Instructions to use opencampus/sign-whisper-german with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use opencampus/sign-whisper-german with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("opencampus/sign-whisper-german", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Christopher H. commited on
Update README.md
Browse files
README.md
CHANGED
|
@@ -106,9 +106,12 @@ val_dataset = YourSignDataset(...)
|
|
| 106 |
# Define training arguments
|
| 107 |
training_args = TrainingArguments(
|
| 108 |
output_dir="./sign-whisper-german",
|
|
|
|
| 109 |
num_train_epochs=3,
|
| 110 |
per_device_train_batch_size=1024,
|
| 111 |
per_device_eval_batch_size=256,
|
|
|
|
|
|
|
| 112 |
warmup_steps=500,
|
| 113 |
weight_decay=0.01,
|
| 114 |
|
|
|
|
| 106 |
# Define training arguments
|
| 107 |
training_args = TrainingArguments(
|
| 108 |
output_dir="./sign-whisper-german",
|
| 109 |
+
|
| 110 |
num_train_epochs=3,
|
| 111 |
per_device_train_batch_size=1024,
|
| 112 |
per_device_eval_batch_size=256,
|
| 113 |
+
|
| 114 |
+
learning_rate=1e-5
|
| 115 |
warmup_steps=500,
|
| 116 |
weight_decay=0.01,
|
| 117 |
|