Training in progress, step 100

all_results.json

@@ -1,20 +1,20 @@
 {
     "epoch": 20.0,
-    "eval_average_metrics": 89.30256169796215,
-    "eval_loss": 0.8872480988502502,
-    "eval_pearson": 89.49792307719518,
-    "eval_runtime": 2.7224,
-    "eval_samples_per_second": 275.492,
-    "eval_spearmanr": 89.10720031872911,
-    "test_average_metrics": 90.79137097057375,
-    "test_loss": 0.8804214000701904,
-    "test_pearson": 90.90568048271486,
-    "test_runtime": 2.8444,
-    "test_samples_per_second": 263.679,
-    "test_spearmanr": 90.67706145843265,
-    "train_loss": 0.8372542282630657,
-    "train_runtime": 586.0447,
+    "eval_average_metrics": 89.89082639431217,
+    "eval_loss": 0.8983659148216248,
+    "eval_pearson": 90.00270057438823,
+    "eval_runtime": 3.833,
+    "eval_samples_per_second": 195.668,
+    "eval_spearmanr": 89.77895221423611,
+    "test_average_metrics": 90.99334935734049,
+    "test_loss": 0.8893818259239197,
+    "test_pearson": 91.06607647471132,
+    "test_runtime": 3.8231,
+    "test_samples_per_second": 196.177,
+    "test_spearmanr": 90.92062223996967,
+    "train_loss": 0.6740480825636121,
+    "train_runtime": 1306.7363,
     "train_samples": 5749,
-    "train_samples_per_second": 196.197,
-    "train_steps_per_second": 1.979
+    "train_samples_per_second": 87.99,
+    "train_steps_per_second": 2.755
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 20.0,
-    "eval_average_metrics": 89.30256169796215,
-    "eval_loss": 0.8872480988502502,
-    "eval_pearson": 89.49792307719518,
-    "eval_runtime": 2.7224,
-    "eval_samples_per_second": 275.492,
-    "eval_spearmanr": 89.10720031872911
+    "eval_average_metrics": 89.89082639431217,
+    "eval_loss": 0.8983659148216248,
+    "eval_pearson": 90.00270057438823,
+    "eval_runtime": 3.833,
+    "eval_samples_per_second": 195.668,
+    "eval_spearmanr": 89.77895221423611
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e6f3f525ae7ff5b2e60a41b766fbedf58090c75cd2cb8e22d35527745ded6aaa
-size 7551621
+oid sha256:512205e7f78c8c9cf316579a829bf531100f40a04095c6cc03ce5a3e028c577a
+size 2631685

runs/Feb01_01-54-14_node2/events.out.tfevents.1643651750.node2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a583ecb026201274f0fd3f0e2c5a65dba6b436ba69d0a24f2f56b6c78c2059fe
-size 18482
+oid sha256:e5d1773ada3fbfae4b2ee4a1692b4db00b36e2781962d7f0e476a66365333ea7
+size 19210

runs/Feb01_01-54-14_node2/events.out.tfevents.1643653061.node2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3f7ec707b58ddf15da5a06c161ece6ef540cfae8bed87bd8b466c7650e57f7b
+size 788

runs/Feb02_16-04-14_node1/1643789122.211467/events.out.tfevents.1643789122.node1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a609709ea354e2ad577ffd9b01e533020f55a5ca1595636296839f31d273b81
+size 5011

runs/Feb02_16-04-14_node1/events.out.tfevents.1643789122.node1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ca643587a88b00378d7b53734fcaf2eeabc788520b4d720fab251891457c363
+size 4300

