Training in progress, step 500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1df6a2e095aa6bbb5d360e7f012cbe4a7929871fd4d4183142dd1f5f93bff5b9
+oid sha256:ab42167ffff4c1e729d547711e458440cd560d09dd08d069574cd46a86c79e0d
 size 268290900

run-1/checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29e2ad483741af1463bbf6cb132bb4200c2aba9a23de248941c7f833939d8507
+oid sha256:992ce6473ead04bcdb67b1d09a58b11791539e3b4a26af0ae5fe8e81758fae52
 size 268290900

run-1/checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:512630753c954c85018b4e199d7a14d77f624b3e2b274e28997f39f4ccb13783
+oid sha256:7c26be44cadfbd774f4b45a36f0fe2da2d795da68a912efacb62db2a9ba05e13
 size 536643898

run-1/checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:60ef01273aaa599804d51a93c6c0c61874ebd50ab9e135a21aab5f7cd0e6487a
+oid sha256:72a4307cca951ef6d42c7f184774e8b1259c78b2bf130b8443ec0b62f6dbf689
 size 1064

run-1/checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:720882e08adee18bc186850828f02514ae1f0c4061921a151ad1b71291c1e3c8
+oid sha256:48a44b718cd5e63b306a632a23ed7f69877efbab8daac27533495007e71fadba
 size 5176

run-1/checkpoint-1500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e55ede6b5d9f3c62cff405499d587825e3459a3973ecd0c37ddb98625d0e3544
+oid sha256:4feb8140b2cb3284d4e3f6147940bbafe0b4183f0c7f4dc8d6e4e99abd845a27
 size 268290900

run-1/checkpoint-1500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:37c6e4e87a5fa18f02ae927d32d069ae525fdaa924b0006ee3d032ee4e6f11ed
+oid sha256:70e3cd8ff7c183a57d4d03e47795322ad975e28f7c51c4c8fd8123b5d2c74aa7
 size 536643898

run-1/checkpoint-1500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55c8d3ce0734337fc0c187ca5543b4c70ca45d996531f199209b3a0c2a798109
+oid sha256:693ec081eb0fa0af7d13e289edb194450dd19fdada23346e2af4292ea228535f
 size 1064

run-1/checkpoint-1500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:720882e08adee18bc186850828f02514ae1f0c4061921a151ad1b71291c1e3c8
+oid sha256:48a44b718cd5e63b306a632a23ed7f69877efbab8daac27533495007e71fadba
 size 5176

run-1/checkpoint-2000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:042c62092408efa9fec63d858bdb2942edf29136586e4197e800b1952421e8d8
+oid sha256:e7c87ecd2ce8a9c08ce1eec28d71fdee2826fa3c4ac9e847a6324f0dae44b424
 size 268290900

run-1/checkpoint-2000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1db3dd137f742f587491edfab8ea4c178010b5522a9445e55791e83681090454
+oid sha256:8779bb51dcb01a55f1e1edafaf0f86a31915b2b7fa47b29976fb522ed71b268c
 size 536643898

run-1/checkpoint-2000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5613af56824b89d2251abc079d69f51d9497a128212db6eb459f4c957e78a29b
+oid sha256:37ccc14cc902672c3d8e770c2a2ce3a532b454981f3118cb18a09274102a99fc
 size 14244

run-1/checkpoint-2000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43d45918bfeb622ba899798234ec7646c7a90dd3f5771db086f5b7dee1d5a530
+oid sha256:67d1a8a9215f18c672cf798b531337b9e0636922d8c4373e76080f8fec19ceb2
 size 1064

run-1/checkpoint-2000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9edb169c95045dfb3d8473928c594ea0196c02cbd8544dc5bc903ed7216c71b
-size 4728
+oid sha256:48a44b718cd5e63b306a632a23ed7f69877efbab8daac27533495007e71fadba
+size 5176

run-1/checkpoint-2500/config.json

@@ -326,6 +326,6 @@
   "sinusoidal_pos_embds": false,
   "tie_weights_": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.37.2",
+  "transformers_version": "4.41.1",
   "vocab_size": 30522
 }

