Training in progress, epoch 20

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0f1c2e57ec1b2c872b589767af2b40a21c239a0b8fe0ea83c2fffcf18bf7e77
+oid sha256:dde74ab64751bd948f2089f033ab9ed4ec866afaeb97f83f3ec4d2380b5f58c2
 size 267832560

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

@@ -53,3 +53,5 @@ You should probably TRAIN this model on a down-stream task to be able to use it
 {'eval_loss': 0.49775975942611694, 'eval_accuracy': 0.921, 'eval_runtime': 37.3442, 'eval_samples_per_second': 26.778, 'eval_steps_per_second': 0.857, 'epoch': 16.0}
 {'eval_loss': 0.5782524347305298, 'eval_accuracy': 0.909, 'eval_runtime': 38.4136, 'eval_samples_per_second': 26.032, 'eval_steps_per_second': 0.833, 'epoch': 17.0}
 {'eval_loss': 0.5907241106033325, 'eval_accuracy': 0.901, 'eval_runtime': 37.9133, 'eval_samples_per_second': 26.376, 'eval_steps_per_second': 0.844, 'epoch': 18.0}
+{'eval_loss': 0.517770528793335, 'eval_accuracy': 0.917, 'eval_runtime': 37.8369, 'eval_samples_per_second': 26.429, 'eval_steps_per_second': 0.846, 'epoch': 19.0}
+{'loss': 0.0187, 'grad_norm': 0.01947682909667492, 'learning_rate': 6e-06, 'epoch': 20.0}

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

@@ -1 +1 @@
-{"eval/loss": 0.517770528793335, "eval/accuracy": 0.917, "eval/runtime": 37.8369, "eval/samples_per_second": 26.429, "eval/steps_per_second": 0.846, "train/epoch": 19.0, "train/global_step": 2375, "_timestamp": 1727252836.361977, "_runtime": 43245.48906803131, "_step": 30, "train/loss": 0.0268, "train/grad_norm": 20.58578109741211, "train/learning_rate": 6.800000000000001e-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": 0.5195603966712952, "eval/accuracy": 0.92, "eval/runtime": 37.9909, "eval/samples_per_second": 26.322, "eval/steps_per_second": 0.842, "train/epoch": 20.0, "train/global_step": 2500, "_timestamp": 1727254719.74196, "_runtime": 45128.869050979614, "_step": 32, "train/loss": 0.0187, "train/grad_norm": 0.01947682909667492, "train/learning_rate": 6e-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}

wandb/run-20240924_172630-x9iddikd/run-x9iddikd.wandb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:225a3466e1f5f63a52cc6e45441bc6533dc9c05caad936ff1a32ff85d9e80380
-size 1184325
+oid sha256:928144a1be5757f0385988ce82942ee4f2429f7e6752e9071b3b840cf07d41a9
+size 1237111