test_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 20.0,
-    "test_average_metrics": 90.79137097057375,
-    "test_loss": 0.8804214000701904,
-    "test_pearson": 90.90568048271486,
-    "test_runtime": 2.8444,
-    "test_samples_per_second": 263.679,
-    "test_spearmanr": 90.67706145843265
+    "test_average_metrics": 90.99334935734049,
+    "test_loss": 0.8893818259239197,
+    "test_pearson": 91.06607647471132,
+    "test_runtime": 3.8231,
+    "test_samples_per_second": 196.177,
+    "test_spearmanr": 90.92062223996967
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.8372542282630657,
-    "train_runtime": 586.0447,
+    "train_loss": 0.6740480825636121,
+    "train_runtime": 1306.7363,
     "train_samples": 5749,
-    "train_samples_per_second": 196.197,
-    "train_steps_per_second": 1.979
+    "train_samples_per_second": 87.99,
+    "train_steps_per_second": 2.755
 }

trainer_state.json

@@ -1,87 +1,427 @@
 {
-  "best_metric": 89.30256169796215,
-  "best_model_checkpoint": "outputs/bitfit/t5-base/stsb/checkpoint-600",
+  "best_metric": 89.89082639431217,
+  "best_model_checkpoint": "outputs/bitfit/t5-base/stsb/checkpoint-900",
   "epoch": 20.0,
-  "global_step": 1160,
+  "global_step": 3600,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 3.45,
-      "eval_average_metrics": 88.9566249550781,
-      "eval_loss": 0.8938751816749573,
-      "eval_pearson": 88.95684320357462,
-      "eval_runtime": 2.4661,
-      "eval_samples_per_second": 304.124,
-      "eval_spearmanr": 88.95640670658156,
+      "epoch": 0.56,
+      "eval_average_metrics": 89.13717477824025,
+      "eval_loss": 0.9123751521110535,
+      "eval_pearson": 89.06474617714444,
+      "eval_runtime": 3.9799,
+      "eval_samples_per_second": 188.447,
+      "eval_spearmanr": 89.20960337933604,
+      "step": 100
+    },
+    {
+      "epoch": 1.11,
+      "eval_average_metrics": 89.23837198704238,
+      "eval_loss": 0.90521639585495,
+      "eval_pearson": 89.21696934784924,
+      "eval_runtime": 2.8408,
+      "eval_samples_per_second": 264.006,
+      "eval_spearmanr": 89.25977462623553,
       "step": 200
     },
     {
-      "epoch": 6.9,
-      "eval_average_metrics": 88.95231484069625,
-      "eval_loss": 0.8857536911964417,
-      "eval_pearson": 89.08767005696347,
-      "eval_runtime": 2.4501,
-      "eval_samples_per_second": 306.105,
-      "eval_spearmanr": 88.81695962442902,
+      "epoch": 1.67,
+      "eval_average_metrics": 88.85409457301812,
+      "eval_loss": 0.8948209881782532,
+      "eval_pearson": 88.76416701427588,
+      "eval_runtime": 3.5011,
+      "eval_samples_per_second": 214.217,
+      "eval_spearmanr": 88.94402213176035,
+      "step": 300
+    },
+    {
+      "epoch": 2.22,
+      "eval_average_metrics": 88.89962091835693,
+      "eval_loss": 0.8898776173591614,
+      "eval_pearson": 88.91820308047936,
+      "eval_runtime": 3.1447,
+      "eval_samples_per_second": 238.499,
+      "eval_spearmanr": 88.8810387562345,
       "step": 400
     },
     {
-      "epoch": 8.62,
-      "learning_rate": 0.0003,
-      "loss": 0.8558,
+      "epoch": 2.78,
+      "learning_rate": 0.00025833333333333334,
+      "loss": 0.8458,
+      "step": 500
+    },
+    {
+      "epoch": 2.78,
+      "eval_average_metrics": 89.72250747135166,
+      "eval_loss": 0.8902953863143921,
+      "eval_pearson": 89.89215489261863,
+      "eval_runtime": 4.5165,
+      "eval_samples_per_second": 166.059,
+      "eval_spearmanr": 89.55286005008469,
       "step": 500
     },
     {
-      "epoch": 10.34,
-      "eval_average_metrics": 89.30256169796215,
-      "eval_loss": 0.8872480988502502,
-      "eval_pearson": 89.49792307719518,
-      "eval_runtime": 2.7534,
-      "eval_samples_per_second": 272.387,
-      "eval_spearmanr": 89.10720031872911,
+      "epoch": 3.33,
+      "eval_average_metrics": 89.46605644881728,
