Training in progress, step 2500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:762b4c02c12b7371af1d828c73566891953cef909ad141bb175ec5901f30c164
+oid sha256:0737b1fa29f840b7030bb5c9b520c81b6908fc48c8a845707681321814b8ff0d
 size 268290900

run-1/checkpoint-2500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47fa5bdd9ebe645b2aa80700c20cbf7176f0e6e274d2b1ea405cf63ea345869d
+oid sha256:0737b1fa29f840b7030bb5c9b520c81b6908fc48c8a845707681321814b8ff0d
 size 268290900

run-1/checkpoint-2500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ebd8c044d2b630d9cc78dcf6c7859ac2646d7dda0503042cc52a692dee954233
+oid sha256:022ba77cf68a848d7df670b4a209d40dfa79dadc3ef5c4d6335b6598b13af6d7
 size 536643898

run-1/checkpoint-2500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b997ea9328158f28ff50e49c476931e703b821360bc8325e0c4d100e032c865
+oid sha256:d116f7be47ce88249a782ce7aafaf5b673a05bfa0ebdf9f14ab4c327c94c6b04
 size 1064

run-1/checkpoint-2500/trainer_state.json

@@ -10,107 +10,107 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5829032258064516,
-      "eval_loss": 0.1985393762588501,
-      "eval_runtime": 5.6203,
-      "eval_samples_per_second": 551.572,
-      "eval_steps_per_second": 11.565,
+      "eval_accuracy": 0.6529032258064517,
+      "eval_loss": 0.28555235266685486,
+      "eval_runtime": 5.4121,
+      "eval_samples_per_second": 572.786,
+      "eval_steps_per_second": 12.01,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.5288612246513367,
-      "learning_rate": 1.606918238993711e-05,
-      "loss": 0.3184,
+      "grad_norm": 0.6621501445770264,
+      "learning_rate": 1.685534591194969e-05,
+      "loss": 0.4579,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8174193548387096,
-      "eval_loss": 0.09726663678884506,
-      "eval_runtime": 6.0438,
-      "eval_samples_per_second": 512.92,
-      "eval_steps_per_second": 10.755,
+      "eval_accuracy": 0.8451612903225807,
+      "eval_loss": 0.1139635443687439,
+      "eval_runtime": 5.9245,
+      "eval_samples_per_second": 523.247,
+      "eval_steps_per_second": 10.971,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8764516129032258,
-      "eval_loss": 0.06577406078577042,
-      "eval_runtime": 5.6348,
-      "eval_samples_per_second": 550.155,
-      "eval_steps_per_second": 11.535,
+      "eval_accuracy": 0.8961290322580645,
+      "eval_loss": 0.06579811871051788,
+      "eval_runtime": 5.4507,
+      "eval_samples_per_second": 568.736,
+      "eval_steps_per_second": 11.925,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.44873112440109253,
-      "learning_rate": 1.2138364779874214e-05,
-      "loss": 0.1127,
+      "grad_norm": 0.5677153468132019,
+      "learning_rate": 1.371069182389937e-05,
+      "loss": 0.1346,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8964516129032258,
-      "eval_loss": 0.05149334296584129,
-      "eval_runtime": 5.7578,
-      "eval_samples_per_second": 538.401,
-      "eval_steps_per_second": 11.289,
+      "eval_accuracy": 0.9138709677419354,
+      "eval_loss": 0.04725706949830055,
+      "eval_runtime": 5.5473,
+      "eval_samples_per_second": 558.83,
+      "eval_steps_per_second": 11.717,
       "step": 1272
     },
     {
       "epoch": 4.716981132075472,
-      "grad_norm": 0.30973368883132935,
-      "learning_rate": 8.207547169811321e-06,
-      "loss": 0.0733,
+      "grad_norm": 0.3639396131038666,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.0751,
       "step": 1500
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9051612903225806,
-      "eval_loss": 0.04290274158120155,
-      "eval_runtime": 5.9031,
-      "eval_samples_per_second": 525.15,
-      "eval_steps_per_second": 11.011,
+      "eval_accuracy": 0.92,
+      "eval_loss": 0.03804958611726761,
+      "eval_runtime": 5.7319,
+      "eval_samples_per_second": 540.831,
+      "eval_steps_per_second": 11.34,
       "step": 1590
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9093548387096774,
-      "eval_loss": 0.038180265575647354,
-      "eval_runtime": 5.6503,
-      "eval_samples_per_second": 548.642,
-      "eval_steps_per_second": 11.504,
+      "eval_accuracy": 0.927741935483871,
+      "eval_loss": 0.03348824381828308,
+      "eval_runtime": 5.4253,
+      "eval_samples_per_second": 571.402,
+      "eval_steps_per_second": 11.981,
       "step": 1908
     },
     {
       "epoch": 6.289308176100629,
-      "grad_norm": 0.27582159638404846,
-      "learning_rate": 4.276729559748428e-06,
-      "loss": 0.0587,
+      "grad_norm": 0.29170453548431396,
+      "learning_rate": 7.421383647798742e-06,
+      "loss": 0.0566,
       "step": 2000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9141935483870968,
-      "eval_loss": 0.03587044030427933,
-      "eval_runtime": 5.962,
-      "eval_samples_per_second": 519.958,
-      "eval_steps_per_second": 10.902,
+      "eval_accuracy": 0.932258064516129,
+      "eval_loss": 0.0307790357619524,
+      "eval_runtime": 5.6345,
+      "eval_samples_per_second": 550.182,
+      "eval_steps_per_second": 11.536,
       "step": 2226
     },
     {
       "epoch": 7.861635220125786,
-      "grad_norm": 0.275245726108551,
-      "learning_rate": 3.459119496855346e-07,
-      "loss": 0.0528,
+      "grad_norm": 0.32524076104164124,
+      "learning_rate": 4.276729559748428e-06,
+      "loss": 0.0488,
       "step": 2500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2544,
+  "max_steps": 3180,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 8,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -128,8 +128,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.819089504077056,
-    "num_train_epochs": 8,
-    "temperature": 12
+    "alpha": 0.31688267813599835,
+    "num_train_epochs": 10,
+    "temperature": 3
   }
 }

run-1/checkpoint-2500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ad3761ee6c049b5f54cddb21d4403152bfaafdbfc2a13b71993419e99660838
+oid sha256:6a074013ebca87018fe538250c2f45bfdeb260149a400a0644508c302966c2ef
 size 5240

runs/Oct12_06-40-39_b76c1be2ae55/events.out.tfevents.1728716534.b76c1be2ae55.1423.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:510b81f142a0bc84f1443c0c290b0cdf5530313ff3ca8445bb1a0e8702c66919
-size 15547
+oid sha256:5a105d06a7edeaac035833951b00d7fe58e68fa10c1b0676f1153501aea079fd
+size 16081