End of training

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 33.8125,
+    "eval_f1": 48.71262499235238,
+    "eval_runtime": 35.5143,
+    "eval_samples": 4007,
+    "eval_samples_per_second": 112.828,
+    "eval_steps_per_second": 4.055,
+    "total_flos": 1.773811812584448e+16,
+    "train_loss": 2.425102486234544,
+    "train_runtime": 2362.6071,
+    "train_samples": 18102,
+    "train_samples_per_second": 76.619,
+    "train_steps_per_second": 2.739
+}

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e315955726e6538f8a3a2c6577ee04b3fc84f8cb55e4cf5d82fd3a4bf5b5690f
-size 29304733
+oid sha256:53bce694f915c6f2c1951a134dced6c2807b5b85add44ad4f3b54ccb44ae60b4
+size 29329966

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 33.8125,
+    "eval_f1": 48.71262499235238,
+    "eval_runtime": 35.5143,
+    "eval_samples": 4007,
+    "eval_samples_per_second": 112.828,
+    "eval_steps_per_second": 4.055
+}

runs/May12_07-35-03_ml-machine/events.out.tfevents.1715501760.ml-machine.36818.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ce0a533adc0ee96768d36888692b55c9a937193bff578aafc198c474289a8ca
+size 412

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 1.773811812584448e+16,
+    "train_loss": 2.425102486234544,
+    "train_runtime": 2362.6071,
+    "train_samples": 18102,
+    "train_samples_per_second": 76.619,
+    "train_steps_per_second": 2.739
+}