run-1/checkpoint-2500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:31b7cc37f0e1279c5a305875fe772217ce53fd999cbd2e73cd6ab6f8e319c839
+oid sha256:f23efffc15ca45d7fa0316048b1dfcc06db888f26436cdb84d4c073b721d67ce
 size 268290900

run-1/checkpoint-2500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:581d4e760b6be92573a6fde20cad95918071694d5ed70a688ce4347862411de0
+oid sha256:4e3ce9861afe4daf681e45155b67576707639a2887d5bcd8db11491be4afc050
 size 536643898

run-1/checkpoint-2500/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:983ad82afbb61099978c5b5a095761d0cc8a5b9889d078c9a1fcb9751b84b6be
+oid sha256:2fe590653dc8c7914203f68b195f5118633e21a9fa30a5fcd6f1518bae980ac9
 size 14244

run-1/checkpoint-2500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d116f7be47ce88249a782ce7aafaf5b673a05bfa0ebdf9f14ab4c327c94c6b04
+oid sha256:b7c943a77651a184c6653dd4706c14c9a27d52136e81942a2b361c74acb9f665
 size 1064

run-1/checkpoint-2500/tokenizer.json

@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 512,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

run-1/checkpoint-2500/trainer_state.json

@@ -10,109 +10,126 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6412903225806451,
-      "eval_loss": 0.230123370885849,
-      "eval_runtime": 1.3595,
-      "eval_samples_per_second": 2280.291,
-      "eval_steps_per_second": 47.813,
+      "eval_accuracy": 0.5951612903225807,
+      "eval_loss": 0.20783506333827972,
+      "eval_runtime": 1.3226,
+      "eval_samples_per_second": 2343.937,
+      "eval_steps_per_second": 49.147,
       "step": 318
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 1.685534591194969e-05,
-      "loss": 0.3693,
+      "epoch": 1.5723270440251573,
+      "grad_norm": 0.5312509536743164,
+      "learning_rate": 1.650593990216632e-05,
+      "loss": 0.3341,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8358064516129032,
-      "eval_loss": 0.10448037087917328,
-      "eval_runtime": 1.3653,
-      "eval_samples_per_second": 2270.569,
-      "eval_steps_per_second": 47.609,
+      "eval_accuracy": 0.8338709677419355,
+      "eval_loss": 0.09971419721841812,
+      "eval_runtime": 1.3325,
+      "eval_samples_per_second": 2326.534,
+      "eval_steps_per_second": 48.782,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8851612903225806,
-      "eval_loss": 0.06413500756025314,
-      "eval_runtime": 1.3667,
-      "eval_samples_per_second": 2268.292,
-      "eval_steps_per_second": 47.561,
+      "eval_accuracy": 0.882258064516129,
+      "eval_loss": 0.06562549620866776,
+      "eval_runtime": 1.3417,
+      "eval_samples_per_second": 2310.56,
+      "eval_steps_per_second": 48.447,
       "step": 954
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.371069182389937e-05,
-      "loss": 0.1204,
+      "epoch": 3.1446540880503147,
+      "grad_norm": 0.4943838119506836,
+      "learning_rate": 1.3011879804332637e-05,
+      "loss": 0.1156,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9041935483870968,
-      "eval_loss": 0.04693836718797684,
-      "eval_runtime": 1.3284,
-      "eval_samples_per_second": 2333.641,
-      "eval_steps_per_second": 48.931,
+      "eval_accuracy": 0.8958064516129032,
+      "eval_loss": 0.04985633119940758,
+      "eval_runtime": 1.3449,
+      "eval_samples_per_second": 2305.032,
+      "eval_steps_per_second": 48.331,
       "step": 1272
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 1.0566037735849058e-05,
-      "loss": 0.0726,
+      "epoch": 4.716981132075472,
+      "grad_norm": 0.3221791386604309,
+      "learning_rate": 9.517819706498952e-06,
+      "loss": 0.0725,
       "step": 1500
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9180645161290323,
-      "eval_loss": 0.03833732381463051,
-      "eval_runtime": 1.2898,
-      "eval_samples_per_second": 2403.423,
-      "eval_steps_per_second": 50.394,
+      "eval_accuracy": 0.9096774193548387,
+      "eval_loss": 0.04061922803521156,
+      "eval_runtime": 1.3564,
+      "eval_samples_per_second": 2285.502,
+      "eval_steps_per_second": 47.922,
       "step": 1590
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9235483870967742,
-      "eval_loss": 0.03355677053332329,
-      "eval_runtime": 1.2765,
-      "eval_samples_per_second": 2428.541,
-      "eval_steps_per_second": 50.921,
+      "eval_accuracy": 0.9187096774193548,
+      "eval_loss": 0.03560462221503258,
+      "eval_runtime": 1.3515,
+      "eval_samples_per_second": 2293.716,
+      "eval_steps_per_second": 48.094,
       "step": 1908
     },
     {
-      "epoch": 6.29,
-      "learning_rate": 7.421383647798742e-06,
-      "loss": 0.0554,
+      "epoch": 6.289308176100629,
+      "grad_norm": 0.2550632953643799,
+      "learning_rate": 6.02375960866527e-06,
+      "loss": 0.0569,
       "step": 2000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9274193548387096,
-      "eval_loss": 0.03028225339949131,
-      "eval_runtime": 1.2767,
-      "eval_samples_per_second": 2428.229,
-      "eval_steps_per_second": 50.914,
+      "eval_accuracy": 0.92,
+      "eval_loss": 0.03291216120123863,
+      "eval_runtime": 1.3475,
+      "eval_samples_per_second": 2300.507,
+      "eval_steps_per_second": 48.236,
       "step": 2226
     },
     {
-      "epoch": 7.86,
-      "learning_rate": 4.276729559748428e-06,
-      "loss": 0.0478,
+      "epoch": 7.861635220125786,
+      "grad_norm": 0.2292759269475937,
+      "learning_rate": 2.5296995108315863e-06,
+      "loss": 0.0497,
       "step": 2500
     }
   ],
   "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,
-  "total_flos": 649131947564688.0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 651155886807636.0,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7207200745295966,
-    "num_train_epochs": 10,
-    "temperature": 5
+    "alpha": 0.7941624119034844,
+    "num_train_epochs": 9,
+    "temperature": 8
   }
 }

