End of training

README.md

@@ -16,8 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [ad019el/Kabyle_xlsr-finetuned-tamasheq.en](https://huggingface.co/ad019el/Kabyle_xlsr-finetuned-tamasheq.en) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6805
-- Wer: 0.8586
+- Loss: 1.5106
+- Wer: 0.8739
 
 ## Model description
 
@@ -37,11 +37,11 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 3e-05
-- train_batch_size: 16
+- train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
 - gradient_accumulation_steps: 2
-- total_train_batch_size: 32
+- total_train_batch_size: 16
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
@@ -51,12 +51,11 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Wer    |
 |:-------------:|:-----:|:----:|:---------------:|:------:|
-| 5.637         | 7.89  | 300  | 2.8478          | 1.0    |
-| 1.8487        | 15.79 | 600  | 1.5027          | 0.9196 |
-| 0.6213        | 23.68 | 900  | 1.4700          | 0.8810 |
-| 0.4539        | 31.58 | 1200 | 1.5753          | 0.8824 |
-| 0.3551        | 39.47 | 1500 | 1.6545          | 0.8690 |
-| 0.3079        | 47.37 | 1800 | 1.6805          | 0.8586 |
+| 3.637         | 3.41  | 300  | 1.9770          | 1.0348 |
+| 1.1027        | 6.82  | 600  | 1.3353          | 0.8913 |
+| 0.7367        | 10.23 | 900  | 1.4064          | 0.8855 |
+| 0.6037        | 13.64 | 1200 | 1.3830          | 0.8768 |
+| 0.529         | 17.05 | 1500 | 1.5106          | 0.8739 |
 
 
 ### Framework versions

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d68fceee99132cd20a8def37f923bacf8dfa77291452bd42eac4f2dfb3b2b346
+oid sha256:5114cf210d7f23f3e9e3a8f958fb1af31ec50a1680a69ada5d8c774d4444a6d9
 size 1262020717