 65 |
-
"grad_norm":
|
| 66 |
"learning_rate": 1.9600000000000002e-05,
|
| 67 |
-
"loss": 1.
|
| 68 |
"step": 100
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch":
|
| 72 |
"step": 100,
|
| 73 |
-
"train/classification_loss": 0.
|
| 74 |
-
"train/contrastive_loss":
|
| 75 |
-
"train/negative_loss":
|
| 76 |
-
"train/num_negatives":
|
| 77 |
-
"train/num_positives":
|
| 78 |
-
"train/positive_loss": 0.
|
| 79 |
-
"train/total_loss": 0.
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
"step": 100,
|
| 84 |
-
"train/classification_loss": 0.
|
| 85 |
-
"train/contrastive_loss":
|
| 86 |
-
"train/negative_loss":
|
| 87 |
-
"train/num_negatives":
|
| 88 |
-
"train/num_positives":
|
| 89 |
-
"train/positive_loss": 0.
|
| 90 |
-
"train/total_loss":
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch":
|
| 94 |
-
"
|
| 95 |
-
"
|
| 96 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
"step": 150
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch":
|
| 101 |
"step": 150,
|
| 102 |
-
"train/classification_loss": 0.
|
| 103 |
-
"train/contrastive_loss": 1.
|
| 104 |
-
"train/negative_loss": 0.
|
| 105 |
-
"train/num_negatives":
|
| 106 |
-
"train/num_positives":
|
| 107 |
-
"train/positive_loss": 0.
|
| 108 |
-
"train/total_loss": 0.
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch":
|
| 112 |
"step": 150,
|
| 113 |
-
"train/classification_loss": 0.
|
| 114 |
-
"train/contrastive_loss":
|
| 115 |
-
"train/negative_loss":
|
| 116 |
-
"train/num_negatives":
|
| 117 |
-
"train/num_positives":
|
| 118 |
-
"train/positive_loss": 0.
|
| 119 |
-
"train/total_loss": 0.
|
| 120 |
},
|
| 121 |
{
|
| 122 |
-
"epoch":
|
| 123 |
-
"grad_norm":
|
| 124 |
-
"learning_rate": 1.
|
| 125 |
-
"loss":
|
| 126 |
"step": 200
|
| 127 |
},
|
| 128 |
{
|
| 129 |
-
"epoch":
|
| 130 |
"step": 200,
|
| 131 |
-
"train/classification_loss": 0.
|
| 132 |
-
"train/contrastive_loss":
|
| 133 |
-
"train/negative_loss":
|
| 134 |
-
"train/num_negatives":
|
| 135 |
-
"train/num_positives":
|
| 136 |
-
"train/positive_loss": 0.
|
| 137 |
-
"train/total_loss": 0.
|
| 138 |
},
|
| 139 |
{
|
| 140 |
-
"epoch":
|
| 141 |
"step": 200,
|
| 142 |
-
"train/classification_loss": 0.
|
| 143 |
-
"train/contrastive_loss":
|
| 144 |
-
"train/negative_loss":
|
| 145 |
-
"train/num_negatives":
|
| 146 |
-
"train/num_positives":
|
| 147 |
-
"train/positive_loss": 0.
|
| 148 |
-
"train/total_loss": 0.
|
| 149 |
},
|
| 150 |
{
|
| 151 |
-
"epoch":
|
| 152 |
-
"
|
| 153 |
-
"
|
| 154 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 155 |
"step": 250
|
| 156 |
},
|
| 157 |
{
|
| 158 |
-
"epoch":
|
| 159 |
"step": 250,
|
| 160 |
-
"train/classification_loss": 0.
|
| 161 |
-
"train/contrastive_loss":
|
| 162 |
-
"train/negative_loss":
|
| 163 |
-
"train/num_negatives":
|
| 164 |
-
"train/num_positives":
|
| 165 |
-
"train/positive_loss": 0.
|
| 166 |
-
"train/total_loss": 0.
|
| 167 |
},
|
| 168 |
{
|
| 169 |
-
"epoch":
|
| 170 |
"step": 250,
|
| 171 |
-
"train/classification_loss": 0.
|
| 172 |
-
"train/contrastive_loss":
|
| 173 |
-
"train/negative_loss":
|
| 174 |
-
"train/num_negatives":
|
| 175 |
-
"train/num_positives":
|
| 176 |
-
"train/positive_loss":
|
| 177 |
-
"train/total_loss": 0.
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch":
|
| 181 |
-
"
|
| 182 |
-
"
|
| 183 |
-
"
|
| 184 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
"train/num_negatives": 46,
|
| 186 |
"train/num_positives": 8,
|
| 187 |
-
"train/positive_loss": 0.
|
| 188 |
-
"train/total_loss": 0.
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 192 |
"eval_exact_match_accuracy": 0.1,
|
| 193 |
"eval_hamming_loss": 0.075,
|
| 194 |
-
"eval_loss": 0.
|
| 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.
|
| 202 |
-
"eval_samples_per_second":
|
| 203 |
-
"eval_steps_per_second":
|
| 204 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 25.0,
|
| 208 |
-
"
|
| 209 |
-
"
|
| 210 |
-
"
|
| 211 |
-
"
|
| 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":
|
| 220 |
-
"train/classification_loss": 0.
|
| 221 |
-
"train/contrastive_loss":
|
| 222 |
-
"train/negative_loss":
|
| 223 |
-
"train/num_negatives":
|
| 224 |
-
"train/num_positives":
|
| 225 |
-
"train/positive_loss": 0.
|
| 226 |
-
"train/total_loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 30.0,
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate":
|
| 232 |
-
"loss": 0.
|
| 233 |
-
"step":
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 30.0,
|
| 237 |
-
"step":
|
| 238 |
-
"train/classification_loss": 0.
|
| 239 |
-
"train/contrastive_loss":
|
| 240 |
-
"train/negative_loss":
|
| 241 |
-
"train/num_negatives":
|
| 242 |
-
"train/num_positives":
|
| 243 |
-
"train/positive_loss": 0.
|
| 244 |
-
"train/total_loss": 0.
|
| 245 |
},
|
| 246 |
{
|
| 247 |
"epoch": 30.0,
|
| 248 |
-
"step":
|
| 249 |
-
"train/classification_loss": 0.
|
| 250 |
-
"train/contrastive_loss":
|
| 251 |
-
"train/negative_loss":
|
| 252 |
-
"train/num_negatives":
|
| 253 |
-
"train/num_positives":
|
| 254 |
-
"train/positive_loss": 0.
|
| 255 |
-
"train/total_loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 35.0,
|
| 259 |
-
"grad_norm": 5.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
-
"step":
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 35.0,
|
| 266 |
-
"step":
|
| 267 |
-
"train/classification_loss": 0.
|
| 268 |
-
"train/contrastive_loss":
|
| 269 |
-
"train/negative_loss":
|
| 270 |
-
"train/num_negatives":
|
| 271 |
-
"train/num_positives":
|
| 272 |
-
"train/positive_loss":
|
| 273 |
-
"train/total_loss": 0.
|
| 274 |
},
|
| 275 |
{
|
| 276 |
"epoch": 35.0,
|
| 277 |
-
"step":
|
| 278 |
-
"train/classification_loss": 0.
|
| 279 |
-
"train/contrastive_loss":
|
| 280 |
-
"train/negative_loss":
|
| 281 |
-
"train/num_negatives":
|
| 282 |
-
"train/num_positives":
|
| 283 |
-
"train/positive_loss": 0.
|
| 284 |
-
"train/total_loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 40.0,
|
| 288 |
-
"grad_norm":
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss": 0.
|
| 291 |
-
"step":
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 40.0,
|
| 295 |
-
"step":
|
| 296 |
-
"train/classification_loss": 0.
|
| 297 |
-
"train/contrastive_loss":
|
| 298 |
-
"train/negative_loss": 0.
|
| 299 |
-
"train/num_negatives":
|
| 300 |
-
"train/num_positives":
|
| 301 |
-
"train/positive_loss": 0.
|
| 302 |
-
"train/total_loss": 0.
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 40.0,
|
| 306 |
-
"step":
|
| 307 |
-
"train/classification_loss": 0.
|
| 308 |
-
"train/contrastive_loss":
|
| 309 |
-
"train/negative_loss":
|
| 310 |
-
"train/num_negatives":
|
| 311 |
-
"train/num_positives":
|
| 312 |
-
"train/positive_loss": 0.
|
| 313 |
-
"train/total_loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 45.0,
|
| 317 |
-
"grad_norm":
|
| 318 |
-
"learning_rate": 2.
|
| 319 |
-
"loss": 0.
|
| 320 |
-
"step":
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 45.0,
|
| 324 |
-
"step":
|
| 325 |
-
"train/classification_loss": 0.
|
| 326 |
-
"train/contrastive_loss":
|
| 327 |
-
"train/negative_loss": 0.
|
| 328 |
-
"train/num_negatives":
|
| 329 |
-
"train/num_positives":
|
| 330 |
-
"train/positive_loss":
|
| 331 |
-
"train/total_loss": 0.