run-1/checkpoint-2500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9edb169c95045dfb3d8473928c594ea0196c02cbd8544dc5bc903ed7216c71b
-size 4728
+oid sha256:48a44b718cd5e63b306a632a23ed7f69877efbab8daac27533495007e71fadba
+size 5176

run-2/checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3cf2f2d4b325795a9e58627b62a7d582d3493767bd8e2c5f4814cf61afa845c9
+oid sha256:344ef733c9a1fb806a98643025fe3967590c0bf1a89d6611b34ad94270482afc
 size 268290900

run-2/checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:311e9ca7c3fa3d5a2f2bd13162fd5bc4217068d013a7555cabf6f05b33ec0f34
+oid sha256:b7bf23a8aaf272e7bc28a3470e51498a16fb9eaae511eafc0f482a281ebba9a4
 size 536643898

run-2/checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed9da1cbe846856af516cf037dd225e7f48e2737ef84c42a9bee0753b8c140ef
+oid sha256:045cf4abe065257226fe11ac1a91dfd4c5b953cb2366365ef8bb18b4cf33e734
 size 1064

run-2/checkpoint-1000/trainer_state.json

@@ -10,50 +10,50 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5741935483870968,
-      "eval_loss": 0.19538737833499908,
-      "eval_runtime": 1.3255,
-      "eval_samples_per_second": 2338.712,
-      "eval_steps_per_second": 49.038,
+      "eval_accuracy": 0.5864516129032258,
+      "eval_loss": 0.2008386254310608,
+      "eval_runtime": 1.3601,
+      "eval_samples_per_second": 2279.22,
+      "eval_steps_per_second": 47.79,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.49559739232063293,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.3117,
+      "grad_norm": 0.5175672173500061,
+      "learning_rate": 1.606918238993711e-05,
+      "loss": 0.3218,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8187096774193549,
-      "eval_loss": 0.0987778976559639,
-      "eval_runtime": 1.3373,
-      "eval_samples_per_second": 2318.165,
-      "eval_steps_per_second": 48.607,
+      "eval_accuracy": 0.8258064516129032,
+      "eval_loss": 0.09898000210523605,
+      "eval_runtime": 1.3658,
+      "eval_samples_per_second": 2269.659,
+      "eval_steps_per_second": 47.59,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8725806451612903,
-      "eval_loss": 0.06809484958648682,
-      "eval_runtime": 1.3423,
-      "eval_samples_per_second": 2309.386,
-      "eval_steps_per_second": 48.423,
+      "eval_accuracy": 0.8777419354838709,
+      "eval_loss": 0.06669934839010239,
+      "eval_runtime": 1.2565,
+      "eval_samples_per_second": 2467.123,
+      "eval_steps_per_second": 51.73,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.46493077278137207,
-      "learning_rate": 1.101527403414196e-05,
-      "loss": 0.1137,
+      "grad_norm": 0.47839590907096863,
+      "learning_rate": 1.2138364779874214e-05,
+      "loss": 0.1144,
       "step": 1000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 2544,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 8,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -71,8 +71,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.2890625148941124,
-    "num_train_epochs": 7,
-    "temperature": 17
+    "alpha": 0.7723229117593493,
+    "num_train_epochs": 8,
+    "temperature": 11
   }
 }

