End of training

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 52.96875,
+    "eval_f1": 66.94393744369789,
+    "eval_runtime": 42.7566,
+    "eval_samples": 3858,
+    "eval_samples_per_second": 90.232,
+    "eval_steps_per_second": 3.228,
+    "total_flos": 3.575911440121344e+16,
+    "train_loss": 1.4707286109222224,
+    "train_runtime": 4171.843,
+    "train_samples": 18247,
+    "train_samples_per_second": 43.738,
+    "train_steps_per_second": 1.563
+}

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3aa5003dacda993b811f094a6af98f42f4a687b4ffb47858dbbedf47442b730d
-size 23994485
+oid sha256:113cd778ada546481f4df0b7e3b6b380d157f57bf2003a6f29ce2d32b0b89b96
+size 24315769

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 52.96875,
+    "eval_f1": 66.94393744369789,
+    "eval_runtime": 42.7566,
+    "eval_samples": 3858,
+    "eval_samples_per_second": 90.232,
+    "eval_steps_per_second": 3.228
+}

runs/May09_13-10-15_ml-machine/events.out.tfevents.1715264539.ml-machine.14546.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33b04a4fe7b9867ca9efc29ab1d064ca1640836f45b2dbc731b7c297df1706f0
+size 412

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 3.575911440121344e+16,
+    "train_loss": 1.4707286109222224,
+    "train_runtime": 4171.843,
+    "train_samples": 18247,
+    "train_samples_per_second": 43.738,
+    "train_steps_per_second": 1.563
+}