+      "eval_loss": 0.927269697189331,
+      "eval_pearson": 89.51595346590146,
+      "eval_runtime": 4.621,
+      "eval_samples_per_second": 162.302,
+      "eval_spearmanr": 89.41615943173312,
       "step": 600
     },
     {
-      "epoch": 13.79,
-      "eval_average_metrics": 89.00274406096713,
-      "eval_loss": 0.8818585872650146,
-      "eval_pearson": 89.11339528387046,
-      "eval_runtime": 2.7298,
-      "eval_samples_per_second": 274.743,
-      "eval_spearmanr": 88.8920928380638,
+      "epoch": 3.89,
+      "eval_average_metrics": 89.25697052652335,
+      "eval_loss": 0.8973984122276306,
+      "eval_pearson": 89.47540644836778,
+      "eval_runtime": 3.2452,
+      "eval_samples_per_second": 231.108,
+      "eval_spearmanr": 89.03853460467892,
+      "step": 700
+    },
+    {
+      "epoch": 4.44,
+      "eval_average_metrics": 89.77994978800227,
+      "eval_loss": 0.9015573859214783,
+      "eval_pearson": 89.98660087713056,
+      "eval_runtime": 5.9734,
+      "eval_samples_per_second": 125.557,
+      "eval_spearmanr": 89.57329869887398,
       "step": 800
     },
     {
-      "epoch": 17.24,
-      "learning_rate": 7.272727272727273e-05,
-      "loss": 0.8251,
+      "epoch": 5.0,
+      "eval_average_metrics": 89.89082639431217,
+      "eval_loss": 0.8983659148216248,
+      "eval_pearson": 90.00270057438823,
+      "eval_runtime": 4.0652,
+      "eval_samples_per_second": 184.494,
+      "eval_spearmanr": 89.77895221423611,
+      "step": 900
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00021666666666666666,
+      "loss": 0.7859,
       "step": 1000
     },
     {
-      "epoch": 17.24,
-      "eval_average_metrics": 89.29097960305327,
-      "eval_loss": 0.8865894079208374,
-      "eval_pearson": 89.46184016190372,
-      "eval_runtime": 2.3462,
-      "eval_samples_per_second": 319.662,
-      "eval_spearmanr": 89.12011904420282,
+      "epoch": 5.56,
+      "eval_average_metrics": 89.15652744561595,
+      "eval_loss": 0.9311385154724121,
+      "eval_pearson": 89.25999907777292,
+      "eval_runtime": 3.1115,
+      "eval_samples_per_second": 241.042,
+      "eval_spearmanr": 89.053055813459,
       "step": 1000
     },
+    {
+      "epoch": 6.11,
+      "eval_average_metrics": 89.36900055711244,
+      "eval_loss": 0.914900541305542,
+      "eval_pearson": 89.48560950002296,
+      "eval_runtime": 4.2641,
+      "eval_samples_per_second": 175.886,
+      "eval_spearmanr": 89.25239161420193,
+      "step": 1100
+    },
+    {
+      "epoch": 6.67,
+      "eval_average_metrics": 89.41410457111328,
+      "eval_loss": 0.9506754279136658,
+      "eval_pearson": 89.45609872168974,
+      "eval_runtime": 3.0012,
+      "eval_samples_per_second": 249.9,
+      "eval_spearmanr": 89.37211042053684,
+      "step": 1200
+    },
+    {
+      "epoch": 7.22,
+      "eval_average_metrics": 89.44094336465075,
+      "eval_loss": 0.953001081943512,
+      "eval_pearson": 89.51193389114643,
+      "eval_runtime": 2.7971,
+      "eval_samples_per_second": 268.139,
+      "eval_spearmanr": 89.36995283815507,
+      "step": 1300
+    },
+    {
+      "epoch": 7.78,
+      "eval_average_metrics": 89.17605484273516,
+      "eval_loss": 0.9723613262176514,
+      "eval_pearson": 89.21895062324246,
+      "eval_runtime": 3.7026,
+      "eval_samples_per_second": 202.559,
+      "eval_spearmanr": 89.13315906222785,
+      "step": 1400
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 0.000175,
+      "loss": 0.724,
+      "step": 1500
+    },
+    {
+      "epoch": 8.33,
+      "eval_average_metrics": 89.48308644886173,
+      "eval_loss": 0.9874196648597717,
+      "eval_pearson": 89.5217033530652,
+      "eval_runtime": 3.886,
+      "eval_samples_per_second": 193.003,
+      "eval_spearmanr": 89.44446954465825,
+      "step": 1500
+    },
+    {
+      "epoch": 8.89,