run-2/checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02895758c800be06b344602e8d1e4451336b524bdce9f82588ca589e12cfef2e
+oid sha256:baea954a8311f77b8b2d4c67cfd083cf6d7e3bc8a41fb699bd5b137d572b44d3
 size 5176

run-2/checkpoint-1500/trainer_state.json

@@ -10,66 +10,66 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5741935483870968,
-      "eval_loss": 0.19538737833499908,
-      "eval_runtime": 1.3255,
-      "eval_samples_per_second": 2338.712,
-      "eval_steps_per_second": 49.038,
+      "eval_accuracy": 0.5864516129032258,
+      "eval_loss": 0.2008386254310608,
+      "eval_runtime": 1.3601,
+      "eval_samples_per_second": 2279.22,
+      "eval_steps_per_second": 47.79,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.49559739232063293,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.3117,
+      "grad_norm": 0.5175672173500061,
+      "learning_rate": 1.606918238993711e-05,
+      "loss": 0.3218,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8187096774193549,
-      "eval_loss": 0.0987778976559639,
-      "eval_runtime": 1.3373,
-      "eval_samples_per_second": 2318.165,
-      "eval_steps_per_second": 48.607,
+      "eval_accuracy": 0.8258064516129032,
+      "eval_loss": 0.09898000210523605,
+      "eval_runtime": 1.3658,
+      "eval_samples_per_second": 2269.659,
+      "eval_steps_per_second": 47.59,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8725806451612903,
-      "eval_loss": 0.06809484958648682,
-      "eval_runtime": 1.3423,
-      "eval_samples_per_second": 2309.386,
-      "eval_steps_per_second": 48.423,
+      "eval_accuracy": 0.8777419354838709,
+      "eval_loss": 0.06669934839010239,
+      "eval_runtime": 1.2565,
+      "eval_samples_per_second": 2467.123,
+      "eval_steps_per_second": 51.73,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.46493077278137207,
-      "learning_rate": 1.101527403414196e-05,
-      "loss": 0.1137,
+      "grad_norm": 0.47839590907096863,
+      "learning_rate": 1.2138364779874214e-05,
+      "loss": 0.1144,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8896774193548387,
-      "eval_loss": 0.053910914808511734,
-      "eval_runtime": 1.3423,
-      "eval_samples_per_second": 2309.462,
-      "eval_steps_per_second": 48.424,
+      "eval_accuracy": 0.8932258064516129,
+      "eval_loss": 0.051668863743543625,
+      "eval_runtime": 1.2584,
+      "eval_samples_per_second": 2463.451,
+      "eval_steps_per_second": 51.653,
       "step": 1272
     },
     {
       "epoch": 4.716981132075472,
-      "grad_norm": 0.32403233647346497,
-      "learning_rate": 6.522911051212939e-06,
-      "loss": 0.0755,
+      "grad_norm": 0.32125934958457947,
+      "learning_rate": 8.207547169811321e-06,
+      "loss": 0.0737,
       "step": 1500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 2544,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 8,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -87,8 +87,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.2890625148941124,
-    "num_train_epochs": 7,
-    "temperature": 17
+    "alpha": 0.7723229117593493,
+    "num_train_epochs": 8,
+    "temperature": 11
   }
 }