trainer_state.json

@@ -0,0 +1,280 @@
+{
+  "best_metric": 66.94393744369789,
+  "best_model_checkpoint": "/root/turkic_qa/en_kaz_models/en_kaz_xlm_roberta_base_model/checkpoint-4564",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 6520,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "step": 652,
+      "train_exact_match": 23.376623376623378,
+      "train_f1": 36.50743629600778,
+      "train_runtime": 13.9023,
+      "train_samples_per_second": 88.259,
+      "train_steps_per_second": 3.165
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.58616256713867,
+      "learning_rate": 5e-06,
+      "loss": 4.6656,
+      "step": 652
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 19.9375,
+      "eval_f1": 32.80148318666002,
+      "eval_runtime": 43.5607,
+      "eval_samples_per_second": 88.566,
+      "eval_steps_per_second": 3.168,
+      "step": 652
+    },
+    {
+      "epoch": 2.0,
+      "step": 1304,
+      "train_exact_match": 47.15284715284715,
+      "train_f1": 62.44620681983165,
+      "train_runtime": 13.8553,
+      "train_samples_per_second": 89.568,
+      "train_steps_per_second": 3.248
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 37.57294845581055,
+      "learning_rate": 1e-05,
+      "loss": 2.2206,
+      "step": 1304
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 43.125,
+      "eval_f1": 59.36157776305659,
+      "eval_runtime": 42.8686,
+      "eval_samples_per_second": 89.996,
+      "eval_steps_per_second": 3.219,
+      "step": 1304
+    },
+    {
+      "epoch": 3.0,
+      "step": 1956,
+      "train_exact_match": 58.841158841158844,
+      "train_f1": 73.32741649456507,
+      "train_runtime": 13.6223,
+      "train_samples_per_second": 88.532,
+      "train_steps_per_second": 3.23
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 28.621103286743164,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.5696,
+      "step": 1956
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 48.875,
+      "eval_f1": 64.04956970752869,
+      "eval_runtime": 42.952,
+      "eval_samples_per_second": 89.821,
+      "eval_steps_per_second": 3.213,
+      "step": 1956
+    },
+    {
+      "epoch": 4.0,
+      "step": 2608,
+      "train_exact_match": 64.63536463536464,
+      "train_f1": 77.93182455149633,
+      "train_runtime": 13.9838,
+      "train_samples_per_second": 90.104,
+      "train_steps_per_second": 3.218
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 32.638771057128906,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.2762,
+      "step": 2608
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 50.21875,
+      "eval_f1": 65.455060310372,
+      "eval_runtime": 42.6703,
+      "eval_samples_per_second": 90.414,
+      "eval_steps_per_second": 3.234,
+      "step": 2608
+    },
+    {
+      "epoch": 5.0,
+      "step": 3260,
+      "train_exact_match": 67.43256743256744,
+      "train_f1": 81.29037094900592,
+      "train_runtime": 13.5595,
+      "train_samples_per_second": 89.236,
+      "train_steps_per_second": 3.245
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 30.29413414001465,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0933,
+      "step": 3260
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 52.09375,
+      "eval_f1": 66.46298976895542,
+      "eval_runtime": 42.7773,
+      "eval_samples_per_second": 90.188,
+      "eval_steps_per_second": 3.226,
+      "step": 3260
+    },
+    {
+      "epoch": 6.0,
+      "step": 3912,
+      "train_exact_match": 72.22777222777223,
+      "train_f1": 85.17963193748265,
+      "train_runtime": 13.2269,
+      "train_samples_per_second": 88.986,
+      "train_steps_per_second": 3.251
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 39.73823547363281,
+      "learning_rate": 5e-06,
+      "loss": 0.9411,
+      "step": 3912
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 52.1875,
+      "eval_f1": 66.65227927064453,
+      "eval_runtime": 42.8644,
+      "eval_samples_per_second": 90.005,
+      "eval_steps_per_second": 3.219,
+      "step": 3912
+    },
+    {
+      "epoch": 7.0,
+      "step": 4564,
+      "train_exact_match": 73.22677322677323,
+      "train_f1": 86.58009920694694,
+      "train_runtime": 13.5591,
+      "train_samples_per_second": 89.239,
+      "train_steps_per_second": 3.245
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 25.444896697998047,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8381,
+      "step": 4564
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 52.96875,
+      "eval_f1": 66.94393744369789,
+      "eval_runtime": 42.8557,
+      "eval_samples_per_second": 90.023,
+      "eval_steps_per_second": 3.22,
+      "step": 4564
+    },
+    {
+      "epoch": 8.0,
+      "step": 5216,
+      "train_exact_match": 77.62237762237763,
+      "train_f1": 88.51013925221713,
+      "train_runtime": 13.9418,
+      "train_samples_per_second": 86.431,
+      "train_steps_per_second": 3.156
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 26.798072814941406,
+      "learning_rate": 2.5e-06,
+      "loss": 0.7554,
+      "step": 5216
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 52.59375,
+      "eval_f1": 66.93975585361125,
+      "eval_runtime": 43.736,
+      "eval_samples_per_second": 88.211,
+      "eval_steps_per_second": 3.155,
+      "step": 5216
+    },
+    {
+      "epoch": 9.0,
+      "step": 5868,
+      "train_exact_match": 77.82217782217782,
+      "train_f1": 90.31779721748207,
+      "train_runtime": 13.6001,
+      "train_samples_per_second": 89.484,
+      "train_steps_per_second": 3.235
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 40.23249816894531,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6921,
+      "step": 5868
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 52.53125,
+      "eval_f1": 66.82837223175379,
+      "eval_runtime": 43.0325,
+      "eval_samples_per_second": 89.653,
+      "eval_steps_per_second": 3.207,
+      "step": 5868
+    },
+    {
+      "epoch": 10.0,
+      "step": 6520,
+      "train_exact_match": 78.82117882117882,
+      "train_f1": 90.52466502877378,
+      "train_runtime": 13.1929,
+      "train_samples_per_second": 89.29,
+      "train_steps_per_second": 3.259
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 33.59892654418945,
+      "learning_rate": 0.0,
+      "loss": 0.6552,
+      "step": 6520
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 52.78125,
+      "eval_f1": 66.90542651172623,
+      "eval_runtime": 42.7756,
+      "eval_samples_per_second": 90.192,
+      "eval_steps_per_second": 3.226,
+      "step": 6520
+    },
+    {
+      "epoch": 10.0,
+      "step": 6520,
+      "total_flos": 3.575911440121344e+16,
+      "train_loss": 1.4707286109222224,
+      "train_runtime": 4171.843,
+      "train_samples_per_second": 43.738,
+      "train_steps_per_second": 1.563
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 6520,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 3.575911440121344e+16,
+  "train_batch_size": 28,
+  "trial_name": null,
+  "trial_params": null
+}