Update README.md
Browse files
README.md
CHANGED
|
@@ -10,6 +10,9 @@ metrics:
|
|
| 10 |
model-index:
|
| 11 |
- name: Whisper large-v2 CV18 Hu
|
| 12 |
results: []
|
|
|
|
|
|
|
|
|
|
| 13 |
---
|
| 14 |
|
| 15 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -18,11 +21,21 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
# Whisper large-v2 CV18 Hu
|
| 19 |
|
| 20 |
This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the fsicoli/common_voice_18_0 dataset.
|
| 21 |
-
It achieves the following results on the evaluation set:
|
| 22 |
- Loss: 0.3493
|
| 23 |
- Wer Ortho: 21.9936
|
| 24 |
- Wer: 16.0057
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
## Model description
|
| 27 |
|
| 28 |
More information needed
|
|
@@ -92,4 +105,4 @@ The following hyperparameters were used during training:
|
|
| 92 |
- Transformers 4.34.1
|
| 93 |
- Pytorch 2.3.0+cu121
|
| 94 |
- Datasets 2.14.6
|
| 95 |
-
- Tokenizers 0.14.1
|
|
|
|
| 10 |
model-index:
|
| 11 |
- name: Whisper large-v2 CV18 Hu
|
| 12 |
results: []
|
| 13 |
+
datasets:
|
| 14 |
+
- fsicoli/common_voice_18_0
|
| 15 |
+
- google/fleurs
|
| 16 |
---
|
| 17 |
|
| 18 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 21 |
# Whisper large-v2 CV18 Hu
|
| 22 |
|
| 23 |
This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the fsicoli/common_voice_18_0 dataset.
|
| 24 |
+
It achieves the following results on the evaluation google/fleurs set:
|
| 25 |
- Loss: 0.3493
|
| 26 |
- Wer Ortho: 21.9936
|
| 27 |
- Wer: 16.0057
|
| 28 |
|
| 29 |
+
Kvantált modellek eredményei:
|
| 30 |
+
| Model | WER | CER | Normalized_WER | Normalized_CER | Database | Split | Runtime |
|
| 31 |
+
|:---------------------------------------------------------|:------|:------|:-----------------|:-----------------|:--------------|:--------|:----------|
|
| 32 |
+
| int8_bfloat16 | 21.49 | 5.93 | 16.04 | 6.21 | google/fleurs | test | 550.18 |
|
| 33 |
+
| bfloat16 | 21.33 | 5.87 | 15.91 | 6.15 | google/fleurs | test | 593.96 |
|
| 34 |
+
| int8 | 21.01 | 5.63 | 15.38 | 5.88 | google/fleurs | test | 668.91 |
|
| 35 |
+
| int8_float32 | 21.01 | 5.63 | 15.38 | 5.88 | google/fleurs | test | 669.81 |
|
| 36 |
+
| int8_float16 | 20.96 | 5.65 | 15.31 | 5.91 | google/fleurs | test | 570.11 |
|
| 37 |
+
| float16 | 20.92 | 5.64 | 15.24 | 5.9 | google/fleurs | test | 589.29 |
|
| 38 |
+
|
| 39 |
## Model description
|
| 40 |
|
| 41 |
More information needed
|
|
|
|
| 105 |
- Transformers 4.34.1
|
| 106 |
- Pytorch 2.3.0+cu121
|
| 107 |
- Datasets 2.14.6
|
| 108 |
+
- Tokenizers 0.14.1
|