End of training
Browse files
README.md
CHANGED
|
@@ -1,8 +1,10 @@
|
|
| 1 |
---
|
| 2 |
-
base_model: NbAiLab/nb-whisper-small-verbatim
|
| 3 |
license: apache-2.0
|
|
|
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: nb-whisper-small-karelian-CodeSwitching
|
| 8 |
results: []
|
|
@@ -15,14 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [NbAiLab/nb-whisper-small-verbatim](https://huggingface.co/NbAiLab/nb-whisper-small-verbatim) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
-
|
| 19 |
-
-
|
| 20 |
-
-
|
| 21 |
-
- eval_runtime: 86.5964
|
| 22 |
-
- eval_samples_per_second: 4.966
|
| 23 |
-
- eval_steps_per_second: 0.162
|
| 24 |
-
- epoch: 0.4535
|
| 25 |
-
- step: 100
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
|
@@ -41,19 +38,43 @@ More information needed
|
|
| 41 |
### Training hyperparameters
|
| 42 |
|
| 43 |
The following hyperparameters were used during training:
|
| 44 |
-
- learning_rate:
|
| 45 |
- train_batch_size: 32
|
| 46 |
- eval_batch_size: 32
|
| 47 |
- seed: 42
|
| 48 |
- distributed_type: multi-GPU
|
| 49 |
-
- gradient_accumulation_steps: 2
|
| 50 |
-
- total_train_batch_size: 64
|
| 51 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
- lr_scheduler_type: linear
|
| 53 |
- lr_scheduler_warmup_steps: 500
|
| 54 |
-
- training_steps:
|
| 55 |
- mixed_precision_training: Native AMP
|
| 56 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
### Framework versions
|
| 58 |
|
| 59 |
- Transformers 4.41.2
|
|
|
|
| 1 |
---
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
+
base_model: NbAiLab/nb-whisper-small-verbatim
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- wer
|
| 8 |
model-index:
|
| 9 |
- name: nb-whisper-small-karelian-CodeSwitching
|
| 10 |
results: []
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [NbAiLab/nb-whisper-small-verbatim](https://huggingface.co/NbAiLab/nb-whisper-small-verbatim) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.7241
|
| 21 |
+
- Wer: 0.3278
|
| 22 |
+
- Cer: 0.0999
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 38 |
### Training hyperparameters
|
| 39 |
|
| 40 |
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 5e-05
|
| 42 |
- train_batch_size: 32
|
| 43 |
- eval_batch_size: 32
|
| 44 |
- seed: 42
|
| 45 |
- distributed_type: multi-GPU
|
|
|
|
|
|
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
- lr_scheduler_warmup_steps: 500
|
| 49 |
+
- training_steps: 10000
|
| 50 |
- mixed_precision_training: Native AMP
|
| 51 |
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|
| 55 |
+
|:-------------:|:-------:|:-----:|:---------------:|:------:|:------:|
|
| 56 |
+
| 0.1848 | 1.1338 | 500 | 0.6758 | 0.4066 | 0.1227 |
|
| 57 |
+
| 0.0796 | 2.2676 | 1000 | 0.6685 | 0.3549 | 0.1015 |
|
| 58 |
+
| 0.0492 | 3.4014 | 1500 | 0.7090 | 0.3758 | 0.1088 |
|
| 59 |
+
| 0.0426 | 4.5351 | 2000 | 0.7095 | 0.3704 | 0.1083 |
|
| 60 |
+
| 0.0311 | 5.6689 | 2500 | 0.7148 | 0.3742 | 0.1138 |
|
| 61 |
+
| 0.027 | 6.8027 | 3000 | 0.7173 | 0.3454 | 0.1004 |
|
| 62 |
+
| 0.0201 | 7.9365 | 3500 | 0.7261 | 0.3813 | 0.1325 |
|
| 63 |
+
| 0.0165 | 9.0703 | 4000 | 0.7158 | 0.3417 | 0.0982 |
|
| 64 |
+
| 0.0179 | 10.2041 | 4500 | 0.7261 | 0.3495 | 0.1036 |
|
| 65 |
+
| 0.0101 | 11.3379 | 5000 | 0.7275 | 0.3315 | 0.0978 |
|
| 66 |
+
| 0.0086 | 12.4717 | 5500 | 0.7437 | 0.3400 | 0.1081 |
|
| 67 |
+
| 0.0058 | 13.6054 | 6000 | 0.7524 | 0.3410 | 0.1026 |
|
| 68 |
+
| 0.0056 | 14.7392 | 6500 | 0.7256 | 0.3407 | 0.1015 |
|
| 69 |
+
| 0.0075 | 15.8730 | 7000 | 0.7202 | 0.3312 | 0.0987 |
|
| 70 |
+
| 0.0047 | 17.0068 | 7500 | 0.7266 | 0.3359 | 0.1025 |
|
| 71 |
+
| 0.0046 | 18.1406 | 8000 | 0.7271 | 0.3312 | 0.0973 |
|
| 72 |
+
| 0.0039 | 19.2744 | 8500 | 0.7334 | 0.3353 | 0.0999 |
|
| 73 |
+
| 0.0025 | 20.4082 | 9000 | 0.7280 | 0.3295 | 0.0987 |
|
| 74 |
+
| 0.0022 | 21.5420 | 9500 | 0.7290 | 0.3254 | 0.0972 |
|
| 75 |
+
| 0.0031 | 22.6757 | 10000 | 0.7241 | 0.3278 | 0.0999 |
|
| 76 |
+
|
| 77 |
+
|
| 78 |
### Framework versions
|
| 79 |
|
| 80 |
- Transformers 4.41.2
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 563189936
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:407a594762d1fa67dfc1c474d5eb1258c96eb16596f954739ec054f0d4cc9aed
|
| 3 |
size 563189936
|
runs/Jun21_15-27-06_maxim-ubuntu3080/events.out.tfevents.1718972838.maxim-ubuntu3080.217531.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d412f642094e04f5266bb876dd458c66438e00c7a31baddbe062b510a27dd523
|
| 3 |
+
size 100161
|