Training in progress, step 500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2cf50cbd134a3066e0d74687b39b2bba0fc2a0cdb34238e44082bdf70216c841
+oid sha256:746732a955ce5ac447aae4ae74e4f595331f971b2b8b78db1b28ff7cc415b7b9
 size 268290900

run-13/checkpoint-1000/trainer_state.json

@@ -10,50 +10,50 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.617741935483871,
-      "eval_loss": 0.2290053814649582,
-      "eval_runtime": 1.3588,
-      "eval_samples_per_second": 2281.395,
-      "eval_steps_per_second": 47.836,
+      "eval_accuracy": 0.6732258064516129,
+      "eval_loss": 0.40791815519332886,
+      "eval_runtime": 1.2355,
+      "eval_samples_per_second": 2509.076,
+      "eval_steps_per_second": 52.61,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.5753514766693115,
-      "learning_rate": 1.685534591194969e-05,
-      "loss": 0.3687,
+      "grad_norm": 1.0064228773117065,
+      "learning_rate": 1.650593990216632e-05,
+      "loss": 0.642,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8435483870967742,
-      "eval_loss": 0.1041906327009201,
-      "eval_runtime": 1.3701,
-      "eval_samples_per_second": 2262.659,
-      "eval_steps_per_second": 47.443,
+      "eval_accuracy": 0.8425806451612903,
+      "eval_loss": 0.14076808094978333,
+      "eval_runtime": 1.2403,
+      "eval_samples_per_second": 2499.46,
+      "eval_steps_per_second": 52.408,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.886774193548387,
-      "eval_loss": 0.06540963798761368,
-      "eval_runtime": 1.3808,
-      "eval_samples_per_second": 2245.156,
-      "eval_steps_per_second": 47.076,
+      "eval_accuracy": 0.8961290322580645,
+      "eval_loss": 0.07251902669668198,
+      "eval_runtime": 1.2468,
+      "eval_samples_per_second": 2486.288,
+      "eval_steps_per_second": 52.132,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.5283789038658142,
-      "learning_rate": 1.371069182389937e-05,
-      "loss": 0.1215,
+      "grad_norm": 0.7264623045921326,
+      "learning_rate": 1.3011879804332637e-05,
+      "loss": 0.1644,
       "step": 1000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3180,
+  "max_steps": 2862,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 9,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -67,12 +67,12 @@
       "attributes": {}
     }
   },
-  "total_flos": 272409458987400.0,
+  "total_flos": 260941334653608.0,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7808198306841543,
-    "num_train_epochs": 10,
-    "temperature": 5
+    "alpha": 0.9929851626814228,
+    "num_train_epochs": 9,
+    "temperature": 2
   }
 }

