Training in progress, epoch 14

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dc3b9887fa00b716011a56980f1a4eeb8eb5eea9854a9a01a86292861d5e4df1
+oid sha256:a0505695cf303da48c7fdc01b3e844c17229610edb63ad1c8b203405e2fd9c2e
 size 267832560

wandb/run-20240924_172630-x9iddikd/files/output.log

@@ -46,3 +46,4 @@ You should probably TRAIN this model on a down-stream task to be able to use it
 {'eval_loss': 0.5506279468536377, 'eval_accuracy': 0.901, 'eval_runtime': 37.4907, 'eval_samples_per_second': 26.673, 'eval_steps_per_second': 0.854, 'epoch': 11.0}
 {'loss': 0.0442, 'grad_norm': 0.6364777684211731, 'learning_rate': 7.600000000000001e-06, 'epoch': 12.0}
 {'eval_loss': 0.578902006149292, 'eval_accuracy': 0.903, 'eval_runtime': 38.2969, 'eval_samples_per_second': 26.112, 'eval_steps_per_second': 0.836, 'epoch': 12.0}
+{'eval_loss': 0.47741687297821045, 'eval_accuracy': 0.92, 'eval_runtime': 37.7268, 'eval_samples_per_second': 26.506, 'eval_steps_per_second': 0.848, 'epoch': 13.0}

wandb/run-20240924_172630-x9iddikd/files/wandb-summary.json

@@ -1 +1 @@
-{"eval/loss": 0.47741687297821045, "eval/accuracy": 0.92, "eval/runtime": 37.7268, "eval/samples_per_second": 26.506, "eval/steps_per_second": 0.848, "train/epoch": 13.0, "train/global_step": 1625, "_timestamp": 1727242574.522822, "_runtime": 32983.64991283417, "_step": 23, "train/loss": 0.0442, "train/grad_norm": 0.6364777684211731, "train/learning_rate": 7.600000000000001e-06, "train_runtime": 8026.8642, "train_samples_per_second": 2.492, "train_steps_per_second": 0.156, "total_flos": 2396475988298112.0, "train_loss": 0.11480112991333008}
+{"eval/loss": 0.5484298467636108, "eval/accuracy": 0.894, "eval/runtime": 38.013, "eval/samples_per_second": 26.307, "eval/steps_per_second": 0.842, "train/epoch": 14.0, "train/global_step": 1750, "_timestamp": 1727244298.3548748, "_runtime": 34707.48196578026, "_step": 24, "train/loss": 0.0442, "train/grad_norm": 0.6364777684211731, "train/learning_rate": 7.600000000000001e-06, "train_runtime": 8026.8642, "train_samples_per_second": 2.492, "train_steps_per_second": 0.156, "total_flos": 2396475988298112.0, "train_loss": 0.11480112991333008}