run-2/checkpoint-2000/trainer_state.json

@@ -10,91 +10,91 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5741935483870968,
-      "eval_loss": 0.19538737833499908,
-      "eval_runtime": 1.3255,
-      "eval_samples_per_second": 2338.712,
-      "eval_steps_per_second": 49.038,
+      "eval_accuracy": 0.5864516129032258,
+      "eval_loss": 0.2008386254310608,
+      "eval_runtime": 1.3601,
+      "eval_samples_per_second": 2279.22,
+      "eval_steps_per_second": 47.79,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.49559739232063293,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.3117,
+      "grad_norm": 0.5175672173500061,
+      "learning_rate": 1.606918238993711e-05,
+      "loss": 0.3218,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8187096774193549,
-      "eval_loss": 0.0987778976559639,
-      "eval_runtime": 1.3373,
-      "eval_samples_per_second": 2318.165,
-      "eval_steps_per_second": 48.607,
+      "eval_accuracy": 0.8258064516129032,
+      "eval_loss": 0.09898000210523605,
+      "eval_runtime": 1.3658,
+      "eval_samples_per_second": 2269.659,
+      "eval_steps_per_second": 47.59,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8725806451612903,
-      "eval_loss": 0.06809484958648682,
-      "eval_runtime": 1.3423,
-      "eval_samples_per_second": 2309.386,
-      "eval_steps_per_second": 48.423,
+      "eval_accuracy": 0.8777419354838709,
+      "eval_loss": 0.06669934839010239,
+      "eval_runtime": 1.2565,
+      "eval_samples_per_second": 2467.123,
+      "eval_steps_per_second": 51.73,
       "step": 954
     },
     {
       "epoch": 3.1446540880503147,
-      "grad_norm": 0.46493077278137207,
-      "learning_rate": 1.101527403414196e-05,
-      "loss": 0.1137,
+      "grad_norm": 0.47839590907096863,
+      "learning_rate": 1.2138364779874214e-05,
+      "loss": 0.1144,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8896774193548387,
-      "eval_loss": 0.053910914808511734,
-      "eval_runtime": 1.3423,
-      "eval_samples_per_second": 2309.462,
-      "eval_steps_per_second": 48.424,
+      "eval_accuracy": 0.8932258064516129,
+      "eval_loss": 0.051668863743543625,
+      "eval_runtime": 1.2584,
+      "eval_samples_per_second": 2463.451,
+      "eval_steps_per_second": 51.653,
       "step": 1272
     },
     {
       "epoch": 4.716981132075472,
-      "grad_norm": 0.32403233647346497,
-      "learning_rate": 6.522911051212939e-06,
-      "loss": 0.0755,
+      "grad_norm": 0.32125934958457947,
+      "learning_rate": 8.207547169811321e-06,
+      "loss": 0.0737,
       "step": 1500
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8990322580645161,
-      "eval_loss": 0.045997604727745056,
-      "eval_runtime": 1.3489,
-      "eval_samples_per_second": 2298.125,
-      "eval_steps_per_second": 48.186,
+      "eval_accuracy": 0.9048387096774193,
+      "eval_loss": 0.04295943304896355,
+      "eval_runtime": 1.2626,
+      "eval_samples_per_second": 2455.311,
+      "eval_steps_per_second": 51.482,
       "step": 1590
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9077419354838709,
-      "eval_loss": 0.042082779109478,
-      "eval_runtime": 1.3634,
-      "eval_samples_per_second": 2273.658,
-      "eval_steps_per_second": 47.673,
+      "eval_accuracy": 0.9125806451612903,
+      "eval_loss": 0.03821730986237526,
+      "eval_runtime": 1.3321,
+      "eval_samples_per_second": 2327.068,
+      "eval_steps_per_second": 48.793,
       "step": 1908
     },
     {
       "epoch": 6.289308176100629,
-      "grad_norm": 0.2889467477798462,
-      "learning_rate": 2.0305480682839176e-06,
-      "loss": 0.0626,
+      "grad_norm": 0.2682598829269409,
+      "learning_rate": 4.276729559748428e-06,
+      "loss": 0.0592,
       "step": 2000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 2544,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 8,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -112,8 +112,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.2890625148941124,
-    "num_train_epochs": 7,
-    "temperature": 17
+    "alpha": 0.7723229117593493,
+    "num_train_epochs": 8,
+    "temperature": 11
   }
 }

