Upload batch 6/7: sst2/mBERT/multi_branch_average/seed_43 ... vsfc/mBERT/hf_sequence_classifier/seed_44

mbert_rtx6000_metrics/result_ablation_mbert/vsfc/mBERT/gated_multi_branch/seed_43/all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9355653821857233,
+    "eval_loss": 0.22262658178806305,
+    "eval_runtime": 5.0451,
+    "eval_samples": 1583,
+    "eval_samples_per_second": 313.768,
+    "eval_steps_per_second": 3.37,
+    "total_flos": 1055522845962000.0,
+    "train_loss": 0.20581150786480976,
+    "train_runtime": 66.9935,
+    "train_samples": 11426,
+    "train_samples_per_second": 511.662,
+    "train_steps_per_second": 10.703
+}

mbert_rtx6000_metrics/result_ablation_mbert/vsfc/mBERT/gated_multi_branch/seed_43/eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9355653821857233,
+    "eval_loss": 0.22262658178806305,
+    "eval_runtime": 5.0451,
+    "eval_samples": 1583,
+    "eval_samples_per_second": 313.768,
+    "eval_steps_per_second": 3.37
+}

mbert_rtx6000_metrics/result_ablation_mbert/vsfc/mBERT/gated_multi_branch/seed_43/train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "total_flos": 1055522845962000.0,
+    "train_loss": 0.20581150786480976,
+    "train_runtime": 66.9935,
+    "train_samples": 11426,
+    "train_samples_per_second": 511.662,
+    "train_steps_per_second": 10.703
+}

mbert_rtx6000_metrics/result_ablation_mbert/vsfc/mBERT/gated_multi_branch/seed_43/trainer_state.json

