Instructions to use ArthurZ/clef-flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ArthurZ/clef-flash with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ArthurZ/clef-flash", device_map="auto") - Notebooks
- Google Colab
- Kaggle
decision_head: the original joint_head.safetensors, loaded through the conversion mapping
Browse files- decision_head/config.json +11 -9
- decision_head/model.safetensors +2 -2
decision_head/config.json
CHANGED
|
@@ -1,17 +1,19 @@
|
|
| 1 |
{
|
| 2 |
-
"activation_function": "gelu",
|
| 3 |
"architectures": [
|
| 4 |
"ClefDecisionHead"
|
| 5 |
],
|
| 6 |
-
"
|
|
|
|
|
|
|
| 7 |
"dtype": "bfloat16",
|
| 8 |
-
"
|
| 9 |
-
"
|
| 10 |
-
"hidden_size": 4096,
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
-
"
|
| 13 |
"model_type": "clef",
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
|
|
|
|
|
|
| 17 |
}
|
|
|
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"ClefDecisionHead"
|
| 4 |
],
|
| 5 |
+
"attention_bias": true,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"backbone_hidden_size": 4096,
|
| 8 |
"dtype": "bfloat16",
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_size": 1024,
|
|
|
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4096,
|
| 13 |
"model_type": "clef",
|
| 14 |
+
"num_attention_heads": 16,
|
| 15 |
+
"num_evidence_layers": 2,
|
| 16 |
+
"num_hidden_layers": 4,
|
| 17 |
+
"num_key_value_heads": 16,
|
| 18 |
+
"transformers_version": "5.19.0.dev0"
|
| 19 |
}
|
decision_head/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19cdcec8c81dc9212be320fff47462ab342fbc1278be4368fb3da71241cf5ba0
|
| 3 |
+
size 243538016
|