Upload model
Browse files- calibration_head.bin +3 -0
- config.json +24 -0
calibration_head.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ac7cca42ec746e9a5b1454ab4231a3aec647393fcfeabeb8b0c156b25a445db
|
| 3 |
+
size 1268842620
|
config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"attention_dropout": 0.0,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-13b-chat-hf",
|
| 4 |
+
"calibration_type": "transformer",
|
| 5 |
+
"feature_key": "hidden_states",
|
| 6 |
+
"freeze_base_model": true,
|
| 7 |
+
"hidden_act": "silu",
|
| 8 |
+
"in_features": 5120,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_temperature": 1.0,
|
| 11 |
+
"intermediate_size": 13824,
|
| 12 |
+
"label_smoothing": 0.0,
|
| 13 |
+
"label_smoothing_type": "topk",
|
| 14 |
+
"layer_idx": 40,
|
| 15 |
+
"log_auxiliary_info": false,
|
| 16 |
+
"loss_type": "xent",
|
| 17 |
+
"max_position_embeddings": 5120,
|
| 18 |
+
"normalize_logits": false,
|
| 19 |
+
"num_attention_heads": 40,
|
| 20 |
+
"num_key_value_heads": 40,
|
| 21 |
+
"smooth_loss_weight": 0.5,
|
| 22 |
+
"smoothing_topk": 5,
|
| 23 |
+
"task_type": "CAUSAL_LM"
|
| 24 |
+
}
|