End of training

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 1.39,
+    "eval_accuracy": 0.838442284952852,
+    "eval_loss": 0.7085304260253906,
+    "eval_runtime": 9.8682,
+    "eval_samples_per_second": 506.676,
+    "eval_steps_per_second": 1.013,
+    "perplexity": 2.031004353305193,
+    "train_loss": 0.18125205249023438,
+    "train_runtime": 269324.8897,
+    "train_samples_per_second": 3802.099,
+    "train_steps_per_second": 3.713
+}

config.json

@@ -19,7 +19,7 @@
   "num_hidden_layers": 12,
   "pad_token_id": 0,
   "position_embedding_type": "absolute",
-  "torch_dtype": "float32",
+  "torch_dtype": "bfloat16",
   "transformers_version": "4.20.1",
   "type_vocab_size": 1,
   "use_cache": true,

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.39,
+    "eval_accuracy": 0.838442284952852,
+    "eval_loss": 0.7085304260253906,
+    "eval_runtime": 9.8682,
+    "eval_samples_per_second": 506.676,
+    "eval_steps_per_second": 1.013,
+    "perplexity": 2.031004353305193
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b4259dfdc9d95d40bfcdd98b7cbe401b1f4e0ac89f3aee63b1c92426a50e86b
-size 442678571
+oid sha256:50b540d5baa5e506eadd065b45c5f621480792a3a73b38c7b1de7abd0cdde37b
+size 221375275

train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.39,
+    "train_loss": 0.18125205249023438,
+    "train_runtime": 269324.8897,
+    "train_samples_per_second": 3802.099,
+    "train_steps_per_second": 3.713
+}