Instructions to use basmalaazab/asl-fingerspelling-transformer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use basmalaazab/asl-fingerspelling-transformer with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://basmalaazab/asl-fingerspelling-transformer") - Notebooks
- Google Colab
- Kaggle
| { | |
| "architecture": "Transformer (HandTalk)", | |
| "framework": "tensorflow/keras", | |
| "num_hid": 200, | |
| "num_head": 4, | |
| "num_feed_forward": 400, | |
| "source_maxlen": 100, | |
| "target_maxlen": 64, | |
| "num_layers_enc": 2, | |
| "num_layers_dec": 1, | |
| "num_classes": 62, | |
| "pad_token_idx": 59, | |
| "start_token_idx": 60, | |
| "end_token_idx": 61, | |
| "frame_len": 128, | |
| "note": "These values were reverse-engineered from the actual weight shapes in transformer_weights.h5 (not the class defaults in modeling.py), so they match your trained model exactly." | |
| } | |