run-13/checkpoint-1500/trainer_state.json

@@ -10,66 +10,66 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.617741935483871,
-      "eval_loss": 0.2290053814649582,
-      "eval_runtime": 1.3588,
-      "eval_samples_per_second": 2281.395,
-      "eval_steps_per_second": 47.836,
+      "eval_accuracy": 0.6732258064516129,
+      "eval_loss": 0.40791815519332886,
+      "eval_runtime": 1.2355,
+      "eval_samples_per_second": 2509.076,
+      "eval_steps_per_second": 52.61,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.5753514766693115,
-      "learning_rate": 1.685534591194969e-05,
-      "loss": 0.3687,
+      "grad_norm": 1.0064228773117065,
+      "learning_rate": 1.650593990216632e-05,
+      "loss": 0.642,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8435483870967742,
-      "eval_loss": 0.1041906327009201,
-      "eval_runtime": 1.3701,
-      "eval_samples_per_second": 2262.659,
-      "eval_steps_per_second": 47.443,
+      "eval_accuracy": 0.8425806451612903,
+      "eval_loss": 0.14076808094978333,
+      "eval_runtime": 1.2403,
+      "eval_samples_per_second": 2499.46,
+      "eval_steps_per_second": 52.408,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.886774193548387,
-      "eval_loss": 0.06540963798761368,
-      "eval_runtime": 1.3808,
-      "eval_samples_per_second": 2245.156,
-      "eval_steps_per_second": 47.076,
+      "eval_accuracy": 0.8961290322580645,
+      "eval_loss": 0.07251902669668198,
+      "eval_runtime": 1.2468,
+      "eval_samples_per_second": 2486.288,
+      "eval_steps_per_second": 52.132,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.5283789038658142,
-      "learning_rate": 1.371069182389937e-05,
-      "loss": 0.1215,
+      "grad_norm": 0.7264623045921326,
+      "learning_rate": 1.3011879804332637e-05,
+      "loss": 0.1644,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8996774193548387,
-      "eval_loss": 0.04832444712519646,
-      "eval_runtime": 1.3844,
-      "eval_samples_per_second": 2239.279,
-      "eval_steps_per_second": 46.953,
+      "eval_accuracy": 0.9141935483870968,
+      "eval_loss": 0.05225619301199913,
+      "eval_runtime": 1.2515,
+      "eval_samples_per_second": 2477.02,
+      "eval_steps_per_second": 51.938,
       "step": 1272
     },
     {
       "epoch": 4.716981132075472,
-      "grad_norm": 0.3390381932258606,
-      "learning_rate": 1.0566037735849058e-05,
-      "loss": 0.0728,
+      "grad_norm": 0.4414461851119995,
+      "learning_rate": 9.517819706498952e-06,
+      "loss": 0.0818,
       "step": 1500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3180,
+  "max_steps": 2862,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 9,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -83,12 +83,12 @@
       "attributes": {}
     }
   },
-  "total_flos": 402837063777120.0,
+  "total_flos": 391368939443328.0,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7808198306841543,
-    "num_train_epochs": 10,
-    "temperature": 5
+    "alpha": 0.9929851626814228,
+    "num_train_epochs": 9,
+    "temperature": 2
   }
 }

run-13/checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d571c93aa3b50a8f6783e224958d2cdce5c10b8402e51b923b085d8491a353ff
+oid sha256:746732a955ce5ac447aae4ae74e4f595331f971b2b8b78db1b28ff7cc415b7b9
 size 268290900

run-13/checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afd0bdeb44b164495ecc3e458e8ebcca3b8c65d77783010337868d68625e5430
+oid sha256:9705b8a6db729e9a5e87252472c23151c8c318a4f1c6679911400c9b1a41b495
 size 536643898

run-13/checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:04366f62f8f88f5a8265df59adb051b320463277845db80e7fa43f13110c18c9
+oid sha256:98041bd7cae455426e290a1a0ee683bd5dd30893f7451fec3a464ae8995b17e4
 size 1064

run-13/checkpoint-500/trainer_state.json

@@ -10,25 +10,25 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.617741935483871,
-      "eval_loss": 0.2290053814649582,
-      "eval_runtime": 1.3588,
-      "eval_samples_per_second": 2281.395,
-      "eval_steps_per_second": 47.836,
+      "eval_accuracy": 0.6732258064516129,
+      "eval_loss": 0.40791815519332886,
+      "eval_runtime": 1.2355,
+      "eval_samples_per_second": 2509.076,
+      "eval_steps_per_second": 52.61,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.5753514766693115,
-      "learning_rate": 1.685534591194969e-05,
-      "loss": 0.3687,
+      "grad_norm": 1.0064228773117065,
+      "learning_rate": 1.650593990216632e-05,
+      "loss": 0.642,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3180,
+  "max_steps": 2862,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 9,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -42,12 +42,12 @@
       "attributes": {}
     }
   },
-  "total_flos": 141540333486132.0,
+  "total_flos": 130072209152340.0,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7808198306841543,
-    "num_train_epochs": 10,
-    "temperature": 5
+    "alpha": 0.9929851626814228,
+    "num_train_epochs": 9,
+    "temperature": 2
   }
 }

run-13/checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f62c7b1cdbc5f04f78a333fc7ad899d41a314edecf47ae9022088c79c529b2e
+oid sha256:825ce3dae72879e90f1e7c8bf9db5978e39cf9b4e9cf4d01acbdb15f1075ac03
 size 5176