Commit
•
80fecf3
1
Parent(s):
1ae1714
Training in progress epoch 0
Browse files- README.md +5 -32
- config.json +2 -2
- tf_model.h5 +1 -1
README.md
CHANGED
|
@@ -15,10 +15,10 @@ probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Train Loss:
|
| 19 |
-
- Validation Loss:
|
| 20 |
-
- Train Accuracy: 0.
|
| 21 |
-
- Epoch:
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -44,34 +44,7 @@ The following hyperparameters were used during training:
|
|
| 44 |
|
| 45 |
| Train Loss | Validation Loss | Train Accuracy | Epoch |
|
| 46 |
|:----------:|:---------------:|:--------------:|:-----:|
|
| 47 |
-
| 4.
|
| 48 |
-
| 4.5685 | 4.3200 | 0.2233 | 1 |
|
| 49 |
-
| 3.9808 | 3.6775 | 0.5833 | 2 |
|
| 50 |
-
| 3.3170 | 3.0567 | 0.72 | 3 |
|
| 51 |
-
| 2.7260 | 2.5223 | 0.7967 | 4 |
|
| 52 |
-
| 2.1996 | 2.0628 | 0.8167 | 5 |
|
| 53 |
-
| 1.7561 | 1.7042 | 0.84 | 6 |
|
| 54 |
-
| 1.4166 | 1.4208 | 0.86 | 7 |
|
| 55 |
-
| 1.1157 | 1.1895 | 0.8733 | 8 |
|
| 56 |
-
| 0.8749 | 0.9915 | 0.8933 | 9 |
|
| 57 |
-
| 0.6874 | 0.8773 | 0.8833 | 10 |
|
| 58 |
-
| 0.5538 | 0.7759 | 0.88 | 11 |
|
| 59 |
-
| 0.4475 | 0.7086 | 0.8833 | 12 |
|
| 60 |
-
| 0.3682 | 0.6501 | 0.8867 | 13 |
|
| 61 |
-
| 0.3097 | 0.6114 | 0.8933 | 14 |
|
| 62 |
-
| 0.2609 | 0.5694 | 0.9 | 15 |
|
| 63 |
-
| 0.2240 | 0.5430 | 0.9 | 16 |
|
| 64 |
-
| 0.1915 | 0.5198 | 0.8967 | 17 |
|
| 65 |
-
| 0.1681 | 0.5091 | 0.9067 | 18 |
|
| 66 |
-
| 0.1482 | 0.4936 | 0.8967 | 19 |
|
| 67 |
-
| 0.1339 | 0.4791 | 0.9067 | 20 |
|
| 68 |
-
| 0.1202 | 0.4821 | 0.8933 | 21 |
|
| 69 |
-
| 0.1090 | 0.4700 | 0.89 | 22 |
|
| 70 |
-
| 0.0987 | 0.4585 | 0.8933 | 23 |
|
| 71 |
-
| 0.0900 | 0.4553 | 0.9 | 24 |
|
| 72 |
-
| 0.0843 | 0.4459 | 0.9033 | 25 |
|
| 73 |
-
| 0.0768 | 0.4447 | 0.9033 | 26 |
|
| 74 |
-
| 0.0717 | 0.4388 | 0.9067 | 27 |
|
| 75 |
|
| 76 |
|
| 77 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [bert-large-uncased](https://huggingface.co/bert-large-uncased) on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Train Loss: 4.9762
|
| 19 |
+
- Validation Loss: 4.7210
|
| 20 |
+
- Train Accuracy: 0.0767
|
| 21 |
+
- Epoch: 0
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 44 |
|
| 45 |
| Train Loss | Validation Loss | Train Accuracy | Epoch |
|
| 46 |
|:----------:|:---------------:|:--------------:|:-----:|
|
| 47 |
+
| 4.9762 | 4.7210 | 0.0767 | 0 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
|
| 49 |
|
| 50 |
### Framework versions
|
config.json
CHANGED
|
@@ -4,10 +4,10 @@
|
|
| 4 |
"BertForSequenceClassification"
|
| 5 |
],
|
| 6 |
"attention_probs_dropout_prob": 0.1,
|
| 7 |
-
"classifier_dropout":
|
| 8 |
"gradient_checkpointing": false,
|
| 9 |
"hidden_act": "gelu",
|
| 10 |
-
"hidden_dropout_prob": 0.
|
| 11 |
"hidden_size": 1024,
|
| 12 |
"id2label": {
|
| 13 |
"0": "LABEL_0",
|
|
|
|
| 4 |
"BertForSequenceClassification"
|
| 5 |
],
|
| 6 |
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": 0.2,
|
| 8 |
"gradient_checkpointing": false,
|
| 9 |
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.0,
|
| 11 |
"hidden_size": 1024,
|
| 12 |
"id2label": {
|
| 13 |
"0": "LABEL_0",
|
tf_model.h5
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1341734528
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:109bc563c3d61ee5d58f29995d7c8506991d78efc579ce4159f0a1ae0cf444bd
|
| 3 |
size 1341734528
|