CuATR-distilbert-LoRA

README.md

@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7010
-- Accuracy: 0.4348
-- F1: 0.0
+- Loss: 0.6890
+- Accuracy: 0.5652
+- F1: 0.7222
 
 ## Model description
 
@@ -52,11 +52,11 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1  |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|:---:|
-| 0.6952        | 0.67  | 1    | 0.7016          | 0.4348   | 0.0 |
-| 0.6907        | 2.0   | 3    | 0.7012          | 0.4348   | 0.0 |
-| 0.719         | 2.67  | 4    | 0.7010          | 0.4348   | 0.0 |
+| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
+| 0.7122        | 0.67  | 1    | 0.6891          | 0.5652   | 0.7222 |
+| 0.7112        | 2.0   | 3    | 0.6890          | 0.5652   | 0.7222 |
+| 0.6955        | 2.67  | 4    | 0.6890          | 0.5652   | 0.7222 |
 
 
 ### Framework versions

adapter_config.json

@@ -2,22 +2,22 @@
   "alpha_pattern": {},
   "auto_mapping": null,
   "base_model_name_or_path": "distilbert-base-uncased",
-  "bias": "all",
+  "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": false,
   "init_lora_weights": true,
   "layers_pattern": null,
   "layers_to_transform": null,
-  "lora_alpha": 32,
-  "lora_dropout": 0.05,
+  "lora_alpha": 16,
+  "lora_dropout": 0.1,
   "modules_to_save": null,
   "peft_type": "LORA",
   "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_lin",
     "v_lin",
+    "k_lin",
     "q_lin"
   ],
   "task_type": "TOKEN_CLS"

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2bcd6bb8a5da04a0abf779344bb4cddd008046bb80a11ee65f83cb468e12a606
-size 1990240
+oid sha256:7b42e653456701e10c61c62e1d9ee2a365ae65a2669fde89dfc4d8dd35ae27cc
+size 1774720

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5c0d322e926af0aab09913b96d93544750b6d0852ba0b4f735925101ec3d6d2
+oid sha256:c85194ab166096f43ad5356cbba09fbe00e9f9707ef5784691ae1932410d2900
 size 4600