trainer_state.json

@@ -0,0 +1,280 @@
+{
+  "best_metric": 48.71262499235238,
+  "best_model_checkpoint": "/root/turkic_qa/tr_kaz_models/orig_kaz_roberta_base_model/checkpoint-5823",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 6470,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "step": 647,
+      "train_exact_match": 3.5964035964035963,
+      "train_f1": 11.56795258837141,
+      "train_runtime": 11.3677,
+      "train_samples_per_second": 115.503,
+      "train_steps_per_second": 4.135
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.74316120147705,
+      "learning_rate": 5e-06,
+      "loss": 4.6868,
+      "step": 647
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 4.4375,
+      "eval_f1": 11.321307166156533,
+      "eval_runtime": 35.34,
+      "eval_samples_per_second": 113.384,
+      "eval_steps_per_second": 4.075,
+      "step": 647
+    },
+    {
+      "epoch": 2.0,
+      "step": 1294,
+      "train_exact_match": 11.688311688311689,
+      "train_f1": 19.152344840271617,
+      "train_runtime": 11.2191,
+      "train_samples_per_second": 112.844,
+      "train_steps_per_second": 4.1
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.642627716064453,
+      "learning_rate": 1e-05,
+      "loss": 3.5751,
+      "step": 1294
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 9.6875,
+      "eval_f1": 17.179159683522684,
+      "eval_runtime": 35.4377,
+      "eval_samples_per_second": 113.072,
+      "eval_steps_per_second": 4.063,
+      "step": 1294
+    },
+    {
+      "epoch": 3.0,
+      "step": 1941,
+      "train_exact_match": 27.47252747252747,
+      "train_f1": 39.09864539474671,
+      "train_runtime": 11.4784,
+      "train_samples_per_second": 112.646,
+      "train_steps_per_second": 4.095
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 15.969060897827148,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 3.0602,
+      "step": 1941
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 20.625,
+      "eval_f1": 31.5365409883315,
+      "eval_runtime": 35.4806,
+      "eval_samples_per_second": 112.935,
+      "eval_steps_per_second": 4.059,
+      "step": 1941
+    },
+    {
+      "epoch": 4.0,
+      "step": 2588,
+      "train_exact_match": 37.46253746253746,
+      "train_f1": 50.75619914926189,
+      "train_runtime": 11.2214,
+      "train_samples_per_second": 114.513,
+      "train_steps_per_second": 4.099
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 23.334590911865234,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.5025,
+      "step": 2588
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 28.5,
+      "eval_f1": 42.64759962559622,
+      "eval_runtime": 35.4891,
+      "eval_samples_per_second": 112.908,
+      "eval_steps_per_second": 4.058,
+      "step": 2588
+    },
+    {
+      "epoch": 5.0,
+      "step": 3235,
+      "train_exact_match": 47.55244755244755,
+      "train_f1": 60.596926593269394,
+      "train_runtime": 11.4376,
+      "train_samples_per_second": 113.922,
+      "train_steps_per_second": 4.109
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 26.927326202392578,
+      "learning_rate": 6.25e-06,
+      "loss": 2.1403,
+      "step": 3235
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 30.9375,
+      "eval_f1": 45.23519557953087,
+      "eval_runtime": 35.5162,
+      "eval_samples_per_second": 112.822,
+      "eval_steps_per_second": 4.054,
+      "step": 3235
+    },
+    {
+      "epoch": 6.0,
+      "step": 3882,
+      "train_exact_match": 53.04695304695305,
+      "train_f1": 65.4389894343176,
+      "train_runtime": 11.2606,
+      "train_samples_per_second": 114.381,
+      "train_steps_per_second": 4.085
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 19.14756965637207,
+      "learning_rate": 5e-06,
+      "loss": 1.9037,
+      "step": 3882
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 32.46875,
+      "eval_f1": 47.29669704036703,
+      "eval_runtime": 35.4202,
+      "eval_samples_per_second": 113.127,
+      "eval_steps_per_second": 4.065,
+      "step": 3882
+    },
+    {
+      "epoch": 7.0,
+      "step": 4529,
+      "train_exact_match": 57.642357642357645,
+      "train_f1": 69.53035836992328,
+      "train_runtime": 11.2232,
+      "train_samples_per_second": 112.802,
+      "train_steps_per_second": 4.099
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 24.628639221191406,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.7438,
+      "step": 4529
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 33.15625,
+      "eval_f1": 47.60121866597253,
+      "eval_runtime": 35.4127,
+      "eval_samples_per_second": 113.152,
+      "eval_steps_per_second": 4.066,
+      "step": 4529
+    },
+    {
+      "epoch": 8.0,
+      "step": 5176,
+      "train_exact_match": 56.54345654345654,
+      "train_f1": 68.42451603920425,
+      "train_runtime": 11.6061,
+      "train_samples_per_second": 113.044,
+      "train_steps_per_second": 4.05
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 25.127649307250977,
+      "learning_rate": 2.5e-06,
+      "loss": 1.6205,
+      "step": 5176
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 33.625,
+      "eval_f1": 48.59237042611232,
+      "eval_runtime": 35.4006,
+      "eval_samples_per_second": 113.19,
+      "eval_steps_per_second": 4.068,
+      "step": 5176
+    },
+    {
+      "epoch": 9.0,
+      "step": 5823,
+      "train_exact_match": 60.53946053946054,
+      "train_f1": 71.41358726367281,
+      "train_runtime": 11.2378,
+      "train_samples_per_second": 113.545,
+      "train_steps_per_second": 4.093
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 22.670251846313477,
+      "learning_rate": 1.25e-06,
+      "loss": 1.5369,
+      "step": 5823
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 33.8125,
+      "eval_f1": 48.71262499235238,
+      "eval_runtime": 35.5251,
+      "eval_samples_per_second": 112.793,
+      "eval_steps_per_second": 4.053,
+      "step": 5823
+    },
+    {
+      "epoch": 10.0,
+      "step": 6470,
+      "train_exact_match": 63.73626373626374,
+      "train_f1": 74.25641909537842,
+      "train_runtime": 11.4055,
+      "train_samples_per_second": 113.892,
+      "train_steps_per_second": 4.121
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 26.434606552124023,
+      "learning_rate": 0.0,
+      "loss": 1.4813,
+      "step": 6470
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 33.78125,
+      "eval_f1": 48.574128923315,
+      "eval_runtime": 35.5406,
+      "eval_samples_per_second": 112.744,
+      "eval_steps_per_second": 4.052,
+      "step": 6470
+    },
+    {
+      "epoch": 10.0,
+      "step": 6470,
+      "total_flos": 1.773811812584448e+16,
+      "train_loss": 2.425102486234544,
+      "train_runtime": 2362.6071,
+      "train_samples_per_second": 76.619,
+      "train_steps_per_second": 2.739
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 6470,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 1.773811812584448e+16,
+  "train_batch_size": 28,
+  "trial_name": null,
+  "trial_params": null
+}