Upload folder using huggingface_hub

README.md

@@ -22,7 +22,7 @@ Parameter-efficient fine-tuning of Mixture-of-Experts using **LoRA (Low-Rank Ada
 
 - **Validation Accuracy**: 0.6400
 - **Dataset**: XSum (topic classification)
-- **Training Samples**: 4,000
+- **Training Samples**: 5,000
 
 ## LoRA Benefits
 

history.csv

@@ -1,6 +1,6 @@
 epoch,train_loss,train_accuracy,val_loss,val_accuracy
-1,0.8074325952529907,0.62525,0.8184478509426117,0.64
-2,0.7937552418708801,0.637,0.7908735847473145,0.64
-3,0.7901616661548615,0.6455,0.798002507686615,0.64
-4,0.7901241521835327,0.6365,0.8332968425750732,0.64
-5,0.7865016897916793,0.6465,0.7994629460573196,0.64
+1,0.8147811661720276,0.6266,0.8106175279617309,0.64
+2,0.8049529413223266,0.6282,0.8057486724853515,0.64
+3,0.79402887840271,0.6384,0.8648435598611832,0.64
+4,0.7893773549079895,0.6438,0.7850593781471252,0.64
+5,0.7901758761405945,0.6414,0.7927370357513428,0.64

lora_adapters.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1cc75c50d7fd0e92374fc126b34a569515c0b2dca5e282ccd8466ce563c41d31
+oid sha256:aab3694f32c175f877bba6583c7ce94772f58557aa09a47a2da7c9d975c64e37
 size 6334282

metrics.json

@@ -2,37 +2,37 @@
   "history": [
     {
       "epoch": 1,
-      "train_loss": 0.8074325952529907,
-      "train_accuracy": 0.62525,
-      "val_loss": 0.8184478509426117,
+      "train_loss": 0.8147811661720276,
+      "train_accuracy": 0.6266,
+      "val_loss": 0.8106175279617309,
       "val_accuracy": 0.64
     },
     {
       "epoch": 2,
-      "train_loss": 0.7937552418708801,
-      "train_accuracy": 0.637,
-      "val_loss": 0.7908735847473145,
+      "train_loss": 0.8049529413223266,
+      "train_accuracy": 0.6282,
+      "val_loss": 0.8057486724853515,
       "val_accuracy": 0.64
     },
     {
       "epoch": 3,
-      "train_loss": 0.7901616661548615,
-      "train_accuracy": 0.6455,
-      "val_loss": 0.798002507686615,
+      "train_loss": 0.79402887840271,
+      "train_accuracy": 0.6384,
+      "val_loss": 0.8648435598611832,
       "val_accuracy": 0.64
     },
     {
       "epoch": 4,
-      "train_loss": 0.7901241521835327,
-      "train_accuracy": 0.6365,
-      "val_loss": 0.8332968425750732,
+      "train_loss": 0.7893773549079895,
+      "train_accuracy": 0.6438,
+      "val_loss": 0.7850593781471252,
       "val_accuracy": 0.64
     },
     {
       "epoch": 5,
-      "train_loss": 0.7865016897916793,
-      "train_accuracy": 0.6465,
-      "val_loss": 0.7994629460573196,
+      "train_loss": 0.7901758761405945,
+      "train_accuracy": 0.6414,
+      "val_loss": 0.7927370357513428,
       "val_accuracy": 0.64
     }
   ],
@@ -57,13 +57,13 @@
     "total": 55228676
   },
   "expert_usage": [
-    270.3500061035156,
-    583.625,
-    598.9650268554688,
-    359.67999267578125,
-    425.7900085449219,
-    603.489990234375,
-    1022.885009765625,
-    231.21499633789062
+    44.994998931884766,
+    48.994998931884766,
+    456.364990234375,
+    263.5,
+    714.1749877929688,
+    1520.7249755859375,
+    810.1400146484375,
+    237.10499572753906
   ]
 }

model.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53be695e8fa32b0f3f871d66bcf00394b365fd99e0c747ad7a2db73979d059cd
+oid sha256:c0f535a9b58723ef46475aa6861a36acc30e6249787e5866a7780cb7fd760907
 size 221009538

tokenizer/tokenizer_config.json

@@ -41,9 +41,10 @@
       "special": true
     }
   },
-  "clean_up_tokenization_spaces": true,
+  "clean_up_tokenization_spaces": false,
   "cls_token": "[CLS]",
   "do_lower_case": true,
+  "extra_special_tokens": {},
   "mask_token": "[MASK]",
   "model_max_length": 512,
   "pad_token": "[PAD]",