End of training

adapter_config.json

@@ -20,11 +20,11 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
+    "gate_proj",
     "q_proj",
-    "o_proj",
     "k_proj",
-    "gate_proj"
+    "o_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

all_results.json

@@ -1,14 +1,14 @@
 {
     "epoch": 5.923076923076923,
     "eval_loss": 1.2003010511398315,
-    "eval_runtime": 30.6491,
+    "eval_runtime": 30.5205,
     "eval_samples": 169,
-    "eval_samples_per_second": 2.904,
-    "eval_steps_per_second": 0.392,
+    "eval_samples_per_second": 2.916,
+    "eval_steps_per_second": 0.393,
     "total_flos": 1.7606154086724403e+17,
     "train_loss": 4.533969319902815e-05,
-    "train_runtime": 4.4385,
+    "train_runtime": 4.3598,
     "train_samples": 1346,
-    "train_samples_per_second": 901.197,
-    "train_steps_per_second": 225.299
+    "train_samples_per_second": 917.469,
+    "train_steps_per_second": 229.367
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.923076923076923,
     "eval_loss": 1.2003010511398315,
-    "eval_runtime": 30.6491,
+    "eval_runtime": 30.5205,
     "eval_samples": 169,
-    "eval_samples_per_second": 2.904,
-    "eval_steps_per_second": 0.392
+    "eval_samples_per_second": 2.916,
+    "eval_steps_per_second": 0.393
 }

train_results.json

@@ -2,8 +2,8 @@
     "epoch": 5.923076923076923,
     "total_flos": 1.7606154086724403e+17,
     "train_loss": 4.533969319902815e-05,
-    "train_runtime": 4.4385,
+    "train_runtime": 4.3598,
     "train_samples": 1346,
-    "train_samples_per_second": 901.197,
-    "train_steps_per_second": 225.299
+    "train_samples_per_second": 917.469,
+    "train_steps_per_second": 229.367
 }

trainer_state.json

@@ -613,9 +613,9 @@
       "step": 1001,
       "total_flos": 1.7606154086724403e+17,
       "train_loss": 4.533969319902815e-05,
-      "train_runtime": 4.4385,
-      "train_samples_per_second": 901.197,
-      "train_steps_per_second": 225.299
+      "train_runtime": 4.3598,
+      "train_samples_per_second": 917.469,
+      "train_steps_per_second": 229.367
     }
   ],
   "logging_steps": 25,