Instructions to use Azimml/Qwen3-8B-trellis-3bit-webgpu with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Azimml/Qwen3-8B-trellis-3bit-webgpu with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Azimml/Qwen3-8B-trellis-3bit-webgpu", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| {"lm_head.weight.bits": {"offset": 0, "shape": [2430976, 32], "dtype": "u32", "shard": "head.bin"}, "lm_head.weight.su": {"offset": 311164928, "shape": [4096], "dtype": "f16", "shard": "head.bin"}, "lm_head.weight.sv": {"offset": 311173120, "shape": [151936], "dtype": "f16", "shard": "head.bin"}, "lm_head.weight.isc": {"offset": 311476992, "shape": [4096], "dtype": "f16", "shard": "head.bin"}, "lm_head.weight.osc": {"offset": 311485184, "shape": [151936], "dtype": "f16", "shard": "head.bin"}, "lm_head.weight.scale": {"offset": 311789056, "shape": [1], "dtype": "f32", "shard": "head.bin"}, "lm_head.weight": {"quant": true, "n_out": 151936, "k_in": 4096, "K": 4, "block": 128, "tiles_out": 9496, "tiles_in": 256, "shard": "head.bin"}} |