+      "eval_average_metrics": 89.79760216722264,
+      "eval_loss": 0.9869381785392761,
+      "eval_pearson": 89.95375320955753,
+      "eval_runtime": 4.168,
+      "eval_samples_per_second": 179.942,
+      "eval_spearmanr": 89.64145112488774,
+      "step": 1600
+    },
+    {
+      "epoch": 9.44,
+      "eval_average_metrics": 89.29279708906819,
+      "eval_loss": 0.9963127970695496,
+      "eval_pearson": 89.46662281864026,
+      "eval_runtime": 3.4947,
+      "eval_samples_per_second": 214.608,
+      "eval_spearmanr": 89.11897135949613,
+      "step": 1700
+    },
+    {
+      "epoch": 10.0,
+      "eval_average_metrics": 89.79299996621938,
+      "eval_loss": 0.9937890768051147,
+      "eval_pearson": 89.88666855723739,
+      "eval_runtime": 4.5917,
+      "eval_samples_per_second": 163.339,
+      "eval_spearmanr": 89.69933137520137,
+      "step": 1800
+    },
+    {
+      "epoch": 10.56,
+      "eval_average_metrics": 89.27048678268574,
+      "eval_loss": 1.006996750831604,
+      "eval_pearson": 89.39560825429157,
+      "eval_runtime": 3.591,
+      "eval_samples_per_second": 208.856,
+      "eval_spearmanr": 89.1453653110799,
+      "step": 1900
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 0.0001333333333333333,
+      "loss": 0.6693,
+      "step": 2000
+    },
+    {
+      "epoch": 11.11,
+      "eval_average_metrics": 89.62053083130016,
+      "eval_loss": 1.0332725048065186,
+      "eval_pearson": 89.69477341302792,
+      "eval_runtime": 3.768,
+      "eval_samples_per_second": 199.043,
+      "eval_spearmanr": 89.54628824957238,
+      "step": 2000
+    },
+    {
+      "epoch": 11.67,
+      "eval_average_metrics": 89.13711909449259,
+      "eval_loss": 1.0669463872909546,
+      "eval_pearson": 89.29635083082506,
+      "eval_runtime": 3.5887,
+      "eval_samples_per_second": 208.99,
+      "eval_spearmanr": 88.97788735816013,
+      "step": 2100
+    },
+    {
+      "epoch": 12.22,
+      "eval_average_metrics": 89.42971483116142,
+      "eval_loss": 1.0573464632034302,
+      "eval_pearson": 89.58674143842262,
+      "eval_runtime": 3.0808,
+      "eval_samples_per_second": 243.441,
+      "eval_spearmanr": 89.2726882239002,
+      "step": 2200
+    },
+    {
+      "epoch": 12.78,
+      "eval_average_metrics": 89.42097809467734,
+      "eval_loss": 1.0824470520019531,
+      "eval_pearson": 89.54344639693154,
+      "eval_runtime": 3.8679,
+      "eval_samples_per_second": 193.903,
+      "eval_spearmanr": 89.29850979242312,
+      "step": 2300
+    },
+    {
+      "epoch": 13.33,
+      "eval_average_metrics": 89.38136709242659,
+      "eval_loss": 1.0812472105026245,
+      "eval_pearson": 89.52312670086297,
+      "eval_runtime": 3.7051,
+      "eval_samples_per_second": 202.426,
+      "eval_spearmanr": 89.23960748399018,
+      "step": 2400
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 9.166666666666667e-05,
+      "loss": 0.6142,
+      "step": 2500
+    },
+    {
+      "epoch": 13.89,
+      "eval_average_metrics": 89.82337151387273,
+      "eval_loss": 1.0920954942703247,
+      "eval_pearson": 89.99504557050292,
+      "eval_runtime": 3.8424,
+      "eval_samples_per_second": 195.19,
+      "eval_spearmanr": 89.65169745724253,
+      "step": 2500
+    },
+    {
+      "epoch": 14.44,
+      "eval_average_metrics": 89.56749485985972,
+      "eval_loss": 1.1302155256271362,
+      "eval_pearson": 89.70885507009902,
+      "eval_runtime": 3.9276,
+      "eval_samples_per_second": 190.954,
+      "eval_spearmanr": 89.42613464962042,
+      "step": 2600
+    },
+    {
+      "epoch": 15.0,
+      "eval_average_metrics": 89.61009900620351,
+      "eval_loss": 1.14012610912323,
+      "eval_pearson": 89.68171447646405,
+      "eval_runtime": 3.7232,
+      "eval_samples_per_second": 201.438,
+      "eval_spearmanr": 89.53848353594297,
