better version
Browse files- config.json +30 -0
- pytorch_model.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "microsoft/deberta-v3-base",
|
| 3 |
+
"hidden_size": 512,
|
| 4 |
+
"ffn_mul": 2,
|
| 5 |
+
"span_mode": "markerV0",
|
| 6 |
+
"dropout": 0.1,
|
| 7 |
+
"max_width": 8,
|
| 8 |
+
"max_top_k": 54,
|
| 9 |
+
"add_top_k": 10,
|
| 10 |
+
"lr_encoder": "1e-5",
|
| 11 |
+
"lr_others": "1e-4",
|
| 12 |
+
"freeze_token_rep": true,
|
| 13 |
+
"num_steps": 5000,
|
| 14 |
+
"warmup_ratio": 500,
|
| 15 |
+
"train_batch_size": 12,
|
| 16 |
+
"eval_every": 2500,
|
| 17 |
+
"max_neg_type_ratio": 3,
|
| 18 |
+
"size_sup": -1,
|
| 19 |
+
"max_len": 384,
|
| 20 |
+
"max_types": 30,
|
| 21 |
+
"shuffle_types": false,
|
| 22 |
+
"random_drop": false,
|
| 23 |
+
"refine_span": false,
|
| 24 |
+
"refine_relation": false,
|
| 25 |
+
"refine_prompt": true,
|
| 26 |
+
"scorer": "dot_norm",
|
| 27 |
+
"fine_tune": true,
|
| 28 |
+
"subtoken_pooling": "first",
|
| 29 |
+
"name": "large"
|
| 30 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4e9d873c9eede2d1bc3735a7b742b7454086679c5cbda06b4bd855ff435a8a9
|
| 3 |
+
size 786662270
|