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 45.0,
|
| 335 |
-
"step":
|
| 336 |
-
"train/classification_loss": 0.
|
| 337 |
-
"train/contrastive_loss":
|
| 338 |
-
"train/negative_loss":
|
| 339 |
-
"train/num_negatives":
|
| 340 |
-
"train/num_positives":
|
| 341 |
-
"train/positive_loss": 0.
|
| 342 |
-
"train/total_loss": 0.
|
| 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":
|
| 353 |
-
"step":
|
| 354 |
-
"train/classification_loss": 0.
|
| 355 |
-
"train/contrastive_loss":
|
| 356 |
-
"train/negative_loss":
|
| 357 |
-
"train/num_negatives":
|
| 358 |
-
"train/num_positives":
|
| 359 |
-
"train/positive_loss":
|
| 360 |
-
"train/total_loss":
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch":
|
| 364 |
-
"step":
|
| 365 |
-
"train/classification_loss": 0.
|
| 366 |
-
"train/contrastive_loss": 3.
|
| 367 |
-
"train/negative_loss":
|
| 368 |
-
"train/num_negatives":
|
| 369 |
-
"train/num_positives":
|
| 370 |
-
"train/positive_loss":
|
| 371 |
-
"train/total_loss": 0.
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch":
|
| 375 |
-
"step":
|
| 376 |
-
"train/classification_loss": 0.
|
| 377 |
-
"train/contrastive_loss":
|
| 378 |
-
"train/negative_loss": 0.
|
| 379 |
"train/num_negatives": 46,
|
| 380 |
"train/num_positives": 8,
|
| 381 |
-
"train/positive_loss":
|
| 382 |
-
"train/total_loss": 0.
|
| 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":
|
| 402 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 403 |
"total_flos": 0.0,
|
| 404 |
-
"train_loss":
|
| 405 |
-
"train_runtime":
|
| 406 |
-
"train_samples_per_second":
|
| 407 |
-
"train_steps_per_second":
|
| 408 |
}
|
| 409 |
],
|
| 410 |
"logging_steps": 50,
|
| 411 |
-
"max_steps":
|
| 412 |
"num_input_tokens_seen": 0,
|
| 413 |
"num_train_epochs": 50,
|
| 414 |
-
"save_steps":
|
| 415 |
"stateful_callbacks": {
|
| 416 |
"EarlyStoppingCallback": {
|
| 417 |
"args": {
|
|
@@ -427,14 +1200,14 @@
|
|
| 427 |
"should_epoch_stop": false,
|
| 428 |
"should_evaluate": false,
|
| 429 |
"should_log": false,
|
| 430 |
-
"should_save":
|
| 431 |
-
"should_training_stop":
|
| 432 |
},
|
| 433 |
"attributes": {}
|
| 434 |
}
|
| 435 |
},
|
| 436 |
"total_flos": 0.0,
|
| 437 |
-
"train_batch_size":
|
| 438 |
"trial_name": null,
|
| 439 |
"trial_params": null
|
| 440 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 600,
|
| 3 |
+
"best_metric": 0.42105263157894735,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 45.0,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 900,
|
| 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": 0.7007833123207092,
|
| 16 |
+
"train/contrastive_loss": 9.572936058044434,
|
| 17 |
+
"train/negative_loss": 9.572856903076172,
|
| 18 |
+
"train/num_negatives": 46,
|
| 19 |
+
"train/num_positives": 10,
|
| 20 |
+
"train/positive_loss": 7.908708357717842e-05,
|
| 21 |
+
"train/total_loss": 2.615370512008667
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0,
|
| 25 |
"step": 0,
|
| 26 |
+
"train/classification_loss": 0.7040252089500427,
|
| 27 |
+
"train/contrastive_loss": 9.725648880004883,
|
| 28 |
+
"train/negative_loss": 9.725597381591797,
|
| 29 |
+
"train/num_negatives": 46,
|
| 30 |
+
"train/num_positives": 10,
|
| 31 |
+
"train/positive_loss": 5.152364246896468e-05,
|
| 32 |
+
"train/total_loss": 2.6491549015045166
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 2.5,
|
| 36 |
+
"grad_norm": 18.165754318237305,
|
| 37 |
"learning_rate": 9.600000000000001e-06,
|
| 38 |
+
"loss": 3.7884,
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 2.5,
|
| 43 |
"step": 50,
|
| 44 |
+
"train/classification_loss": 0.6516271233558655,
|
| 45 |
+
"train/contrastive_loss": 1.8623473644256592,
|
| 46 |
+
"train/negative_loss": 1.4998806715011597,
|
| 47 |
+
"train/num_negatives": 38,
|
| 48 |
+
"train/num_positives": 18,
|
| 49 |
+
"train/positive_loss": 0.36246663331985474,
|
| 50 |
+
"train/total_loss": 1.0240966081619263
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 2.5,
|
| 54 |
"step": 50,
|
| 55 |
+
"train/classification_loss": 0.6451266407966614,
|
| 56 |
+
"train/contrastive_loss": 1.7446768283843994,
|
| 57 |
+
"train/negative_loss": 1.668500304222107,
|
| 58 |
+
"train/num_negatives": 50,
|
| 59 |
+
"train/num_positives": 6,
|
| 60 |
+
"train/positive_loss": 0.07617650926113129,
|
| 61 |
+
"train/total_loss": 0.9940620064735413
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 5.0,
|
| 65 |
+
"grad_norm": 16.30499839782715,
|
| 66 |
"learning_rate": 1.9600000000000002e-05,
|
| 67 |
+
"loss": 1.7889,
|
| 68 |
"step": 100
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 5.0,
|
| 72 |
"step": 100,
|
| 73 |
+
"train/classification_loss": 0.4927652180194855,
|
| 74 |
+
"train/contrastive_loss": 2.06559419631958,
|
| 75 |
+
"train/negative_loss": 1.8223981857299805,
|
| 76 |
+
"train/num_negatives": 44,
|
| 77 |
+
"train/num_positives": 12,
|
| 78 |
+
"train/positive_loss": 0.24319612979888916,
|
| 79 |
+
"train/total_loss": 0.9058840274810791
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 5.0,
|
| 83 |
"step": 100,
|
| 84 |
+
"train/classification_loss": 0.5166366100311279,
|
| 85 |
+
"train/contrastive_loss": 3.1131491661071777,
|
| 86 |
+
"train/negative_loss": 3.112612724304199,
|
| 87 |
+
"train/num_negatives": 44,
|
| 88 |
+
"train/num_positives": 12,
|
| 89 |
+
"train/positive_loss": 0.0005365243996493518,
|
| 90 |
+
"train/total_loss": 1.1392664909362793
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"epoch": 5.0,
|
| 94 |
+
"step": 100,
|
| 95 |
+
"train/classification_loss": 0.4922243058681488,
|
| 96 |
+
"train/contrastive_loss": 1.7773994207382202,
|
| 97 |
+
"train/negative_loss": 1.7757622003555298,
|
| 98 |
+
"train/num_negatives": 52,
|
| 99 |
+
"train/num_positives": 4,
|
| 100 |
+
"train/positive_loss": 0.0016372093232348561,
|
| 101 |
+
"train/total_loss": 0.8477041721343994
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 5.0,
|
| 105 |
+
"step": 100,
|
| 106 |
+
"train/classification_loss": 0.5021852850914001,
|
| 107 |
+
"train/contrastive_loss": 3.2608611583709717,
|
| 108 |
+
"train/negative_loss": 2.6507816314697266,
|
| 109 |
+
"train/num_negatives": 48,
|
| 110 |
+
"train/num_positives": 8,
|
| 111 |
+
"train/positive_loss": 0.6100795269012451,
|
| 112 |
+
"train/total_loss": 1.1543575525283813
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 5.0,
|
| 116 |
+
"step": 100,
|
| 117 |
+
"train/classification_loss": 0.4944652020931244,
|
| 118 |
+
"train/contrastive_loss": 2.343458652496338,
|
| 119 |
+
"train/negative_loss": 1.1261494159698486,
|
| 120 |
+
"train/num_negatives": 46,
|
| 121 |
+
"train/num_positives": 8,
|
| 122 |
+
"train/positive_loss": 1.2173092365264893,
|
| 123 |
+
"train/total_loss": 0.9631569385528564
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 5.0,
|
| 127 |
+
"eval_exact_match_accuracy": 0.0,
|
| 128 |
+
"eval_hamming_loss": 0.08529411764705883,
|
| 129 |
+
"eval_loss": 1.002073884010315,
|
| 130 |
+
"eval_macro_f1": 0.0,
|
| 131 |