+      "step": 2700
+    },
+    {
+      "epoch": 15.56,
+      "eval_average_metrics": 89.27459685358238,
+      "eval_loss": 1.1592134237289429,
+      "eval_pearson": 89.32968690818579,
+      "eval_runtime": 4.52,
+      "eval_samples_per_second": 165.929,
+      "eval_spearmanr": 89.21950679897895,
+      "step": 2800
+    },
+    {
+      "epoch": 16.11,
+      "eval_average_metrics": 89.65586988882487,
+      "eval_loss": 1.1740944385528564,
+      "eval_pearson": 89.72543322190315,
+      "eval_runtime": 3.2065,
+      "eval_samples_per_second": 233.903,
+      "eval_spearmanr": 89.5863065557466,
+      "step": 2900
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 4.9999999999999996e-05,
+      "loss": 0.5671,
+      "step": 3000
+    },
+    {
+      "epoch": 16.67,
+      "eval_average_metrics": 89.74259369919744,
+      "eval_loss": 1.1778316497802734,
+      "eval_pearson": 89.82066967172904,
+      "eval_runtime": 3.705,
+      "eval_samples_per_second": 202.432,
+      "eval_spearmanr": 89.66451772666586,
+      "step": 3000
+    },
+    {
+      "epoch": 17.22,
+      "eval_average_metrics": 89.67436613731013,
+      "eval_loss": 1.1869001388549805,
+      "eval_pearson": 89.822535941685,
+      "eval_runtime": 4.6778,
+      "eval_samples_per_second": 160.331,
+      "eval_spearmanr": 89.52619633293524,
+      "step": 3100
+    },
+    {
+      "epoch": 17.78,
+      "eval_average_metrics": 89.64929997632476,
+      "eval_loss": 1.1901123523712158,
+      "eval_pearson": 89.78058123437394,
+      "eval_runtime": 3.3067,
+      "eval_samples_per_second": 226.812,
+      "eval_spearmanr": 89.51801871827558,
+      "step": 3200
+    },
+    {
+      "epoch": 18.33,
+      "eval_average_metrics": 89.78282294446872,
+      "eval_loss": 1.1919935941696167,
+      "eval_pearson": 89.9585327889659,
+      "eval_runtime": 4.0203,
+      "eval_samples_per_second": 186.553,
+      "eval_spearmanr": 89.60711309997154,
+      "step": 3300
+    },
+    {
+      "epoch": 18.89,
+      "eval_average_metrics": 89.64203906594571,
+      "eval_loss": 1.1938636302947998,
+      "eval_pearson": 89.76878899591361,
+      "eval_runtime": 2.5879,
+      "eval_samples_per_second": 289.807,
+      "eval_spearmanr": 89.51528913597781,
+      "step": 3400
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 8.333333333333332e-06,
+      "loss": 0.5407,
+      "step": 3500
+    },
+    {
+      "epoch": 19.44,
+      "eval_average_metrics": 89.85580070489866,
+      "eval_loss": 1.2061032056808472,
+      "eval_pearson": 89.99507172793925,
+      "eval_runtime": 3.3291,
+      "eval_samples_per_second": 225.288,
+      "eval_spearmanr": 89.71652968185808,
+      "step": 3500
+    },
+    {
+      "epoch": 20.0,
+      "eval_average_metrics": 89.66212005188657,
+      "eval_loss": 1.2073893547058105,
+      "eval_pearson": 89.79939157390831,
+      "eval_runtime": 3.5775,
+      "eval_samples_per_second": 209.645,
+      "eval_spearmanr": 89.52484852986483,
+      "step": 3600
+    },
     {
       "epoch": 20.0,
-      "step": 1160,
-      "total_flos": 1.3780641944564136e+16,
-      "train_loss": 0.8372542282630657,
-      "train_runtime": 586.0447,
-      "train_samples_per_second": 196.197,
-      "train_steps_per_second": 1.979
+      "step": 3600,
+      "total_flos": 1.2081375923975424e+16,
+      "train_loss": 0.6740480825636121,
+      "train_runtime": 1306.7363,
+      "train_samples_per_second": 87.99,
+      "train_steps_per_second": 2.755
     }
   ],
-  "max_steps": 1160,
+  "max_steps": 3600,
   "num_train_epochs": 20,
-  "total_flos": 1.3780641944564136e+16,
+  "total_flos": 1.2081375923975424e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4532ae8df18f200354792e9d75508579d85eda3d88a52d81a7dd97a8a6facb9c
+oid sha256:1dae689eba1a2f300b89fa3369aa940714da48e58d5e0706f64c6a408dfeddb2
 size 3183

