Training completed

README.md

@@ -38,10 +38,10 @@ The following hyperparameters were used during training:
 - eval_batch_size: 8
 - seed: 42
 - distributed_type: multi-GPU
-- num_devices: 2
+- num_devices: 4
 - gradient_accumulation_steps: 10
-- total_train_batch_size: 40
-- total_eval_batch_size: 16
+- total_train_batch_size: 80
+- total_eval_batch_size: 32
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.05

all_results.json

@@ -1,11 +1,11 @@
 {
-    "avg_step_time": 6.689460206031799,
-    "epoch": 0.01,
-    "total_flos": 4810910442979328.0,
-    "total_training_time": 79.5918402671814,
-    "total_training_time_mins": 1.32653067111969,
-    "train_loss": 3.181192398071289,
-    "train_runtime": 72.5058,
-    "train_samples_per_second": 5.517,
-    "train_steps_per_second": 0.138
+    "avg_step_time": 6.874203062057495,
+    "epoch": 0.02,
+    "total_flos": 9621820885958656.0,
+    "total_training_time": 90.24143290519714,
+    "total_training_time_mins": 1.5040238817532858,
+    "train_loss": 3.0087697982788084,
+    "train_runtime": 75.0646,
+    "train_samples_per_second": 10.657,
+    "train_steps_per_second": 0.133
 }

train_results.json

@@ -1,11 +1,11 @@
 {
-    "avg_step_time": 6.689460206031799,
-    "epoch": 0.01,
-    "total_flos": 4810910442979328.0,
-    "total_training_time": 79.5918402671814,
-    "total_training_time_mins": 1.32653067111969,
-    "train_loss": 3.181192398071289,
-    "train_runtime": 72.5058,
-    "train_samples_per_second": 5.517,
-    "train_steps_per_second": 0.138
+    "avg_step_time": 6.874203062057495,
+    "epoch": 0.02,
+    "total_flos": 9621820885958656.0,
+    "total_training_time": 90.24143290519714,
+    "total_training_time_mins": 1.5040238817532858,
+    "train_loss": 3.0087697982788084,
+    "train_runtime": 75.0646,
+    "train_samples_per_second": 10.657,
+    "train_steps_per_second": 0.133
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.01,
+  "epoch": 0.02,
   "eval_steps": 0,
   "global_step": 10,
   "is_hyper_param_search": false,
@@ -10,20 +10,20 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "grad_norm": 0.5649898052215576,
+      "epoch": 0.02,
+      "grad_norm": 0.561686098575592,
       "learning_rate": 0.0001,
-      "loss": 3.1812,
+      "loss": 3.0088,
       "step": 10
     },
     {
-      "epoch": 0.01,
+      "epoch": 0.02,
       "step": 10,
-      "total_flos": 4810910442979328.0,
-      "train_loss": 3.181192398071289,
-      "train_runtime": 72.5058,
-      "train_samples_per_second": 5.517,
-      "train_steps_per_second": 0.138
+      "total_flos": 9621820885958656.0,
+      "train_loss": 3.0087697982788084,
+      "train_runtime": 75.0646,
+      "train_samples_per_second": 10.657,
+      "train_steps_per_second": 0.133
     }
   ],
   "logging_steps": 10,
@@ -43,7 +43,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4810910442979328.0,
+  "total_flos": 9621820885958656.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null