End of training

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 56.53125,
+    "eval_f1": 70.80451562424408,
+    "eval_runtime": 44.2214,
+    "eval_samples": 3858,
+    "eval_samples_per_second": 87.243,
+    "eval_steps_per_second": 3.121,
+    "total_flos": 3.575911440121344e+16,
+    "train_loss": 0.8889602286684001,
+    "train_runtime": 4193.6079,
+    "train_samples": 18247,
+    "train_samples_per_second": 43.511,
+    "train_steps_per_second": 1.555
+}

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b588600389c604d0d570cfd25ed7b12d75ec850122a87de096fae4b15e0480c
-size 23845343
+oid sha256:bb5d84200be2d392afe7c7056148f683d4ae044060013f5656bbd2c979a0baad
+size 24138494

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "eval_exact_match": 56.53125,
+    "eval_f1": 70.80451562424408,
+    "eval_runtime": 44.2214,
+    "eval_samples": 3858,
+    "eval_samples_per_second": 87.243,
+    "eval_steps_per_second": 3.121
+}

runs/May09_14-26-37_ml-machine/events.out.tfevents.1715269146.ml-machine.35123.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9b69b2d9dcf8d3b7e1fecd1e3fd8edbe8301b4db9705ea749bbee6148958722
+size 412

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 3.575911440121344e+16,
+    "train_loss": 0.8889602286684001,
+    "train_runtime": 4193.6079,
+    "train_samples": 18247,
+    "train_samples_per_second": 43.511,
+    "train_steps_per_second": 1.555
+}