training_config.json

@@ -1 +1 @@
-{"bottleneck_dim": 24, "dataset_config_name": ["en"], "delta_type": "adapter", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "stsb", "eval_steps": 100, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "max_source_length": 128, "metric_for_best_model": "average_metrics", "model_name_or_path": "../../../../plm_cache/t5-base", "num_train_epochs": 20, "output_dir": "outputs/bitfit/t5-base/stsb", "overwrite_output_dir": true, "per_device_eval_batch_size": 32, "per_device_train_batch_size": 32, "predict_with_generate": true, "push_to_hub": true, "save_steps": 100, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "split_validation_test": true, "task_name": "stsb", "test_dataset_config_name": ["en"], "test_dataset_name": "stsb", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "warmup_steps": 0}
+{"dataset_config_name": ["en"], "delta_type": "lora", "do_eval": true, "do_test": true, "do_train": true, "eval_dataset_config_name": ["en"], "eval_dataset_name": "stsb", "eval_steps": 100, "evaluation_strategy": "steps", "greater_is_better": true, "learning_rate": 0.0003, "load_best_model_at_end": true, "lora_r": 8, "max_source_length": 128, "metric_for_best_model": "average_metrics", "model_name_or_path": "../../../../plm_cache/t5-base", "num_train_epochs": 20, "output_dir": "outputs/bitfit/t5-base/stsb", "overwrite_output_dir": true, "per_device_eval_batch_size": 32, "per_device_train_batch_size": 32, "predict_with_generate": true, "push_to_hub": true, "save_steps": 100, "save_strategy": "steps", "save_total_limit": 1, "seed": 42, "split_validation_test": true, "task_name": "stsb", "test_dataset_config_name": ["en"], "test_dataset_name": "stsb", "tokenizer_name": "../../../../plm_cache/t5-base", "unfrozen_modules": ["deltas", "layer_norm", "final_layer_norm"], "warmup_steps": 0}