End of training
Browse files- README.md +12 -13
- generation_config.json +1 -1
- model.safetensors +1 -1
README.md
CHANGED
|
@@ -11,17 +11,17 @@ datasets:
|
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
model-index:
|
| 14 |
-
- name: Whisper Tiny
|
| 15 |
results:
|
| 16 |
- task:
|
| 17 |
name: Automatic Speech Recognition
|
| 18 |
type: automatic-speech-recognition
|
| 19 |
dataset:
|
| 20 |
-
name: Moroccan
|
| 21 |
type: Yassinevic/Moroccan-Darija-HSIO
|
| 22 |
config: default
|
| 23 |
split: None
|
| 24 |
-
args: 'config:
|
| 25 |
metrics:
|
| 26 |
- name: Wer
|
| 27 |
type: wer
|
|
@@ -31,11 +31,11 @@ model-index:
|
|
| 31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 32 |
should probably proofread and complete it, then remove this comment. -->
|
| 33 |
|
| 34 |
-
# Whisper Tiny
|
| 35 |
|
| 36 |
-
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Moroccan
|
| 37 |
It achieves the following results on the evaluation set:
|
| 38 |
-
- Loss: 0.
|
| 39 |
- Wer: 0.0
|
| 40 |
|
| 41 |
## Model description
|
|
@@ -61,19 +61,18 @@ The following hyperparameters were used during training:
|
|
| 61 |
- seed: 42
|
| 62 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 63 |
- lr_scheduler_type: linear
|
| 64 |
-
- lr_scheduler_warmup_steps:
|
| 65 |
-
- training_steps:
|
| 66 |
- mixed_precision_training: Native AMP
|
| 67 |
|
| 68 |
### Training results
|
| 69 |
|
| 70 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 71 |
|:-------------:|:--------:|:----:|:---------------:|:---:|
|
| 72 |
-
| 0.
|
| 73 |
-
| 0.
|
| 74 |
-
| 0.
|
| 75 |
-
| 0.
|
| 76 |
-
| 0.0001 | 333.3333 | 1000 | 0.0001 | 0.0 |
|
| 77 |
|
| 78 |
|
| 79 |
### Framework versions
|
|
|
|
| 11 |
metrics:
|
| 12 |
- wer
|
| 13 |
model-index:
|
| 14 |
+
- name: Whisper Tiny AR - Yassinevic
|
| 15 |
results:
|
| 16 |
- task:
|
| 17 |
name: Automatic Speech Recognition
|
| 18 |
type: automatic-speech-recognition
|
| 19 |
dataset:
|
| 20 |
+
name: Moroccan Darija HSIO
|
| 21 |
type: Yassinevic/Moroccan-Darija-HSIO
|
| 22 |
config: default
|
| 23 |
split: None
|
| 24 |
+
args: 'config: hi, split: test'
|
| 25 |
metrics:
|
| 26 |
- name: Wer
|
| 27 |
type: wer
|
|
|
|
| 31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 32 |
should probably proofread and complete it, then remove this comment. -->
|
| 33 |
|
| 34 |
+
# Whisper Tiny AR - Yassinevic
|
| 35 |
|
| 36 |
+
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Moroccan Darija HSIO dataset.
|
| 37 |
It achieves the following results on the evaluation set:
|
| 38 |
+
- Loss: 0.0000
|
| 39 |
- Wer: 0.0
|
| 40 |
|
| 41 |
## Model description
|
|
|
|
| 61 |
- seed: 42
|
| 62 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 63 |
- lr_scheduler_type: linear
|
| 64 |
+
- lr_scheduler_warmup_steps: 500
|
| 65 |
+
- training_steps: 4000
|
| 66 |
- mixed_precision_training: Native AMP
|
| 67 |
|
| 68 |
### Training results
|
| 69 |
|
| 70 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 71 |
|:-------------:|:--------:|:----:|:---------------:|:---:|
|
| 72 |
+
| 0.0001 | 166.6667 | 1000 | 0.0001 | 0.0 |
|
| 73 |
+
| 0.0 | 333.3333 | 2000 | 0.0001 | 0.0 |
|
| 74 |
+
| 0.0 | 500.0 | 3000 | 0.0000 | 0.0 |
|
| 75 |
+
| 0.0 | 666.6667 | 4000 | 0.0000 | 0.0 |
|
|
|
|
| 76 |
|
| 77 |
|
| 78 |
### Framework versions
|
generation_config.json
CHANGED
|
@@ -134,7 +134,7 @@
|
|
| 134 |
"<|yo|>": 50325,
|
| 135 |
"<|zh|>": 50260
|
| 136 |
},
|
| 137 |
-
"language": "
|
| 138 |
"max_initial_timestamp_index": 50,
|
| 139 |
"max_length": 448,
|
| 140 |
"no_timestamps_token_id": 50363,
|
|
|
|
| 134 |
"<|yo|>": 50325,
|
| 135 |
"<|zh|>": 50260
|
| 136 |
},
|
| 137 |
+
"language": "Arabic",
|
| 138 |
"max_initial_timestamp_index": 50,
|
| 139 |
"max_length": 448,
|
| 140 |
"no_timestamps_token_id": 50363,
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 151061672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:865af8df61d29d728cdbb2a9a015b63d568f588e5ef0890a41c845e9e1242eb5
|
| 3 |
size 151061672
|