Training in progress, step 3

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:560bb317099cbfe2d5a2fc87cdb4df54e4f37bc40064df9cdc34a07d0e3cc736
+oid sha256:9bbd2a48fe8221eda4ca0d47649d9c04798a5d29816af5091eb83e54d25699ce
 size 4965805240

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a367faf1e0584c346a822843f222f4188524466fd216ad4783987e3dfad1e4b2
+oid sha256:651c655175c1186eb3ba3b64d30ac535205aabd835d67d6752abf7aeb8374d1e
 size 1459729952

sft_sentimentbackdoors_sft_Llama-3.2-3B_experiment_3.3.txt

@@ -1,7 +1,7 @@
 Steps per epoch: 20.0, Eval steps: 1 Save steps: 3.0
 Creating repository sentimentbackdoors_sft_Llama-3.2-3B_experiment_3.3 under organization martian-mech-interp-grant...
 {'loss': 2.6332, 'grad_norm': 26.125, 'learning_rate': 1.0000000000000002e-06, 'epoch': 0.02}
-{'eval_loss': 2.6233537197113037, 'eval_runtime': 114.1124, 'eval_samples_per_second': 14.109, 'eval_steps_per_second': 1.77, 'epoch': 0.02}
+{'eval_loss': 2.6233537197113037, 'eval_runtime': 113.9386, 'eval_samples_per_second': 14.13, 'eval_steps_per_second': 1.773, 'epoch': 0.02}
 {'loss': 2.678, 'grad_norm': 27.0, 'learning_rate': 2.0000000000000003e-06, 'epoch': 0.03}
-{'eval_loss': 2.604088306427002, 'eval_runtime': 114.5221, 'eval_samples_per_second': 14.058, 'eval_steps_per_second': 1.764, 'epoch': 0.03}
-{'loss': 2.5736, 'grad_norm': 25.0, 'learning_rate': 3e-06, 'epoch': 0.05}
+{'eval_loss': 2.6063108444213867, 'eval_runtime': 114.1691, 'eval_samples_per_second': 14.102, 'eval_steps_per_second': 1.769, 'epoch': 0.03}
+{'loss': 2.5759, 'grad_norm': 25.125, 'learning_rate': 3e-06, 'epoch': 0.05}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd1356b158db215f1c041aef6ae43b2c74c5318b27778a9e94c11c6b027c35ba
+oid sha256:4a25a3659e15d8f4679cf94badf9f9eac33bd2510ae6bb015f34eeed4c71fdfd
 size 5560