Training in progress, step 500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bd8a2f3d3b78a273316722aa1b2f3f93bfa9a4935b4aae13a5055dd3a5dea73
+oid sha256:71e0adabe57431758e6a472d31246e096a12256d0c48319058002b04b0696af1
 size 268290900

run-0/checkpoint-1000/trainer_state.json

@@ -10,50 +10,50 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6674193548387096,
-      "eval_loss": 0.416048139333725,
-      "eval_runtime": 1.3187,
-      "eval_samples_per_second": 2350.748,
-      "eval_steps_per_second": 49.29,
+      "eval_accuracy": 0.5841935483870968,
+      "eval_loss": 0.2228970229625702,
+      "eval_runtime": 1.3275,
+      "eval_samples_per_second": 2335.158,
+      "eval_steps_per_second": 48.963,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.9905994534492493,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.6477,
+      "grad_norm": 0.5151006579399109,
+      "learning_rate": 1.371069182389937e-05,
+      "loss": 0.3486,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8332258064516129,
-      "eval_loss": 0.15088918805122375,
-      "eval_runtime": 1.3272,
-      "eval_samples_per_second": 2335.773,
-      "eval_steps_per_second": 48.976,
+      "eval_accuracy": 0.8129032258064516,
+      "eval_loss": 0.11430680006742477,
+      "eval_runtime": 1.3269,
+      "eval_samples_per_second": 2336.263,
+      "eval_steps_per_second": 48.986,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.886774193548387,
-      "eval_loss": 0.07951628416776657,
-      "eval_runtime": 1.3591,
-      "eval_samples_per_second": 2280.965,
-      "eval_steps_per_second": 47.827,
+      "eval_accuracy": 0.864516129032258,
+      "eval_loss": 0.08068062365055084,
+      "eval_runtime": 1.339,
+      "eval_samples_per_second": 2315.116,
+      "eval_steps_per_second": 48.543,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.7344270348548889,
-      "learning_rate": 1.101527403414196e-05,
-      "loss": 0.1749,
+      "grad_norm": 0.4880180358886719,
+      "learning_rate": 7.421383647798742e-06,
+      "loss": 0.1317,
       "step": 1000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 1590,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -71,8 +71,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.5999229797501479,
-    "num_train_epochs": 7,
-    "temperature": 2
+    "alpha": 0.33163246742013985,
+    "num_train_epochs": 5,
+    "temperature": 7
   }
 }

run-0/checkpoint-1500/trainer_state.json

@@ -10,66 +10,66 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6674193548387096,
-      "eval_loss": 0.416048139333725,
-      "eval_runtime": 1.3187,
-      "eval_samples_per_second": 2350.748,
-      "eval_steps_per_second": 49.29,
+      "eval_accuracy": 0.5841935483870968,
+      "eval_loss": 0.2228970229625702,
+      "eval_runtime": 1.3275,
+      "eval_samples_per_second": 2335.158,
+      "eval_steps_per_second": 48.963,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.9905994534492493,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.6477,
+      "grad_norm": 0.5151006579399109,
+      "learning_rate": 1.371069182389937e-05,
+      "loss": 0.3486,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8332258064516129,
-      "eval_loss": 0.15088918805122375,
-      "eval_runtime": 1.3272,
-      "eval_samples_per_second": 2335.773,
-      "eval_steps_per_second": 48.976,
+      "eval_accuracy": 0.8129032258064516,
+      "eval_loss": 0.11430680006742477,
+      "eval_runtime": 1.3269,
+      "eval_samples_per_second": 2336.263,
+      "eval_steps_per_second": 48.986,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.886774193548387,
-      "eval_loss": 0.07951628416776657,
-      "eval_runtime": 1.3591,
-      "eval_samples_per_second": 2280.965,
-      "eval_steps_per_second": 47.827,
+      "eval_accuracy": 0.864516129032258,
+      "eval_loss": 0.08068062365055084,
+      "eval_runtime": 1.339,
+      "eval_samples_per_second": 2315.116,
+      "eval_steps_per_second": 48.543,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.7344270348548889,
-      "learning_rate": 1.101527403414196e-05,
-      "loss": 0.1749,
+      "grad_norm": 0.4880180358886719,
+      "learning_rate": 7.421383647798742e-06,
+      "loss": 0.1317,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9090322580645162,
-      "eval_loss": 0.05769029259681702,
-      "eval_runtime": 1.3435,
-      "eval_samples_per_second": 2307.35,
-      "eval_steps_per_second": 48.38,
+      "eval_accuracy": 0.8783870967741936,
+      "eval_loss": 0.06695606559515,
+      "eval_runtime": 1.3349,
+      "eval_samples_per_second": 2322.196,
+      "eval_steps_per_second": 48.691,
       "step": 1272
     },
     {
       "epoch": 4.716981132075472,
-      "grad_norm": 0.4701627790927887,
-      "learning_rate": 6.522911051212939e-06,
-      "loss": 0.0902,
+      "grad_norm": 0.3954662084579468,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.0931,
       "step": 1500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 1590,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -87,8 +87,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.5999229797501479,
-    "num_train_epochs": 7,
-    "temperature": 2
+    "alpha": 0.33163246742013985,
+    "num_train_epochs": 5,
+    "temperature": 7
   }
 }

run-0/checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b64fce98847974a1fbc15a5211620e911f29d0a6fa5bc6a5bf2e0acabc13c361
+oid sha256:71e0adabe57431758e6a472d31246e096a12256d0c48319058002b04b0696af1
 size 268290900

run-0/checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8b58f2178d777caeda6638dad5deb905c5d807ad067e89066f0357dbd988962e
+oid sha256:95b0a41e7fe17b4677a3a2a292701cf38e3b072bd5db1f6291353f28538029de
 size 536643898

run-0/checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:11e744277c61f87520794334442fae36c5f9ff6e10cb79d4bfee5176ca7eafe2
+oid sha256:9e1264523e958cf7990dc5f42d876cc12129475c4603804cf66868aaf25c2c24
 size 1064

run-0/checkpoint-500/trainer_state.json

@@ -10,25 +10,25 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6674193548387096,
-      "eval_loss": 0.416048139333725,
-      "eval_runtime": 1.3187,
-      "eval_samples_per_second": 2350.748,
-      "eval_steps_per_second": 49.29,
+      "eval_accuracy": 0.5841935483870968,
+      "eval_loss": 0.2228970229625702,
+      "eval_runtime": 1.3275,
+      "eval_samples_per_second": 2335.158,
+      "eval_steps_per_second": 48.963,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.9905994534492493,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.6477,
+      "grad_norm": 0.5151006579399109,
+      "learning_rate": 1.371069182389937e-05,
+      "loss": 0.3486,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 1590,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -46,8 +46,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.5999229797501479,
-    "num_train_epochs": 7,
-    "temperature": 2
+    "alpha": 0.33163246742013985,
+    "num_train_epochs": 5,
+    "temperature": 7
   }
 }

run-0/checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3995b387f2d0291be709cdbf61f96ec0665c9f1f556146e986cb9b22d69b84bd
+oid sha256:6765f5cb88961cbc2387298f0752026d7dd9093834e5e5d36d91152176870a41
 size 5176

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2a30697f0cf510c2a11fd1d808c468bcdeb16c48bed1099567859e4e853e3ad9
+oid sha256:6765f5cb88961cbc2387298f0752026d7dd9093834e5e5d36d91152176870a41
 size 5176