+
"eval_macro_precision": 0.0,
|
| 132 |
+
"eval_macro_recall": 0.0,
|
| 133 |
+
"eval_micro_f1": 0.0,
|
| 134 |
+
"eval_micro_precision": 0.0,
|
| 135 |
+
"eval_micro_recall": 0.0,
|
| 136 |
+
"eval_runtime": 0.2122,
|
| 137 |
+
"eval_samples_per_second": 188.524,
|
| 138 |
+
"eval_steps_per_second": 23.566,
|
| 139 |
+
"step": 100
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 5.0,
|
| 143 |
+
"step": 100,
|
| 144 |
+
"train/classification_loss": 0.5136489272117615,
|
| 145 |
+
"train/contrastive_loss": 1.3931580781936646,
|
| 146 |
+
"train/negative_loss": 1.0857487916946411,
|
| 147 |
+
"train/num_negatives": 42,
|
| 148 |
+
"train/num_positives": 12,
|
| 149 |
+
"train/positive_loss": 0.30740925669670105,
|
| 150 |
+
"train/total_loss": 0.7922805547714233
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 5.0,
|
| 154 |
+
"step": 100,
|
| 155 |
+
"train/classification_loss": 0.5019432902336121,
|
| 156 |
+
"train/contrastive_loss": 0.7777740955352783,
|
| 157 |
+
"train/negative_loss": 0.5738040804862976,
|
| 158 |
+
"train/num_negatives": 40,
|
| 159 |
+
"train/num_positives": 16,
|
| 160 |
+
"train/positive_loss": 0.20396998524665833,
|
| 161 |
+
"train/total_loss": 0.6574981212615967
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 7.5,
|
| 165 |
+
"grad_norm": 12.862069129943848,
|
| 166 |
+
"learning_rate": 1.8933333333333334e-05,
|
| 167 |
+
"loss": 1.396,
|
| 168 |
"step": 150
|
| 169 |
},
|
| 170 |
{
|
| 171 |
+
"epoch": 7.5,
|
| 172 |
"step": 150,
|
| 173 |
+
"train/classification_loss": 0.3993939161300659,
|
| 174 |
+
"train/contrastive_loss": 1.1298128366470337,
|
| 175 |
+
"train/negative_loss": 0.8979541659355164,
|
| 176 |
+
"train/num_negatives": 44,
|
| 177 |
+
"train/num_positives": 8,
|
| 178 |
+
"train/positive_loss": 0.23185870051383972,
|
| 179 |
+
"train/total_loss": 0.6253564953804016
|
| 180 |
},
|
| 181 |
{
|
| 182 |
+
"epoch": 7.5,
|
| 183 |
"step": 150,
|
| 184 |
+
"train/classification_loss": 0.393746018409729,
|
| 185 |
+
"train/contrastive_loss": 1.2817302942276,
|
| 186 |
+
"train/negative_loss": 1.1390491724014282,
|
| 187 |
+
"train/num_negatives": 44,
|
| 188 |
+
"train/num_positives": 10,
|
| 189 |
+
"train/positive_loss": 0.14268112182617188,
|
| 190 |
+
"train/total_loss": 0.6500921249389648
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"epoch": 10.0,
|
| 194 |
+
"grad_norm": 23.76511573791504,
|
| 195 |
+
"learning_rate": 1.782222222222222e-05,
|
| 196 |
+
"loss": 1.156,
|
| 197 |
"step": 200
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 10.0,
|
| 201 |
"step": 200,
|
| 202 |
+
"train/classification_loss": 0.30127307772636414,
|
| 203 |
+
"train/contrastive_loss": 1.8966163396835327,
|
| 204 |
+
"train/negative_loss": 1.7765212059020996,
|
| 205 |
+
"train/num_negatives": 44,
|
| 206 |
+
"train/num_positives": 12,
|
| 207 |
+
"train/positive_loss": 0.12009509652853012,
|
| 208 |
+
"train/total_loss": 0.6805963516235352
|
| 209 |
},
|
| 210 |
{
|
| 211 |
+
"epoch": 10.0,
|
| 212 |
"step": 200,
|
| 213 |
+
"train/classification_loss": 0.3545684218406677,
|
| 214 |
+
"train/contrastive_loss": 3.0673155784606934,
|
| 215 |
+
"train/negative_loss": 3.067002058029175,
|
| 216 |
+
"train/num_negatives": 44,
|
| 217 |
+
"train/num_positives": 12,
|
| 218 |
+
"train/positive_loss": 0.0003136020968668163,
|
| 219 |
+
"train/total_loss": 0.9680315256118774
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 10.0,
|
| 223 |
+
"step": 200,
|
| 224 |
+
"train/classification_loss": 0.32993215322494507,
|
| 225 |
+
"train/contrastive_loss": 2.723947048187256,
|
| 226 |
+
"train/negative_loss": 2.7232000827789307,
|
| 227 |
+
"train/num_negatives": 52,
|
| 228 |
+
"train/num_positives": 4,
|
| 229 |
+
"train/positive_loss": 0.0007470683194696903,
|
| 230 |
+
"train/total_loss": 0.8747215867042542
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 10.0,
|
| 234 |
+
"step": 200,
|
| 235 |
+
"train/classification_loss": 0.32662659883499146,
|
| 236 |
+
"train/contrastive_loss": 3.6181204319000244,
|
| 237 |
+
"train/negative_loss": 3.4868171215057373,
|
| 238 |
+
"train/num_negatives": 48,
|
| 239 |
+
"train/num_positives": 8,
|
| 240 |
+
"train/positive_loss": 0.1313033103942871,
|
| 241 |
+
"train/total_loss": 1.050250768661499
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 10.0,
|
| 245 |
+
"step": 200,
|
| 246 |
+
"train/classification_loss": 0.33877384662628174,
|
| 247 |
+
"train/contrastive_loss": 2.0167577266693115,
|
| 248 |
+
"train/negative_loss": 1.4009835720062256,
|
| 249 |
+
"train/num_negatives": 46,
|
| 250 |
+
"train/num_positives": 8,
|
| 251 |
+
"train/positive_loss": 0.6157740950584412,
|
| 252 |
+
"train/total_loss": 0.742125391960144
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 10.0,
|
| 256 |
+
"eval_exact_match_accuracy": 0.0,
|
| 257 |
+
"eval_hamming_loss": 0.0838235294117647,
|
| 258 |
+
"eval_loss": 0.8631451725959778,
|
| 259 |
+
"eval_macro_f1": 0.0,
|
| 260 |
+
"eval_macro_precision": 0.0,
|
| 261 |
+
"eval_macro_recall": 0.0,
|
| 262 |
+
"eval_micro_f1": 0.0,
|
| 263 |
+
"eval_micro_precision": 0.0,
|
| 264 |
+
"eval_micro_recall": 0.0,
|
| 265 |
+
"eval_runtime": 0.202,
|
| 266 |
+
"eval_samples_per_second": 198.018,
|
| 267 |
+
"eval_steps_per_second": 24.752,
|
| 268 |
+
"step": 200
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 10.0,
|
| 272 |
+
"step": 200,
|
| 273 |
+
"train/classification_loss": 0.34425634145736694,
|
| 274 |
+
"train/contrastive_loss": 1.217971682548523,
|
| 275 |
+
"train/negative_loss": 0.8478565216064453,
|
| 276 |
+
"train/num_negatives": 42,
|
| 277 |
+
"train/num_positives": 14,
|
| 278 |
+
"train/positive_loss": 0.37011516094207764,
|
| 279 |
+
"train/total_loss": 0.5878506898880005
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 10.0,
|
| 283 |
+
"step": 200,
|
| 284 |
+
"train/classification_loss": 0.3122542202472687,
|
| 285 |
+
"train/contrastive_loss": 0.6875693798065186,
|
| 286 |
+
"train/negative_loss": 0.686568558216095,
|
| 287 |
+
"train/num_negatives": 42,
|
| 288 |
+
"train/num_positives": 14,
|
| 289 |
+
"train/positive_loss": 0.0010008324170485139,
|
| 290 |
+
"train/total_loss": 0.4497680962085724
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 12.5,
|
| 294 |
+
"grad_norm": 7.754025459289551,
|
| 295 |
+
"learning_rate": 1.6711111111111112e-05,
|
| 296 |
+
"loss": 1.0042,
|
| 297 |
"step": 250
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 12.5,
|
| 301 |
"step": 250,
|
| 302 |
+
"train/classification_loss": 0.24464763700962067,
|
| 303 |
+
"train/contrastive_loss": 0.6364108920097351,
|
| 304 |
+
"train/negative_loss": 0.4732590615749359,
|
| 305 |
+
"train/num_negatives": 44,
|
| 306 |
+
"train/num_positives": 12,
|
| 307 |
+
"train/positive_loss": 0.163151815533638,
|
| 308 |
+
"train/total_loss": 0.3719298243522644
|
| 309 |
},
|
| 310 |
{
|
| 311 |
+
"epoch": 12.5,
|
| 312 |
"step": 250,
|
| 313 |
+
"train/classification_loss": 0.3077278137207031,
|
| 314 |
+
