End of training

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [mikhail-panzo/malay_full_checkpoint](https://huggingface.co/mikhail-panzo/malay_full_checkpoint) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4194
+- Loss: 0.4181
 
 ## Model description
 
@@ -34,30 +34,26 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 1e-06
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
-- gradient_accumulation_steps: 2
-- total_train_batch_size: 32
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 64
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- training_steps: 4000
+- training_steps: 2000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 0.5034        | 11.24 | 500  | 0.4641          |
-| 0.4676        | 22.47 | 1000 | 0.4385          |
-| 0.4584        | 33.71 | 1500 | 0.4276          |
-| 0.4456        | 44.94 | 2000 | 0.4247          |
-| 0.448         | 56.18 | 2500 | 0.4216          |
-| 0.4431        | 67.42 | 3000 | 0.4199          |
-| 0.4358        | 78.65 | 3500 | 0.4198          |
-| 0.4314        | 89.89 | 4000 | 0.4194          |
+| 0.4345        | 22.47 | 500  | 0.4195          |
+| 0.4333        | 44.94 | 1000 | 0.4181          |
+| 0.4303        | 67.42 | 1500 | 0.4189          |
+| 0.4354        | 89.89 | 2000 | 0.4181          |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c078a2d862275fd8501374372313d1198ccc5c6c8ea8287cfe14f66afe8063e
+oid sha256:510be518f032cce9d25770f36142d87fe70d7ec2ccb8b9892914081cb5fa3a20
 size 577789320