trainer_state.json

@@ -0,0 +1,280 @@
+{
+  "best_metric": 70.80451562424408,
+  "best_model_checkpoint": "/root/turkic_qa/en_kaz_models/en_kaz_xlm_roberta_base_squad_model/checkpoint-3260",
+  "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": 54.645354645354644,
+      "train_f1": 71.43920046551005,
+      "train_runtime": 13.6364,
+      "train_samples_per_second": 89.98,
+      "train_steps_per_second": 3.227
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 42.09185028076172,
+      "learning_rate": 5e-06,
+      "loss": 1.6894,
+      "step": 652
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 51.0,
+      "eval_f1": 66.75535596211667,
+      "eval_runtime": 42.9083,
+      "eval_samples_per_second": 89.913,
+      "eval_steps_per_second": 3.216,
+      "step": 652
+    },
+    {
+      "epoch": 2.0,
+      "step": 1304,
+      "train_exact_match": 60.03996003996004,
+      "train_f1": 74.97586392957278,
+      "train_runtime": 13.9152,
+      "train_samples_per_second": 89.183,
+      "train_steps_per_second": 3.234
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.847925186157227,
+      "learning_rate": 1e-05,
+      "loss": 1.3571,
+      "step": 1304
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 54.0625,
+      "eval_f1": 69.25389311582437,
+      "eval_runtime": 43.0882,
+      "eval_samples_per_second": 89.537,
+      "eval_steps_per_second": 3.203,
+      "step": 1304
+    },
+    {
+      "epoch": 3.0,
+      "step": 1956,
+      "train_exact_match": 68.23176823176823,
+      "train_f1": 81.76253341059837,
+      "train_runtime": 13.5561,
+      "train_samples_per_second": 88.964,
+      "train_steps_per_second": 3.246
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 34.729644775390625,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.1643,
+      "step": 1956
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 54.9375,
+      "eval_f1": 69.89098890413919,
+      "eval_runtime": 43.0049,
+      "eval_samples_per_second": 89.711,
+      "eval_steps_per_second": 3.209,
+      "step": 1956
+    },
+    {
+      "epoch": 4.0,
+      "step": 2608,
+      "train_exact_match": 71.82817182817183,
+      "train_f1": 85.225535613071,
+      "train_runtime": 14.2911,
+      "train_samples_per_second": 88.167,
+      "train_steps_per_second": 3.149
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 33.257789611816406,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9703,
+      "step": 2608
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 56.1875,
+      "eval_f1": 70.69884423902388,
+      "eval_runtime": 43.7869,
+      "eval_samples_per_second": 88.109,
+      "eval_steps_per_second": 3.152,
+      "step": 2608
+    },
+    {
+      "epoch": 5.0,
+      "step": 3260,
+      "train_exact_match": 76.42357642357642,
+      "train_f1": 88.7083439838053,
+      "train_runtime": 13.5599,
+      "train_samples_per_second": 89.234,
+      "train_steps_per_second": 3.245
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 29.723169326782227,
+      "learning_rate": 6.25e-06,
+      "loss": 0.8246,
+      "step": 3260
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 56.53125,
+      "eval_f1": 70.80451562424408,
+      "eval_runtime": 43.1752,
+      "eval_samples_per_second": 89.357,
+      "eval_steps_per_second": 3.196,
+      "step": 3260
+    },
+    {
+      "epoch": 6.0,
+      "step": 3912,
+      "train_exact_match": 77.32267732267732,
+      "train_f1": 89.41850450829368,
+      "train_runtime": 13.3805,
+      "train_samples_per_second": 87.964,
+      "train_steps_per_second": 3.214
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 43.614688873291016,
+      "learning_rate": 5e-06,
+      "loss": 0.7106,
+      "step": 3912
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 56.28125,
+      "eval_f1": 70.69128089372614,
+      "eval_runtime": 43.3271,
+      "eval_samples_per_second": 89.044,
+      "eval_steps_per_second": 3.185,
+      "step": 3912
+    },
+    {
+      "epoch": 7.0,
+      "step": 4564,
+      "train_exact_match": 80.31968031968032,
+      "train_f1": 90.9932970975269,
+      "train_runtime": 13.7154,
+      "train_samples_per_second": 88.222,
+      "train_steps_per_second": 3.208
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 16.68464469909668,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6206,
+      "step": 4564
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 56.28125,
+      "eval_f1": 70.50206404433744,
+      "eval_runtime": 43.2324,
+      "eval_samples_per_second": 89.239,
+      "eval_steps_per_second": 3.192,
+      "step": 4564
+    },
+    {
+      "epoch": 8.0,
+      "step": 5216,
+      "train_exact_match": 82.41758241758242,
+      "train_f1": 92.16600808934162,
+      "train_runtime": 13.6769,
+      "train_samples_per_second": 88.105,
+      "train_steps_per_second": 3.217
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 30.79519271850586,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5603,
+      "step": 5216
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 56.40625,
+      "eval_f1": 70.75408379302462,
+      "eval_runtime": 43.2887,
+      "eval_samples_per_second": 89.123,
+      "eval_steps_per_second": 3.188,
+      "step": 5216
+    },
+    {
+      "epoch": 9.0,
+      "step": 5868,
+      "train_exact_match": 83.61638361638362,
+      "train_f1": 93.40487188804495,
+      "train_runtime": 13.7502,
+      "train_samples_per_second": 88.508,
+      "train_steps_per_second": 3.2
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 45.799346923828125,
+      "learning_rate": 1.25e-06,
+      "loss": 0.5138,
+      "step": 5868
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 56.46875,
+      "eval_f1": 70.54295738591223,
+      "eval_runtime": 43.2261,
+      "eval_samples_per_second": 89.252,
+      "eval_steps_per_second": 3.193,
+      "step": 5868
+    },
+    {
+      "epoch": 10.0,
+      "step": 6520,
+      "train_exact_match": 84.31568431568432,
+      "train_f1": 93.73109223264476,
+      "train_runtime": 13.3013,
+      "train_samples_per_second": 88.563,
+      "train_steps_per_second": 3.233
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 22.400699615478516,
+      "learning_rate": 0.0,
+      "loss": 0.4785,
+      "step": 6520
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 56.34375,
+      "eval_f1": 70.56177522205788,
+      "eval_runtime": 43.1408,
+      "eval_samples_per_second": 89.428,
+      "eval_steps_per_second": 3.199,
+      "step": 6520
+    },
+    {
+      "epoch": 10.0,
+      "step": 6520,
+      "total_flos": 3.575911440121344e+16,
+      "train_loss": 0.8889602286684001,
+      "train_runtime": 4193.6079,
+      "train_samples_per_second": 43.511,
+      "train_steps_per_second": 1.555
+    }
+  ],
+  "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
+}