"train/contrastive_loss": 0.4854884147644043,
|
| 315 |
+
"train/negative_loss": 0.33287519216537476,
|
| 316 |
+
"train/num_negatives": 42,
|
| 317 |
+
"train/num_positives": 8,
|
| 318 |
+
"train/positive_loss": 0.15261322259902954,
|
| 319 |
+
"train/total_loss": 0.40482550859451294
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 15.0,
|
| 323 |
+
"grad_norm": 16.385419845581055,
|
| 324 |
+
"learning_rate": 1.5600000000000003e-05,
|
| 325 |
+
"loss": 0.8775,
|
| 326 |
+
"step": 300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 15.0,
|
| 330 |
+
"step": 300,
|
| 331 |
+
"train/classification_loss": 0.23176752030849457,
|
| 332 |
+
"train/contrastive_loss": 3.0300073623657227,
|
| 333 |
+
"train/negative_loss": 2.5657095909118652,
|
| 334 |
+
"train/num_negatives": 44,
|
| 335 |
+
"train/num_positives": 12,
|
| 336 |
+
"train/positive_loss": 0.4642978310585022,
|
| 337 |
+
"train/total_loss": 0.8377690315246582
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 15.0,
|
| 341 |
+
"step": 300,
|
| 342 |
+
"train/classification_loss": 0.3002067804336548,
|
| 343 |
+
"train/contrastive_loss": 3.911193609237671,
|
| 344 |
+
"train/negative_loss": 3.911159038543701,
|
| 345 |
+
"train/num_negatives": 44,
|
| 346 |
+
"train/num_positives": 12,
|
| 347 |
+
"train/positive_loss": 3.4572090953588486e-05,
|
| 348 |
+
"train/total_loss": 1.082445502281189
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 15.0,
|
| 352 |
+
"step": 300,
|
| 353 |
+
"train/classification_loss": 0.2759508788585663,
|
| 354 |
+
"train/contrastive_loss": 2.758004665374756,
|
| 355 |
+
"train/negative_loss": 2.7472095489501953,
|
| 356 |
+
"train/num_negatives": 52,
|
| 357 |
+
"train/num_positives": 4,
|
| 358 |
+
"train/positive_loss": 0.010795066133141518,
|
| 359 |
+
"train/total_loss": 0.8275518417358398
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 15.0,
|
| 363 |
+
"step": 300,
|
| 364 |
+
"train/classification_loss": 0.27436333894729614,
|
| 365 |
+
"train/contrastive_loss": 4.6799540519714355,
|
| 366 |
+
"train/negative_loss": 4.251977920532227,
|
| 367 |
+
"train/num_negatives": 48,
|
| 368 |
+
"train/num_positives": 8,
|
| 369 |
+
"train/positive_loss": 0.4279760420322418,
|
| 370 |
+
"train/total_loss": 1.2103540897369385
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 15.0,
|
| 374 |
+
"step": 300,
|
| 375 |
+
"train/classification_loss": 0.2907729744911194,
|
| 376 |
+
"train/contrastive_loss": 2.066659927368164,
|
| 377 |
+
"train/negative_loss": 1.4751646518707275,
|
| 378 |
"train/num_negatives": 46,
|
| 379 |
"train/num_positives": 8,
|
| 380 |
+
"train/positive_loss": 0.5914952754974365,
|
| 381 |
+
"train/total_loss": 0.7041049599647522
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 15.0,
|
| 385 |
+
"eval_exact_match_accuracy": 0.05,
|
| 386 |
+
"eval_hamming_loss": 0.07647058823529412,
|
| 387 |
+
"eval_loss": 0.9324451684951782,
|
| 388 |
+
"eval_macro_f1": 0.03676470588235294,
|
| 389 |
+
"eval_macro_precision": 0.058823529411764705,
|
| 390 |
+
"eval_macro_recall": 0.026737967914438502,
|
| 391 |
+
"eval_micro_f1": 0.16129032258064516,
|
| 392 |
+
"eval_micro_precision": 1.0,
|
| 393 |
+
"eval_micro_recall": 0.08771929824561403,
|
| 394 |
+
"eval_runtime": 0.207,
|
| 395 |
+
"eval_samples_per_second": 193.242,
|
| 396 |
+
"eval_steps_per_second": 24.155,
|
| 397 |
+
"step": 300
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 15.0,
|
| 401 |
+
"step": 300,
|
| 402 |
+
"train/classification_loss": 0.26088976860046387,
|
| 403 |
+
"train/contrastive_loss": 0.5876651406288147,
|
| 404 |
+
"train/negative_loss": 0.5801823139190674,
|
| 405 |
+
"train/num_negatives": 32,
|
| 406 |
+
"train/num_positives": 20,
|
| 407 |
+
"train/positive_loss": 0.007482839282602072,
|
| 408 |
+
"train/total_loss": 0.3784227967262268
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 15.0,
|
| 412 |
+
"step": 300,
|
| 413 |
+
"train/classification_loss": 0.2548993229866028,
|
| 414 |
+
"train/contrastive_loss": 1.50383722782135,
|
| 415 |
+
"train/negative_loss": 1.4029262065887451,
|
| 416 |
+
"train/num_negatives": 46,
|
| 417 |
+
"train/num_positives": 10,
|
| 418 |
+
"train/positive_loss": 0.10091102123260498,
|
| 419 |
+
"train/total_loss": 0.5556668043136597
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 17.5,
|
| 423 |
+
"grad_norm": 16.8145809173584,
|
| 424 |
+
"learning_rate": 1.448888888888889e-05,
|
| 425 |
+
"loss": 0.827,
|
| 426 |
+
"step": 350
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 17.5,
|
| 430 |
+
"step": 350,
|
| 431 |
+
"train/classification_loss": 0.24073848128318787,
|
| 432 |
+
"train/contrastive_loss": 0.6602705717086792,
|
| 433 |
+
"train/negative_loss": 0.4261236786842346,
|
| 434 |
+
"train/num_negatives": 42,
|
| 435 |
+
"train/num_positives": 14,
|
| 436 |
+
"train/positive_loss": 0.23414692282676697,
|
| 437 |
+
"train/total_loss": 0.3727926015853882
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 17.5,
|
| 441 |
+
"step": 350,
|
| 442 |
+
"train/classification_loss": 0.2786425054073334,
|
| 443 |
+
"train/contrastive_loss": 0.29787567257881165,
|
| 444 |
+
"train/negative_loss": 0.2970171868801117,
|
| 445 |
+
"train/num_negatives": 44,
|
| 446 |
+
"train/num_positives": 6,
|
| 447 |
+
"train/positive_loss": 0.0008584868628531694,
|
| 448 |
+
"train/total_loss": 0.3382176458835602
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 20.0,
|
| 452 |
+
"grad_norm": 18.7548828125,
|
| 453 |
+
"learning_rate": 1.3377777777777778e-05,
|
| 454 |
+
"loss": 0.7747,
|
| 455 |
+
"step": 400
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 20.0,
|
| 459 |
+
"step": 400,
|
| 460 |
+
"train/classification_loss": 0.20234902203083038,
|
| 461 |
+
"train/contrastive_loss": 1.7551348209381104,
|
| 462 |
+
"train/negative_loss": 1.1972665786743164,
|
| 463 |
+
"train/num_negatives": 44,
|
| 464 |
+
"train/num_positives": 12,
|
| 465 |
+
"train/positive_loss": 0.557868242263794,
|
| 466 |
+
"train/total_loss": 0.5533760190010071
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 20.0,
|
| 470 |
+
"step": 400,
|
| 471 |
+
"train/classification_loss": 0.27940884232521057,
|
| 472 |
+
"train/contrastive_loss": 3.1547234058380127,
|
| 473 |
+
"train/negative_loss": 3.1508476734161377,
|
| 474 |
+
"train/num_negatives": 44,
|
| 475 |
+
"train/num_positives": 12,
|
| 476 |
+
"train/positive_loss": 0.0038756858557462692,
|
| 477 |
+
"train/total_loss": 0.9103535413742065
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 20.0,
|
| 481 |
+
"step": 400,
|
| 482 |
+
"train/classification_loss": 0.2563062608242035,
|
| 483 |
+
"train/contrastive_loss": 3.433103084564209,
|
| 484 |
+
"train/negative_loss": 3.146430253982544,
|
| 485 |
+
"train/num_negatives": 52,
|
| 486 |
+
"train/num_positives": 4,
|
| 487 |
+
"train/positive_loss": 0.2866727411746979,
|
| 488 |
+
"train/total_loss": 0.9429268836975098
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 20.0,
|
| 492 |
+
"step": 400,
|
| 493 |
+