run-2/checkpoint-2500/config.json

@@ -326,6 +326,6 @@
   "sinusoidal_pos_embds": false,
   "tie_weights_": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.37.2",
+  "transformers_version": "4.41.1",
   "vocab_size": 30522
 }

run-2/checkpoint-2500/tokenizer.json

@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 512,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

run-2/checkpoint-2500/trainer_state.json

@@ -10,109 +10,126 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6041935483870968,
-      "eval_loss": 0.19521339237689972,
-      "eval_runtime": 1.4292,
-      "eval_samples_per_second": 2168.979,
-      "eval_steps_per_second": 45.479,
+      "eval_accuracy": 0.5864516129032258,
+      "eval_loss": 0.2008386254310608,
+      "eval_runtime": 1.3601,
+      "eval_samples_per_second": 2279.22,
+      "eval_steps_per_second": 47.79,
       "step": 318
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 1.650593990216632e-05,
-      "loss": 0.313,
+      "epoch": 1.5723270440251573,
+      "grad_norm": 0.5175672173500061,
+      "learning_rate": 1.606918238993711e-05,
+      "loss": 0.3218,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8187096774193549,
-      "eval_loss": 0.09592956304550171,
-      "eval_runtime": 1.3897,
-      "eval_samples_per_second": 2230.662,
-      "eval_steps_per_second": 46.772,
+      "eval_accuracy": 0.8258064516129032,
+      "eval_loss": 0.09898000210523605,
+      "eval_runtime": 1.3658,
+      "eval_samples_per_second": 2269.659,
+      "eval_steps_per_second": 47.59,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8793548387096775,
-      "eval_loss": 0.06310795992612839,
-      "eval_runtime": 1.4045,
-      "eval_samples_per_second": 2207.192,
-      "eval_steps_per_second": 46.28,
+      "eval_accuracy": 0.8777419354838709,
+      "eval_loss": 0.06669934839010239,
+      "eval_runtime": 1.2565,
+      "eval_samples_per_second": 2467.123,
+      "eval_steps_per_second": 51.73,
       "step": 954
     },
     {
-      "epoch": 3.14,
-      "learning_rate": 1.3011879804332637e-05,
-      "loss": 0.1093,
+      "epoch": 3.1446540880503147,
+      "grad_norm": 0.47839590907096863,
+      "learning_rate": 1.2138364779874214e-05,
+      "loss": 0.1144,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8993548387096775,
-      "eval_loss": 0.04842004552483559,
-      "eval_runtime": 1.4043,
-      "eval_samples_per_second": 2207.516,
-      "eval_steps_per_second": 46.287,
+      "eval_accuracy": 0.8932258064516129,
+      "eval_loss": 0.051668863743543625,
+      "eval_runtime": 1.2584,
+      "eval_samples_per_second": 2463.451,
+      "eval_steps_per_second": 51.653,
       "step": 1272
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 9.517819706498952e-06,
-      "loss": 0.0703,
+      "epoch": 4.716981132075472,
+      "grad_norm": 0.32125934958457947,
+      "learning_rate": 8.207547169811321e-06,
+      "loss": 0.0737,
       "step": 1500
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9070967741935484,
-      "eval_loss": 0.04050072282552719,
-      "eval_runtime": 1.4216,
-      "eval_samples_per_second": 2180.697,
-      "eval_steps_per_second": 45.724,
+      "eval_accuracy": 0.9048387096774193,
+      "eval_loss": 0.04295943304896355,
+      "eval_runtime": 1.2626,
+      "eval_samples_per_second": 2455.311,
+      "eval_steps_per_second": 51.482,
       "step": 1590
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9138709677419354,
-      "eval_loss": 0.0358940027654171,
-      "eval_runtime": 1.526,
-      "eval_samples_per_second": 2031.447,
-      "eval_steps_per_second": 42.595,
+      "eval_accuracy": 0.9125806451612903,
+      "eval_loss": 0.03821730986237526,
+      "eval_runtime": 1.3321,
+      "eval_samples_per_second": 2327.068,
+      "eval_steps_per_second": 48.793,
       "step": 1908
     },
     {
-      "epoch": 6.29,
-      "learning_rate": 6.02375960866527e-06,
-      "loss": 0.0554,
+      "epoch": 6.289308176100629,
+      "grad_norm": 0.2682598829269409,
+      "learning_rate": 4.276729559748428e-06,
+      "loss": 0.0592,
       "step": 2000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9170967741935484,
-      "eval_loss": 0.03284876048564911,
-      "eval_runtime": 1.3918,
-      "eval_samples_per_second": 2227.252,
-      "eval_steps_per_second": 46.7,
+      "eval_accuracy": 0.9151612903225806,
+      "eval_loss": 0.036075200885534286,
+      "eval_runtime": 1.3604,
+      "eval_samples_per_second": 2278.679,
+      "eval_steps_per_second": 47.779,
       "step": 2226
     },
     {
-      "epoch": 7.86,
-      "learning_rate": 2.5296995108315863e-06,
-      "loss": 0.0488,
+      "epoch": 7.861635220125786,
+      "grad_norm": 0.23691096901893616,
+      "learning_rate": 3.459119496855346e-07,
+      "loss": 0.0531,
       "step": 2500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2862,
+  "max_steps": 2544,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 9,
+  "num_train_epochs": 8,
   "save_steps": 500,
-  "total_flos": 649131947564688.0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 651155886807636.0,
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.8841117631705209,
-    "num_train_epochs": 9,
-    "temperature": 15
+    "alpha": 0.7723229117593493,
+    "num_train_epochs": 8,
+    "temperature": 11
   }
 }

