End of training

adapter_config.json

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

all_results.json

@@ -1,14 +1,14 @@
 {
     "epoch": 5.923076923076923,
     "eval_loss": 1.2003010511398315,
-    "eval_runtime": 31.1108,
+    "eval_runtime": 30.8916,
     "eval_samples": 169,
-    "eval_samples_per_second": 2.861,
-    "eval_steps_per_second": 0.386,
+    "eval_samples_per_second": 2.881,
+    "eval_steps_per_second": 0.388,
     "total_flos": 1.7606154086724403e+17,
     "train_loss": 4.533969319902815e-05,
-    "train_runtime": 4.4642,
+    "train_runtime": 4.5265,
     "train_samples": 1346,
-    "train_samples_per_second": 896.022,
-    "train_steps_per_second": 224.006
+    "train_samples_per_second": 883.679,
+    "train_steps_per_second": 220.92
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.923076923076923,
     "eval_loss": 1.2003010511398315,
-    "eval_runtime": 31.1108,
+    "eval_runtime": 30.8916,
     "eval_samples": 169,
-    "eval_samples_per_second": 2.861,
-    "eval_steps_per_second": 0.386
+    "eval_samples_per_second": 2.881,
+    "eval_steps_per_second": 0.388
 }

train_results.json

@@ -2,8 +2,8 @@
     "epoch": 5.923076923076923,
     "total_flos": 1.7606154086724403e+17,
     "train_loss": 4.533969319902815e-05,
-    "train_runtime": 4.4642,
+    "train_runtime": 4.5265,
     "train_samples": 1346,
-    "train_samples_per_second": 896.022,
-    "train_steps_per_second": 224.006
+    "train_samples_per_second": 883.679,
+    "train_steps_per_second": 220.92
 }

trainer_state.json

@@ -613,9 +613,9 @@
       "step": 1001,
       "total_flos": 1.7606154086724403e+17,
       "train_loss": 4.533969319902815e-05,
-      "train_runtime": 4.4642,
-      "train_samples_per_second": 896.022,
-      "train_steps_per_second": 224.006
+      "train_runtime": 4.5265,
+      "train_samples_per_second": 883.679,
+      "train_steps_per_second": 220.92
     }
   ],
   "logging_steps": 25,