"train/classification_loss": 0.24611052870750427,
|
| 494 |
+
"train/contrastive_loss": 3.2940289974212646,
|
| 495 |
+
"train/negative_loss": 2.822859525680542,
|
| 496 |
+
"train/num_negatives": 48,
|
| 497 |
+
"train/num_positives": 8,
|
| 498 |
+
"train/positive_loss": 0.47116944193840027,
|
| 499 |
+
"train/total_loss": 0.9049162864685059
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 20.0,
|
| 503 |
+
"step": 400,
|
| 504 |
+
"train/classification_loss": 0.26658472418785095,
|
| 505 |
+
"train/contrastive_loss": 0.9518164992332458,
|
| 506 |
+
"train/negative_loss": 0.90028977394104,
|
| 507 |
+
"train/num_negatives": 46,
|
| 508 |
+
"train/num_positives": 8,
|
| 509 |
+
"train/positive_loss": 0.051526736468076706,
|
| 510 |
+
"train/total_loss": 0.45694804191589355
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 20.0,
|
| 514 |
"eval_exact_match_accuracy": 0.1,
|
| 515 |
"eval_hamming_loss": 0.075,
|
| 516 |
+
"eval_loss": 0.7537041902542114,
|
| 517 |
"eval_macro_f1": 0.0784313725490196,
|
| 518 |
"eval_macro_precision": 0.1092436974789916,
|
| 519 |
"eval_macro_recall": 0.06149732620320855,
|
| 520 |
"eval_micro_f1": 0.2153846153846154,
|
| 521 |
"eval_micro_precision": 0.875,
|
| 522 |
"eval_micro_recall": 0.12280701754385964,
|
| 523 |
+
"eval_runtime": 0.2075,
|
| 524 |
+
"eval_samples_per_second": 192.749,
|
| 525 |
+
"eval_steps_per_second": 24.094,
|
| 526 |
+
"step": 400
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 20.0,
|
| 530 |
+
"step": 400,
|
| 531 |
+
"train/classification_loss": 0.24341967701911926,
|
| 532 |
+
"train/contrastive_loss": 0.7751690149307251,
|
| 533 |
+
"train/negative_loss": 0.517412543296814,
|
| 534 |
+
"train/num_negatives": 42,
|
| 535 |
+
"train/num_positives": 10,
|
| 536 |
+
"train/positive_loss": 0.25775647163391113,
|
| 537 |
+
"train/total_loss": 0.3984534740447998
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 20.0,
|
| 541 |
+
"step": 400,
|
| 542 |
+
"train/classification_loss": 0.2490834891796112,
|
| 543 |
+
"train/contrastive_loss": 0.44699349999427795,
|
| 544 |
+
"train/negative_loss": 0.4466739594936371,
|
| 545 |
+
"train/num_negatives": 52,
|
| 546 |
+
"train/num_positives": 4,
|
| 547 |
+
"train/positive_loss": 0.0003195433528162539,
|
| 548 |
+
"train/total_loss": 0.33848220109939575
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 22.5,
|
| 552 |
+
"grad_norm": 3.832901954650879,
|
| 553 |
+
"learning_rate": 1.2266666666666667e-05,
|
| 554 |
+
"loss": 0.6929,
|
| 555 |
+
"step": 450
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 22.5,
|
| 559 |
+
"step": 450,
|
| 560 |
+
"train/classification_loss": 0.23125219345092773,
|
| 561 |
+
"train/contrastive_loss": 0.7171761393547058,
|
| 562 |
+
"train/negative_loss": 0.7147800922393799,
|
| 563 |
+
"train/num_negatives": 50,
|
| 564 |
+
"train/num_positives": 4,
|
| 565 |
+
"train/positive_loss": 0.0023960734251886606,
|
| 566 |
+
"train/total_loss": 0.37468743324279785
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 22.5,
|
| 570 |
+
"step": 450,
|
| 571 |
+
"train/classification_loss": 0.20611771941184998,
|
| 572 |
+
"train/contrastive_loss": 0.7421404123306274,
|
| 573 |
+
"train/negative_loss": 0.4822021722793579,
|
| 574 |
+
"train/num_negatives": 50,
|
| 575 |
+
"train/num_positives": 6,
|
| 576 |
+
"train/positive_loss": 0.2599382698535919,
|
| 577 |
+
"train/total_loss": 0.35454580187797546
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 25.0,
|
| 581 |
+
"grad_norm": 11.741353988647461,
|
| 582 |
+
"learning_rate": 1.1155555555555556e-05,
|
| 583 |
+
"loss": 0.7074,
|
| 584 |
+
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 25.0,
|
| 588 |
+
"step": 500,
|
| 589 |
+
"train/classification_loss": 0.1859707236289978,
|
| 590 |
+
"train/contrastive_loss": 2.448401689529419,
|
| 591 |
+
"train/negative_loss": 1.6338316202163696,
|
| 592 |
+
"train/num_negatives": 44,
|
| 593 |
+
"train/num_positives": 12,
|
| 594 |
+
"train/positive_loss": 0.8145700693130493,
|
| 595 |
+
"train/total_loss": 0.6756510734558105
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 25.0,
|
| 599 |
+
"step": 500,
|
| 600 |
+
"train/classification_loss": 0.2645534574985504,
|
| 601 |
+
"train/contrastive_loss": 2.718876838684082,
|
| 602 |
+
"train/negative_loss": 2.718696355819702,
|
| 603 |
+
"train/num_negatives": 44,
|
| 604 |
+
"train/num_positives": 12,
|
| 605 |
+
"train/positive_loss": 0.00018060117145068944,
|
| 606 |
+
"train/total_loss": 0.8083288669586182
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 25.0,
|
| 610 |
+
"step": 500,
|
| 611 |
+
"train/classification_loss": 0.244839608669281,
|
| 612 |
+
"train/contrastive_loss": 4.527173042297363,
|
| 613 |
+
"train/negative_loss": 2.9302120208740234,
|
| 614 |
+
"train/num_negatives": 52,
|
| 615 |
+
"train/num_positives": 4,
|
| 616 |
+
"train/positive_loss": 1.5969611406326294,
|
| 617 |
+
"train/total_loss": 1.1502742767333984
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 25.0,
|
| 621 |
+
"step": 500,
|
| 622 |
+
"train/classification_loss": 0.22313973307609558,
|
| 623 |
+
"train/contrastive_loss": 3.153029203414917,
|
| 624 |
+
"train/negative_loss": 2.193068504333496,
|
| 625 |
+
"train/num_negatives": 48,
|
| 626 |
+
"train/num_positives": 8,
|
| 627 |
+
"train/positive_loss": 0.9599607586860657,
|
| 628 |
+
"train/total_loss": 0.8537455797195435
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 25.0,
|
| 632 |
+
"step": 500,
|
| 633 |
+
"train/classification_loss": 0.25405386090278625,
|
| 634 |
+
"train/contrastive_loss": 1.7682042121887207,
|
| 635 |
+
"train/negative_loss": 0.7665292620658875,
|
| 636 |
+
"train/num_negatives": 46,
|
| 637 |
+
"train/num_positives": 8,
|
| 638 |
+
"train/positive_loss": 1.001675009727478,
|
| 639 |
+
"train/total_loss": 0.6076947450637817
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 25.0,
|
| 643 |
+
"eval_exact_match_accuracy": 0.175,
|
| 644 |
+
"eval_hamming_loss": 0.07205882352941176,
|
| 645 |
+
"eval_loss": 0.8191388845443726,
|
| 646 |
+
"eval_macro_f1": 0.10560224089635854,
|
| 647 |
+
"eval_macro_precision": 0.14869281045751634,
|
| 648 |
+
"eval_macro_recall": 0.08449197860962566,
|
| 649 |
+
"eval_micro_f1": 0.30985915492957744,
|
| 650 |
+
"eval_micro_precision": 0.7857142857142857,
|
| 651 |
+
"eval_micro_recall": 0.19298245614035087,
|
| 652 |
+
"eval_runtime": 0.2062,
|
| 653 |
+
"eval_samples_per_second": 193.983,
|
| 654 |
+
"eval_steps_per_second": 24.248,
|
| 655 |
+
"step": 500
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 25.0,
|
| 659 |
+
"step": 500,
|
| 660 |
+
"train/classification_loss": 0.16842614114284515,
|
| 661 |
+
"train/contrastive_loss": 0.2924913763999939,
|
| 662 |
+
"train/negative_loss": 0.28081196546554565,
|
| 663 |
+
"train/num_negatives": 42,
|
| 664 |
+
"train/num_positives": 14,
|
| 665 |
+
"train/positive_loss": 0.011679417453706264,
|
| 666 |
+
"train/total_loss": 0.22692441940307617
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 25.0,
