./whisper-medium-gabriel_fold_1

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1796
-- Wer: 3.8664
+- Loss: 0.2852
+- Wer: 15.3917
 
 ## Model description
 
@@ -49,12 +49,12 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 4.8035        | 1.0   | 7    | 2.7446          | 37.4341 |
-| 2.254         | 2.0   | 14   | 1.4997          | 17.9262 |
-| 1.2891        | 3.0   | 21   | 1.0274          | 23.2865 |
-| 0.8979        | 4.0   | 28   | 0.7098          | 20.6503 |
-| 0.5824        | 5.0   | 35   | 0.4147          | 5.4482  |
-| 0.3019        | 6.0   | 42   | 0.1796          | 3.8664  |
+| 0.7658        | 1.0   | 169  | 0.3021          | 14.4355 |
+| 0.1396        | 2.0   | 338  | 0.2658          | 13.8186 |
+| 0.0566        | 3.0   | 507  | 0.2666          | 14.0037 |
+| 0.0252        | 4.0   | 676  | 0.2820          | 18.3837 |
+| 0.0118        | 5.0   | 845  | 0.2801          | 13.6336 |
+| 0.0067        | 6.0   | 1014 | 0.2852          | 15.3917 |
 
 
 ### Framework versions

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ddacab5c42bf5e89d30e05e516b3e8130127fb1c214a9cc2589eebda32213a7c
+oid sha256:e638d0ceded8b75bc996eb1263f195684f26b4058e0d1d378a989d46af4a0401
 size 3055754969