MHGanainy/mgpt-lora-multi-shared-1024

README.md

@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
 # mgpt-lora-multi-shared-1024
 
 This model is a fine-tuned version of [ai-forever/mGPT](https://huggingface.co/ai-forever/mGPT) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.8008
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 1.0,
-    "eval_loss": 3.7963714599609375,
-    "eval_runtime": 10.5803,
-    "eval_samples_per_second": 9.451,
-    "eval_steps_per_second": 9.451,
-    "perplexity": 44.53927837463248,
+    "eval_loss": 3.800776720046997,
+    "eval_runtime": 10.5511,
+    "eval_samples_per_second": 9.478,
+    "eval_steps_per_second": 9.478,
+    "perplexity": 44.73591828689196,
     "total_flos": 1527651080601600.0,
-    "train_loss": 4.171203002929688,
-    "train_runtime": 157.8425,
-    "train_samples_per_second": 0.634,
-    "train_steps_per_second": 0.634
+    "train_loss": 4.172185363769532,
+    "train_runtime": 144.7805,
+    "train_samples_per_second": 0.691,
+    "train_steps_per_second": 0.691
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "eval_loss": 3.7963714599609375,
-    "eval_runtime": 10.5803,
-    "eval_samples_per_second": 9.451,
-    "eval_steps_per_second": 9.451,
-    "perplexity": 44.53927837463248
+    "eval_loss": 3.800776720046997,
+    "eval_runtime": 10.5511,
+    "eval_samples_per_second": 9.478,
+    "eval_steps_per_second": 9.478,
+    "perplexity": 44.73591828689196
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
     "total_flos": 1527651080601600.0,
-    "train_loss": 4.171203002929688,
-    "train_runtime": 157.8425,
-    "train_samples_per_second": 0.634,
-    "train_steps_per_second": 0.634
+    "train_loss": 4.172185363769532,
+    "train_runtime": 144.7805,
+    "train_samples_per_second": 0.691,
+    "train_steps_per_second": 0.691
 }

trainer_state.json

@@ -10,19 +10,19 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 3.4095587730407715,
+      "grad_norm": 3.3964641094207764,
       "learning_rate": 0.0,
-      "loss": 4.1712,
+      "loss": 4.1722,
       "step": 100
     },
     {
       "epoch": 1.0,
       "step": 100,
       "total_flos": 1527651080601600.0,
-      "train_loss": 4.171203002929688,
-      "train_runtime": 157.8425,
-      "train_samples_per_second": 0.634,
-      "train_steps_per_second": 0.634
+      "train_loss": 4.172185363769532,
+      "train_runtime": 144.7805,
+      "train_samples_per_second": 0.691,
+      "train_steps_per_second": 0.691
     }
   ],
   "logging_steps": 100,