|
| 670 |
+
"step": 500,
|
| 671 |
+
"train/classification_loss": 0.24865391850471497,
|
| 672 |
+
"train/contrastive_loss": 1.3489311933517456,
|
| 673 |
+
"train/negative_loss": 0.7319411635398865,
|
| 674 |
+
"train/num_negatives": 44,
|
| 675 |
+
"train/num_positives": 12,
|
| 676 |
+
"train/positive_loss": 0.6169900298118591,
|
| 677 |
+
"train/total_loss": 0.5184401273727417
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 27.5,
|
| 681 |
+
"grad_norm": 6.354713439941406,
|
| 682 |
+
"learning_rate": 1.0044444444444446e-05,
|
| 683 |
+
"loss": 0.6366,
|
| 684 |
+
"step": 550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 27.5,
|
| 688 |
+
"step": 550,
|
| 689 |
+
"train/classification_loss": 0.23189660906791687,
|
| 690 |
+
"train/contrastive_loss": 0.8256056308746338,
|
| 691 |
+
"train/negative_loss": 0.8255752325057983,
|
| 692 |
+
"train/num_negatives": 54,
|
| 693 |
+
"train/num_positives": 2,
|
| 694 |
+
"train/positive_loss": 3.039883085875772e-05,
|
| 695 |
+
"train/total_loss": 0.3970177173614502
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 27.5,
|
| 699 |
+
"step": 550,
|
| 700 |
+
"train/classification_loss": 0.2541985809803009,
|
| 701 |
+
"train/contrastive_loss": 1.3727329969406128,
|
| 702 |
+
"train/negative_loss": 1.0316259860992432,
|
| 703 |
+
"train/num_negatives": 40,
|
| 704 |
+
"train/num_positives": 12,
|
| 705 |
+
"train/positive_loss": 0.34110698103904724,
|
| 706 |
+
"train/total_loss": 0.528745174407959
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 30.0,
|
| 710 |
+
"grad_norm": 12.662666320800781,
|
| 711 |
+
"learning_rate": 8.933333333333333e-06,
|
| 712 |
+
"loss": 0.6281,
|
| 713 |
+
"step": 600
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 30.0,
|
| 717 |
+
"step": 600,
|
| 718 |
+
"train/classification_loss": 0.17969225347042084,
|
| 719 |
+
"train/contrastive_loss": 2.1998844146728516,
|
| 720 |
+
"train/negative_loss": 1.262956976890564,
|
| 721 |
+
"train/num_negatives": 44,
|
| 722 |
+
"train/num_positives": 12,
|
| 723 |
+
"train/positive_loss": 0.9369274973869324,
|
| 724 |
+
"train/total_loss": 0.6196691393852234
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 30.0,
|
| 728 |
+
"step": 600,
|
| 729 |
+
"train/classification_loss": 0.2561105787754059,
|
| 730 |
+
"train/contrastive_loss": 3.2061843872070312,
|
| 731 |
+
"train/negative_loss": 3.2013988494873047,
|
| 732 |
+
"train/num_negatives": 44,
|
| 733 |
+
"train/num_positives": 12,
|
| 734 |
+
"train/positive_loss": 0.00478551909327507,
|
| 735 |
+
"train/total_loss": 0.8973474502563477
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 30.0,
|
| 739 |
+
"step": 600,
|
| 740 |
+
"train/classification_loss": 0.24141749739646912,
|
| 741 |
+
"train/contrastive_loss": 4.734986305236816,
|
| 742 |
+
"train/negative_loss": 3.0736501216888428,
|
| 743 |
+
"train/num_negatives": 52,
|
| 744 |
+
"train/num_positives": 4,
|
| 745 |
+
"train/positive_loss": 1.6613364219665527,
|
| 746 |
+
"train/total_loss": 1.1884148120880127
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 30.0,
|
| 750 |
+
"step": 600,
|
| 751 |
+
"train/classification_loss": 0.22462235391139984,
|
| 752 |
+
"train/contrastive_loss": 4.2080254554748535,
|
| 753 |
+
"train/negative_loss": 3.6183528900146484,
|
| 754 |
+
"train/num_negatives": 48,
|
| 755 |
+
"train/num_positives": 8,
|
| 756 |
+
"train/positive_loss": 0.5896727442741394,
|
| 757 |
+
"train/total_loss": 1.0662274360656738
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 30.0,
|
| 761 |
+
"step": 600,
|
| 762 |
+
"train/classification_loss": 0.24234618246555328,
|
| 763 |
+
"train/contrastive_loss": 1.1968506574630737,
|
| 764 |
+
"train/negative_loss": 0.8532204031944275,
|
| 765 |
+
"train/num_negatives": 46,
|
| 766 |
+
"train/num_positives": 8,
|
| 767 |
+
"train/positive_loss": 0.34363028407096863,
|
| 768 |
+
"train/total_loss": 0.4817163348197937
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 30.0,
|
| 772 |
+
"eval_exact_match_accuracy": 0.275,
|
| 773 |
+
"eval_hamming_loss": 0.06470588235294118,
|
| 774 |
+
"eval_loss": 0.8506749868392944,
|
| 775 |
+
"eval_macro_f1": 0.12978524743230624,
|
| 776 |
+
"eval_macro_precision": 0.15735294117647058,
|
| 777 |
+
"eval_macro_recall": 0.11336898395721925,
|
| 778 |
+
"eval_micro_f1": 0.42105263157894735,
|
| 779 |
+
"eval_micro_precision": 0.8421052631578947,
|
| 780 |
+
"eval_micro_recall": 0.2807017543859649,
|
| 781 |
+
"eval_runtime": 0.2062,
|
| 782 |
+
"eval_samples_per_second": 193.977,
|
| 783 |
+
"eval_steps_per_second": 24.247,
|
| 784 |
+
"step": 600
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 30.0,
|
| 788 |
+
"step": 600,
|
| 789 |
+
"train/classification_loss": 0.16461151838302612,
|
| 790 |
+
"train/contrastive_loss": 0.5128712058067322,
|
| 791 |
+
"train/negative_loss": 0.23724108934402466,
|
| 792 |
+
"train/num_negatives": 38,
|
| 793 |
+
"train/num_positives": 18,
|
| 794 |
+
"train/positive_loss": 0.2756301164627075,
|
| 795 |
+
"train/total_loss": 0.2671857476234436
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 30.0,
|
| 799 |
+
"step": 600,
|
| 800 |
+
"train/classification_loss": 0.2038976103067398,
|
| 801 |
+
"train/contrastive_loss": 1.0636850595474243,
|
| 802 |
+
"train/negative_loss": 0.5897871255874634,
|
| 803 |
+
"train/num_negatives": 48,
|
| 804 |
+
"train/num_positives": 8,
|
| 805 |
+
"train/positive_loss": 0.47389790415763855,
|
| 806 |
+
"train/total_loss": 0.41663461923599243
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 32.5,
|
| 810 |
+
"grad_norm": 11.408817291259766,
|
| 811 |
+
"learning_rate": 7.822222222222224e-06,
|
| 812 |
+
"loss": 0.5854,
|
| 813 |
+
"step": 650
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 32.5,
|
| 817 |
+
"step": 650,
|
| 818 |
+
"train/classification_loss": 0.1786508709192276,
|
| 819 |
+
"train/contrastive_loss": 0.5145746469497681,
|
| 820 |
+
"train/negative_loss": 0.14455223083496094,
|
| 821 |
+
"train/num_negatives": 36,
|
| 822 |
+
"train/num_positives": 18,
|
| 823 |
+
"train/positive_loss": 0.3700224459171295,
|
| 824 |
+
"train/total_loss": 0.28156578540802
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 32.5,
|
| 828 |
+
"step": 650,
|
| 829 |
+
"train/classification_loss": 0.2088967263698578,
|
| 830 |
+
"train/contrastive_loss": 1.686006784439087,
|
| 831 |
+
"train/negative_loss": 1.5359704494476318,
|
| 832 |
+
"train/num_negatives": 50,
|
| 833 |
+
"train/num_positives": 6,
|
| 834 |
+
"train/positive_loss": 0.1500363051891327,
|
| 835 |
+
"train/total_loss": 0.5460981130599976
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 35.0,
|
| 839 |
+
"grad_norm": 5.847558975219727,
|
| 840 |
+
"learning_rate": 6.711111111111111e-06,
|
| 841 |
+
"loss": 0.5506,
|
| 842 |
+
"step": 700
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 35.0,
|
| 846 |
+
"step": 700,
|
| 847 |
+
"train/classification_loss": 0.17107558250427246,
|
| 848 |
+
"train/contrastive_loss": 2.287914276123047,
|
| 849 |
+