run-2/checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ae147e2e48feced86b55c6b6d628db5362b3c2138b33f8fba746e0ebe68edd51
+oid sha256:ab42167ffff4c1e729d547711e458440cd560d09dd08d069574cd46a86c79e0d
 size 268290900

run-2/checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fe14e21d335d9c04e1a34d733657319a77b7fdb6d8b81be0edccd3467f7f4d4
+oid sha256:c88303ba5ba42b09afa68c73b5722a85e01c14c7abc0f512d64493229c84ab41
 size 536643898

run-2/checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:11e744277c61f87520794334442fae36c5f9ff6e10cb79d4bfee5176ca7eafe2
+oid sha256:90dc4637e972cc69c745eebddd8a7560dca27d2318df3e23f8e145abbf236536
 size 1064

run-2/checkpoint-500/trainer_state.json

@@ -10,25 +10,25 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5741935483870968,
-      "eval_loss": 0.19538737833499908,
-      "eval_runtime": 1.3255,
-      "eval_samples_per_second": 2338.712,
-      "eval_steps_per_second": 49.038,
+      "eval_accuracy": 0.5864516129032258,
+      "eval_loss": 0.2008386254310608,
+      "eval_runtime": 1.3601,
+      "eval_samples_per_second": 2279.22,
+      "eval_steps_per_second": 47.79,
       "step": 318
     },
     {
       "epoch": 1.5723270440251573,
-      "grad_norm": 0.49559739232063293,
-      "learning_rate": 1.550763701707098e-05,
-      "loss": 0.3117,
+      "grad_norm": 0.5175672173500061,
+      "learning_rate": 1.606918238993711e-05,
+      "loss": 0.3218,
       "step": 500
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2226,
+  "max_steps": 2544,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 8,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -46,8 +46,8 @@
   "train_batch_size": 48,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.2890625148941124,
-    "num_train_epochs": 7,
-    "temperature": 17
+    "alpha": 0.7723229117593493,
+    "num_train_epochs": 8,
+    "temperature": 11
   }
 }

run-2/checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02895758c800be06b344602e8d1e4451336b524bdce9f82588ca589e12cfef2e
+oid sha256:baea954a8311f77b8b2d4c67cfd083cf6d7e3bc8a41fb699bd5b137d572b44d3
 size 5176

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48a44b718cd5e63b306a632a23ed7f69877efbab8daac27533495007e71fadba
+oid sha256:baea954a8311f77b8b2d4c67cfd083cf6d7e3bc8a41fb699bd5b137d572b44d3
 size 5176