End of training

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/rembert](https://huggingface.co/google/rembert) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6085
+- Loss: 0.8181
 - Accuracy: 0.9441
 - 1-f1: 0.0
 - 1-recall: 0.0
@@ -48,20 +48,18 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.06
 - num_epochs: 20
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Accuracy | 1-f1 | 1-recall | 1-precision | Balanced Acc |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|:----:|:--------:|:-----------:|:------------:|
-| 0.4918        | 1.0   | 108  | 0.6222          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.6322        | 2.0   | 216  | 0.6152          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.575         | 3.0   | 324  | 0.6253          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.5098        | 4.0   | 432  | 0.6090          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.6741        | 5.0   | 540  | 0.6079          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.3941        | 6.0   | 648  | 0.6104          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
-| 0.3901        | 7.0   | 756  | 0.6085          | 0.9441   | 0.0  | 0.0      | 0.0         | 0.5          |
+| Training Loss | Epoch | Step | Validation Loss | Accuracy | 1-f1   | 1-recall | 1-precision | Balanced Acc |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:--------:|:-----------:|:------------:|
+| 0.2947        | 1.0   | 108  | 0.6166          | 0.9441   | 0.0    | 0.0      | 0.0         | 0.5          |
+| 0.9524        | 2.0   | 216  | 0.5859          | 0.9441   | 0.0    | 0.0      | 0.0         | 0.5          |
+| 0.5436        | 3.0   | 324  | 0.7579          | 0.2564   | 0.1064 | 0.7917   | 0.0571      | 0.5082       |
+| 0.7444        | 4.0   | 432  | 0.8181          | 0.9441   | 0.0    | 0.0      | 0.0         | 0.5          |
 
 
 ### Framework versions

config.json

@@ -3,14 +3,14 @@
   "architectures": [
     "RemBertForSequenceClassification"
   ],
-  "attention_probs_dropout_prob": 0,
+  "attention_probs_dropout_prob": 0.2,
   "bos_token_id": 312,
   "classifier_dropout_prob": 0.1,
   "embedding_dropout_prob": 0,
   "embedding_size": 256,
   "eos_token_id": 313,
   "hidden_act": "gelu",
-  "hidden_dropout_prob": 0,
+  "hidden_dropout_prob": 0.2,
   "hidden_size": 1152,
   "initializer_range": 0.02,
   "input_embedding_size": 256,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e52b04a0fa505d77c83ac30f1dc67c895a654c6498bd9b3bc38b73a510fa4af6
+oid sha256:4ed2092e93c5e67d88a929bb420f3779e35801f5dc26861d6c66d1f32fd3b745
 size 2303755328

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:deb4a9b024c0be243b1b5673c1015b3767c8d01a4a0118ca675c95baeb863dc5
+oid sha256:b14041930874d06f07d4627b98daac4585fa4bc4618a50414dfa96e30aa11dfa
 size 5432