"train/negative_loss": 1.0503000020980835,
|
| 850 |
+
"train/num_negatives": 44,
|
| 851 |
+
"train/num_positives": 12,
|
| 852 |
+
"train/positive_loss": 1.237614393234253,
|
| 853 |
+
"train/total_loss": 0.6286584138870239
|
| 854 |
},
|
| 855 |
{
|
| 856 |
"epoch": 35.0,
|
| 857 |
+
"step": 700,
|
| 858 |
+
"train/classification_loss": 0.2500559389591217,
|
| 859 |
+
"train/contrastive_loss": 2.9361915588378906,
|
| 860 |
+
"train/negative_loss": 2.824162721633911,
|
| 861 |
+
"train/num_negatives": 44,
|
| 862 |
+
"train/num_positives": 12,
|
| 863 |
+
"train/positive_loss": 0.1120288297533989,
|
| 864 |
+
"train/total_loss": 0.8372942209243774
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 35.0,
|
| 868 |
+
"step": 700,
|
| 869 |
+
"train/classification_loss": 0.23473431169986725,
|
| 870 |
+
"train/contrastive_loss": 4.627296447753906,
|
| 871 |
+
"train/negative_loss": 3.135880708694458,
|
| 872 |
+
"train/num_negatives": 52,
|
| 873 |
+
"train/num_positives": 4,
|
| 874 |
+
"train/positive_loss": 1.4914155006408691,
|
| 875 |
+
"train/total_loss": 1.160193681716919
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 35.0,
|
| 879 |
+
"step": 700,
|
| 880 |
+
"train/classification_loss": 0.20871424674987793,
|
| 881 |
+
"train/contrastive_loss": 2.308566093444824,
|
| 882 |
+
"train/negative_loss": 1.4599714279174805,
|
| 883 |
+
"train/num_negatives": 48,
|
| 884 |
+
"train/num_positives": 8,
|
| 885 |
+
"train/positive_loss": 0.8485947251319885,
|
| 886 |
+
"train/total_loss": 0.6704274415969849
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 35.0,
|
| 890 |
+
"step": 700,
|
| 891 |
+
"train/classification_loss": 0.23494853079319,
|
| 892 |
+
"train/contrastive_loss": 0.9399895071983337,
|
| 893 |
+
"train/negative_loss": 0.4285624623298645,
|
| 894 |
+
"train/num_negatives": 46,
|
| 895 |
+
"train/num_positives": 8,
|
| 896 |
+
"train/positive_loss": 0.5114270448684692,
|
| 897 |
+
"train/total_loss": 0.4229464530944824
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 35.0,
|
| 901 |
+
"eval_exact_match_accuracy": 0.25,
|
| 902 |
+
"eval_hamming_loss": 0.0661764705882353,
|
| 903 |
+
"eval_loss": 0.7439039945602417,
|
| 904 |
+
"eval_macro_f1": 0.12555610479485912,
|
| 905 |
+
"eval_macro_precision": 0.15630252100840336,
|
| 906 |
+
"eval_macro_recall": 0.10748663101604278,
|
| 907 |
+
"eval_micro_f1": 0.4,
|
| 908 |
+
"eval_micro_precision": 0.8333333333333334,
|
| 909 |
+
"eval_micro_recall": 0.2631578947368421,
|
| 910 |
+
"eval_runtime": 0.2043,
|
| 911 |
+
"eval_samples_per_second": 195.76,
|
| 912 |
+
"eval_steps_per_second": 24.47,
|
| 913 |
+
"step": 700
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 35.0,
|
| 917 |
+
"step": 700,
|
| 918 |
+
"train/classification_loss": 0.18580235540866852,
|
| 919 |
+
"train/contrastive_loss": 0.5447431802749634,
|
| 920 |
+
"train/negative_loss": 0.26964136958122253,
|
| 921 |
+
"train/num_negatives": 40,
|
| 922 |
+
"train/num_positives": 16,
|
| 923 |
+
"train/positive_loss": 0.27510178089141846,
|
| 924 |
+
"train/total_loss": 0.29475098848342896
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 35.0,
|
| 928 |
+
"step": 700,
|
| 929 |
+
"train/classification_loss": 0.19202794134616852,
|
| 930 |
+
"train/contrastive_loss": 1.1511749029159546,
|
| 931 |
+
"train/negative_loss": 0.8905836343765259,
|
| 932 |
+
"train/num_negatives": 40,
|
| 933 |
+
"train/num_positives": 16,
|
| 934 |
+
"train/positive_loss": 0.2605912685394287,
|
| 935 |
+
"train/total_loss": 0.42226290702819824
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 37.5,
|
| 939 |
+
"grad_norm": 5.041801452636719,
|
| 940 |
+
"learning_rate": 5.600000000000001e-06,
|
| 941 |
+
"loss": 0.5486,
|
| 942 |
+
"step": 750
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 37.5,
|
| 946 |
+
"step": 750,
|
| 947 |
+
"train/classification_loss": 0.18688128888607025,
|
| 948 |
+
"train/contrastive_loss": 0.7704952955245972,
|
| 949 |
+
"train/negative_loss": 0.7704557180404663,
|
| 950 |
+
"train/num_negatives": 48,
|
| 951 |
+
"train/num_positives": 8,
|
| 952 |
+
"train/positive_loss": 3.9578346331836656e-05,
|
| 953 |
+
"train/total_loss": 0.3409803509712219
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 37.5,
|
| 957 |
+
"step": 750,
|
| 958 |
+
"train/classification_loss": 0.19349302351474762,
|
| 959 |
+
"train/contrastive_loss": 0.4310402274131775,
|
| 960 |
+
"train/negative_loss": 0.4191313683986664,
|
| 961 |
+
"train/num_negatives": 46,
|
| 962 |
+
"train/num_positives": 10,
|
| 963 |
+
"train/positive_loss": 0.011908866465091705,
|
| 964 |
+
"train/total_loss": 0.2797010540962219
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 40.0,
|
| 968 |
+
"grad_norm": 8.75462532043457,
|
| 969 |
+
"learning_rate": 4.488888888888889e-06,
|
| 970 |
+
"loss": 0.5091,
|
| 971 |
+
"step": 800
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 40.0,
|
| 975 |
+
"step": 800,
|
| 976 |
+
"train/classification_loss": 0.1726197898387909,
|
| 977 |
+
"train/contrastive_loss": 1.8750531673431396,
|
| 978 |
+
"train/negative_loss": 0.9292571544647217,
|
| 979 |
+
"train/num_negatives": 44,
|
| 980 |
+
"train/num_positives": 12,
|
| 981 |
+
"train/positive_loss": 0.9457959532737732,
|
| 982 |
+
"train/total_loss": 0.5476304292678833
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 40.0,
|
| 986 |
+
"step": 800,
|
| 987 |
+
"train/classification_loss": 0.24574041366577148,
|
| 988 |
+
"train/contrastive_loss": 3.074068784713745,
|
| 989 |
+
"train/negative_loss": 3.055785894393921,
|
| 990 |
+
"train/num_negatives": 44,
|
| 991 |
+
"train/num_positives": 12,
|
| 992 |
+
"train/positive_loss": 0.018282821401953697,
|
| 993 |
+
"train/total_loss": 0.8605541586875916
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 40.0,
|
| 997 |
+
"step": 800,
|
| 998 |
+
"train/classification_loss": 0.2323407232761383,
|
| 999 |
+
"train/contrastive_loss": 4.7350592613220215,
|
| 1000 |
+
"train/negative_loss": 2.995204210281372,
|
| 1001 |
+
"train/num_negatives": 52,
|
| 1002 |
+
"train/num_positives": 4,
|
| 1003 |
+
"train/positive_loss": 1.739855170249939,
|
| 1004 |
+
"train/total_loss": 1.1793526411056519
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 40.0,
|
| 1008 |
+
"step": 800,
|
| 1009 |
+
"train/classification_loss": 0.20833879709243774,
|
| 1010 |
+
"train/contrastive_loss": 3.4109816551208496,
|
| 1011 |
+
"train/negative_loss": 2.4856531620025635,
|
| 1012 |
+
"train/num_negatives": 48,
|
| 1013 |
+
"train/num_positives": 8,
|
| 1014 |
+
"train/positive_loss": 0.9253284931182861,
|
| 1015 |
+
"train/total_loss": 0.8905351161956787
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 40.0,
|
| 1019 |
+
"step": 800,
|
| 1020 |
+
"train/classification_loss": 0.23090128600597382,
|
| 1021 |
+
"train/contrastive_loss": 1.3861477375030518,
|
| 1022 |
+
"train/negative_loss": 0.5846720933914185,
|
| 1023 |
+
"train/num_negatives": 46,
|
| 1024 |
+
"train/num_positives": 8,
|
| 1025 |
+