@@ -0,0 +1,265 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 717,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.10460251046025104,
+      "grad_norm": 5.261765480041504,
+      "learning_rate": 2.8953974895397488e-05,
+      "loss": 0.4915,
+      "step": 25
+    },
+    {
+      "epoch": 0.20920502092050208,
+      "grad_norm": 3.6981773376464844,
+      "learning_rate": 2.7907949790794982e-05,
+      "loss": 0.3149,
+      "step": 50
+    },
+    {
+      "epoch": 0.3138075313807531,
+      "grad_norm": 7.417821884155273,
+      "learning_rate": 2.686192468619247e-05,
+      "loss": 0.3092,
+      "step": 75
+    },
+    {
+      "epoch": 0.41841004184100417,
+      "grad_norm": 5.29278564453125,
+      "learning_rate": 2.5815899581589957e-05,
+      "loss": 0.2702,
+      "step": 100
+    },
+    {
+      "epoch": 0.5230125523012552,
+      "grad_norm": 9.274847984313965,
+      "learning_rate": 2.4769874476987448e-05,
+      "loss": 0.3182,
+      "step": 125
+    },
+    {
+      "epoch": 0.6276150627615062,
+      "grad_norm": 5.753255367279053,
+      "learning_rate": 2.372384937238494e-05,
+      "loss": 0.3011,
+      "step": 150
+    },
+    {
+      "epoch": 0.7322175732217573,
+      "grad_norm": 4.454889297485352,
+      "learning_rate": 2.2677824267782426e-05,
+      "loss": 0.2741,
+      "step": 175
+    },
+    {
+      "epoch": 0.8368200836820083,
+      "grad_norm": 5.513760566711426,
+      "learning_rate": 2.1631799163179917e-05,
+      "loss": 0.2326,
+      "step": 200
+    },
+    {
+      "epoch": 0.9414225941422594,
+      "grad_norm": 6.699063777923584,
+      "learning_rate": 2.0585774058577408e-05,
+      "loss": 0.241,
+      "step": 225
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.930511686670878,
+      "eval_loss": 0.23169396817684174,
+      "eval_runtime": 5.2812,
+      "eval_samples_per_second": 299.741,
+      "eval_steps_per_second": 3.219,
+      "step": 239
+    },
+    {
+      "epoch": 1.0460251046025104,
+      "grad_norm": 1.1250585317611694,
+      "learning_rate": 1.9539748953974895e-05,
+      "loss": 0.2292,
+      "step": 250
+    },
+    {
+      "epoch": 1.1506276150627615,
+      "grad_norm": 4.790164947509766,
+      "learning_rate": 1.8493723849372386e-05,
+      "loss": 0.1835,
+      "step": 275
+    },
+    {
+      "epoch": 1.2552301255230125,
+      "grad_norm": 6.640511989593506,
+      "learning_rate": 1.7447698744769877e-05,
+      "loss": 0.2212,
+      "step": 300
+    },
+    {
+      "epoch": 1.3598326359832635,
+      "grad_norm": 7.408748149871826,
+      "learning_rate": 1.6401673640167365e-05,
+      "loss": 0.1845,
+      "step": 325
+    },
+    {
+      "epoch": 1.4644351464435146,
+      "grad_norm": 3.5393199920654297,
+      "learning_rate": 1.5355648535564852e-05,
+      "loss": 0.1667,
+      "step": 350
+    },
+    {
+      "epoch": 1.5690376569037658,
+      "grad_norm": 4.460846424102783,
+      "learning_rate": 1.4309623430962344e-05,
+      "loss": 0.1983,
+      "step": 375
+    },
+    {
+      "epoch": 1.6736401673640167,
+      "grad_norm": 3.503453493118286,
+      "learning_rate": 1.3263598326359832e-05,
+      "loss": 0.1726,
+      "step": 400
+    },
+    {
+      "epoch": 1.778242677824268,
+      "grad_norm": 2.8081846237182617,
+      "learning_rate": 1.2217573221757323e-05,
+      "loss": 0.2135,
+      "step": 425
+    },
+    {
+      "epoch": 1.8828451882845187,
+      "grad_norm": 4.33410120010376,
+      "learning_rate": 1.1171548117154812e-05,
+      "loss": 0.1675,
+      "step": 450
+    },
+    {
+      "epoch": 1.98744769874477,
+      "grad_norm": 3.7940282821655273,
+      "learning_rate": 1.0125523012552301e-05,
+      "loss": 0.1761,
+      "step": 475
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9336702463676564,
+      "eval_loss": 0.20952366292476654,
+      "eval_runtime": 5.0796,
+      "eval_samples_per_second": 311.638,
+      "eval_steps_per_second": 3.347,
+      "step": 478
+    },
+    {
+      "epoch": 2.092050209205021,
+      "grad_norm": 5.969915390014648,
+      "learning_rate": 9.079497907949792e-06,
+      "loss": 0.1174,
+      "step": 500
+    },
+    {
+      "epoch": 2.196652719665272,
+      "grad_norm": 8.860109329223633,
+      "learning_rate": 8.03347280334728e-06,
+      "loss": 0.1213,
+      "step": 525
+    },
+    {
+      "epoch": 2.301255230125523,
+      "grad_norm": 1.0917476415634155,
+      "learning_rate": 6.98744769874477e-06,
+      "loss": 0.14,
+      "step": 550
+    },
+    {
+      "epoch": 2.405857740585774,
+      "grad_norm": 3.1612932682037354,
+      "learning_rate": 5.941422594142259e-06,
+      "loss": 0.13,
+      "step": 575
+    },
+    {
+      "epoch": 2.510460251046025,
+      "grad_norm": 3.376967430114746,
+      "learning_rate": 4.895397489539749e-06,
+      "loss": 0.135,
+      "step": 600
+    },
+    {
+      "epoch": 2.6150627615062763,
+      "grad_norm": 4.779818534851074,
+      "learning_rate": 3.849372384937238e-06,
+      "loss": 0.1299,
+      "step": 625
+    },
+    {
+      "epoch": 2.719665271966527,
+      "grad_norm": 0.9145883321762085,
+      "learning_rate": 2.803347280334728e-06,
+      "loss": 0.1277,
+      "step": 650
+    },
+    {
+      "epoch": 2.8242677824267783,
+      "grad_norm": 4.260244369506836,
+      "learning_rate": 1.7573221757322176e-06,
+      "loss": 0.1337,
+      "step": 675
+    },
+    {
+      "epoch": 2.928870292887029,
+      "grad_norm": 2.3931667804718018,
+      "learning_rate": 7.112970711297071e-07,
+      "loss": 0.1308,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9355653821857233,
+      "eval_loss": 0.22262658178806305,
+      "eval_runtime": 4.9595,
+      "eval_samples_per_second": 319.185,
+      "eval_steps_per_second": 3.428,
+      "step": 717
+    },
+    {
+      "epoch": 3.0,
+      "step": 717,
+      "total_flos": 1055522845962000.0,
+      "train_loss": 0.20581150786480976,
+      "train_runtime": 66.9935,
+      "train_samples_per_second": 511.662,
+      "train_steps_per_second": 10.703
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 717,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1055522845962000.0,
+  "train_batch_size": 48,
+  "trial_name": null,
+  "trial_params": null
+}