Instructions to use ArthurZ/d3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ArthurZ/d3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("zero-shot-classification", model="ArthurZ/d3")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ArthurZ/d3", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Keep the readout in float32
Browse files- config.json +3 -0
config.json
CHANGED
|
@@ -206,6 +206,9 @@
|
|
| 206 |
"dtype": "bfloat16",
|
| 207 |
"hidden_size": 5120,
|
| 208 |
"initializer_range": 0.02,
|
|
|
|
|
|
|
|
|
|
| 209 |
"model_type": "jev",
|
| 210 |
"num_verbalizers": 255,
|
| 211 |
"offsets": {
|
|
|
|
| 206 |
"dtype": "bfloat16",
|
| 207 |
"hidden_size": 5120,
|
| 208 |
"initializer_range": 0.02,
|
| 209 |
+
"keep_in_fp32_modules": [
|
| 210 |
+
"^proj"
|
| 211 |
+
],
|
| 212 |
"model_type": "jev",
|
| 213 |
"num_verbalizers": 255,
|
| 214 |
"offsets": {
|