Finished training.

README.md

@@ -23,7 +23,7 @@ model-index:
       args: irony
     metrics:
     - type: accuracy
-      value: 0.6471204188481675
+      value: 0.6418848167539267
       name: accuracy
 ---
 
@@ -34,7 +34,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [neibla/distilbert-base-uncased-finetuned-emotion](https://huggingface.co/neibla/distilbert-base-uncased-finetuned-emotion) on the tweet_eval dataset.
 It achieves the following results on the evaluation set:
-- accuracy: 0.6471
+- accuracy: 0.6419
 
 ## Model description
 
@@ -65,15 +65,15 @@ The following hyperparameters were used during training:
 
 | accuracy | train_loss | epoch |
 |:--------:|:----------:|:-----:|
-| 0.5529   | None       | 0     |
-| 0.5654   | 0.6994     | 0     |
-| 0.5937   | 0.6503     | 1     |
-| 0.6335   | 0.6247     | 2     |
-| 0.6346   | 0.5982     | 3     |
-| 0.6230   | 0.5733     | 4     |
-| 0.6304   | 0.5528     | 5     |
-| 0.6440   | 0.5347     | 6     |
-| 0.6471   | 0.5227     | 7     |
+| 0.5183   | None       | 0     |
+| 0.5791   | 0.7012     | 0     |
+| 0.5822   | 0.6512     | 1     |
+| 0.6230   | 0.6258     | 2     |
+| 0.6513   | 0.5896     | 3     |
+| 0.6293   | 0.5683     | 4     |
+| 0.6356   | 0.5461     | 5     |
+| 0.6304   | 0.5312     | 6     |
+| 0.6419   | 0.5166     | 7     |
 
 
 ### Framework versions

adapter_config.json

@@ -19,8 +19,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_lin",
-    "v_lin"
+    "v_lin",
+    "q_lin"
   ],
   "task_type": "SEQ_CLS",
   "use_rslora": false

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9de7c8bce603e7ecccd3167ac8365bb169084cf9a35fa4288c49d9645b493f7
+oid sha256:41de0476718748e7ae89ae770b1f558318be85298a6ea455b58ea4bdc750ebed
 size 2446056