How to use SaraHe/aya_compress_0_31_two_layers with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("SaraHe/aya_compress_0_31_two_layers", device_map="auto")