"train/positive_loss": 0.8014755845069885,
|
| 1026 |
+
"train/total_loss": 0.5081308484077454
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 40.0,
|
| 1030 |
+
"eval_exact_match_accuracy": 0.275,
|
| 1031 |
+
"eval_hamming_loss": 0.06470588235294118,
|
| 1032 |
+
"eval_loss": 0.7972406148910522,
|
| 1033 |
+
"eval_macro_f1": 0.12978524743230624,
|
| 1034 |
+
"eval_macro_precision": 0.15735294117647058,
|
| 1035 |
+
"eval_macro_recall": 0.11336898395721925,
|
| 1036 |
+
"eval_micro_f1": 0.42105263157894735,
|
| 1037 |
+
"eval_micro_precision": 0.8421052631578947,
|
| 1038 |
+
"eval_micro_recall": 0.2807017543859649,
|
| 1039 |
+
"eval_runtime": 0.2048,
|
| 1040 |
+
"eval_samples_per_second": 195.328,
|
| 1041 |
+
"eval_steps_per_second": 24.416,
|
| 1042 |
+
"step": 800
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 40.0,
|
| 1046 |
+
"step": 800,
|
| 1047 |
+
"train/classification_loss": 0.16540196537971497,
|
| 1048 |
+
"train/contrastive_loss": 0.28263404965400696,
|
| 1049 |
+
"train/negative_loss": 0.28260505199432373,
|
| 1050 |
+
"train/num_negatives": 44,
|
| 1051 |
+
"train/num_positives": 12,
|
| 1052 |
+
"train/positive_loss": 2.9008200726821087e-05,
|
| 1053 |
+
"train/total_loss": 0.22192877531051636
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 40.0,
|
| 1057 |
+
"step": 800,
|
| 1058 |
+
"train/classification_loss": 0.12730641663074493,
|
| 1059 |
+
"train/contrastive_loss": 0.2957398295402527,
|
| 1060 |
+
"train/negative_loss": 0.29565563797950745,
|
| 1061 |
+
"train/num_negatives": 30,
|
| 1062 |
+
"train/num_positives": 26,
|
| 1063 |
+
"train/positive_loss": 8.41914297780022e-05,
|
| 1064 |
+
"train/total_loss": 0.1864543855190277
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 42.5,
|
| 1068 |
+
"grad_norm": 6.981760501861572,
|
| 1069 |
+
"learning_rate": 3.377777777777778e-06,
|
| 1070 |
+
"loss": 0.4948,
|
| 1071 |
+
"step": 850
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 42.5,
|
| 1075 |
+
"step": 850,
|
| 1076 |
+
"train/classification_loss": 0.14926917850971222,
|
| 1077 |
+
"train/contrastive_loss": 0.2675209641456604,
|
| 1078 |
+
"train/negative_loss": 0.22713389992713928,
|
| 1079 |
+
"train/num_negatives": 40,
|
| 1080 |
+
"train/num_positives": 16,
|
| 1081 |
+
"train/positive_loss": 0.04038705304265022,
|
| 1082 |
+
"train/total_loss": 0.20277337729930878
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 42.5,
|
| 1086 |
+
"step": 850,
|
| 1087 |
+
"train/classification_loss": 0.17404112219810486,
|
| 1088 |
+
"train/contrastive_loss": 0.08631884306669235,
|
| 1089 |
+
"train/negative_loss": 0.08508215099573135,
|
| 1090 |
+
"train/num_negatives": 40,
|
| 1091 |
+
"train/num_positives": 16,
|
| 1092 |
+
"train/positive_loss": 0.0012366925366222858,
|
| 1093 |
+
"train/total_loss": 0.19130489230155945
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 45.0,
|
| 1097 |
+
"grad_norm": 3.230358839035034,
|
| 1098 |
+
"learning_rate": 2.266666666666667e-06,
|
| 1099 |
+
"loss": 0.5038,
|
| 1100 |
+
"step": 900
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 45.0,
|
| 1104 |
+
"step": 900,
|
| 1105 |
+
"train/classification_loss": 0.17026303708553314,
|
| 1106 |
+
"train/contrastive_loss": 2.131587266921997,
|
| 1107 |
+
"train/negative_loss": 0.9242226481437683,
|
| 1108 |
+
"train/num_negatives": 44,
|
| 1109 |
+
"train/num_positives": 12,
|
| 1110 |
+
"train/positive_loss": 1.207364559173584,
|
| 1111 |
+
"train/total_loss": 0.5965805053710938
|
| 1112 |
},
|
| 1113 |
{
|
| 1114 |
"epoch": 45.0,
|
| 1115 |
+
"step": 900,
|
| 1116 |
+
"train/classification_loss": 0.24070139229297638,
|
| 1117 |
+
"train/contrastive_loss": 3.244175910949707,
|
| 1118 |
+
"train/negative_loss": 3.2388288974761963,
|
| 1119 |
+
"train/num_negatives": 44,
|
| 1120 |
+
"train/num_positives": 12,
|
| 1121 |
+
"train/positive_loss": 0.005346930585801601,
|
| 1122 |
+
"train/total_loss": 0.8895365595817566
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
+
"epoch": 45.0,
|
| 1126 |
+
"step": 900,
|
| 1127 |
+
"train/classification_loss": 0.2271755486726761,
|
| 1128 |
+
"train/contrastive_loss": 4.544618129730225,
|
| 1129 |
+
"train/negative_loss": 2.900235891342163,
|
| 1130 |
+
"train/num_negatives": 52,
|
| 1131 |
+
"train/num_positives": 4,
|
| 1132 |
+
"train/positive_loss": 1.6443822383880615,
|
| 1133 |
+
"train/total_loss": 1.1360992193222046
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
+
"epoch": 45.0,
|
| 1137 |
+
"step": 900,
|
| 1138 |
+
"train/classification_loss": 0.20477482676506042,
|
| 1139 |
+
"train/contrastive_loss": 3.36401104927063,
|
| 1140 |
+
"train/negative_loss": 2.652341365814209,
|
| 1141 |
+
"train/num_negatives": 48,
|
| 1142 |
+
"train/num_positives": 8,
|
| 1143 |
+
"train/positive_loss": 0.7116697430610657,
|
| 1144 |
+
"train/total_loss": 0.8775770664215088
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
+
"epoch": 45.0,
|
| 1148 |
+
"step": 900,
|
| 1149 |
+
"train/classification_loss": 0.22650422155857086,
|
| 1150 |
+
"train/contrastive_loss": 1.7576167583465576,
|
| 1151 |
+
"train/negative_loss": 0.546459436416626,
|
| 1152 |
"train/num_negatives": 46,
|
| 1153 |
"train/num_positives": 8,
|
| 1154 |
+
"train/positive_loss": 1.2111573219299316,
|
| 1155 |
+
"train/total_loss": 0.578027606010437
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
+
"epoch": 45.0,
|
| 1159 |
+
"eval_exact_match_accuracy": 0.275,
|
| 1160 |
+
"eval_hamming_loss": 0.06470588235294118,
|
| 1161 |
+
"eval_loss": 0.8155641555786133,
|
| 1162 |
+
"eval_macro_f1": 0.12978524743230624,
|
| 1163 |
+
"eval_macro_precision": 0.15735294117647058,
|
| 1164 |
+
"eval_macro_recall": 0.11336898395721925,
|
| 1165 |
+
"eval_micro_f1": 0.42105263157894735,
|
| 1166 |
+
"eval_micro_precision": 0.8421052631578947,
|
| 1167 |
+
"eval_micro_recall": 0.2807017543859649,
|
| 1168 |
+
"eval_runtime": 0.2057,
|
| 1169 |
+
"eval_samples_per_second": 194.486,
|
| 1170 |
+
"eval_steps_per_second": 24.311,
|
| 1171 |
+
"step": 900
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 45.0,
|
| 1175 |
+
"step": 900,
|
| 1176 |
"total_flos": 0.0,
|
| 1177 |
+
"train_loss": 0.9705644819471572,
|
| 1178 |
+
"train_runtime": 232.6541,
|
| 1179 |
+
"train_samples_per_second": 67.912,
|
| 1180 |
+
"train_steps_per_second": 4.298
|
| 1181 |
}
|
| 1182 |
],
|
| 1183 |
"logging_steps": 50,
|
| 1184 |
+
"max_steps": 1000,
|
| 1185 |
"num_input_tokens_seen": 0,
|
| 1186 |
"num_train_epochs": 50,
|
| 1187 |
+
"save_steps": 50000,
|
| 1188 |
"stateful_callbacks": {
|
| 1189 |
"EarlyStoppingCallback": {
|
| 1190 |
"args": {
|
|
|
|
| 1200 |
"should_epoch_stop": false,
|
| 1201 |
"should_evaluate": false,
|
| 1202 |
"should_log": false,
|
| 1203 |
+
"should_save": false,
|
| 1204 |
+
"should_training_stop": false
|
| 1205 |
},
|
| 1206 |
"attributes": {}
|
| 1207 |
}
|
| 1208 |
},
|
| 1209 |
"total_flos": 0.0,
|
| 1210 |
+
"train_batch_size": 8,
|
| 1211 |
"trial_name": null,
|
| 1212 |
"trial_params": null
|
| 1213 |
}
|