Training in progress, epoch 1

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b93dca996de1e2a65bd2b8805ca12671d565a8759726dc118ad227d9b8a09a73
+oid sha256:32f2b6344dbeefedd77c7295f5f43e3c60235bc524019549e354574302929153
 size 267829484

run-2/checkpoint-72/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:249941c828507120d54adf6c5c3e3fcb54092c15aabd84dbcf3cb5310dc37911
+oid sha256:29762bedee45da925779652901931aaaa997ceed3f77ded17a0fc6f3db058cb0
 size 267829484

run-2/checkpoint-72/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77633fbec22f842c94d0497b7fc329bd24ed1768be551c33e31789f8f18c0dc3
+oid sha256:30e04af797359d85d3bcbd20632d6df2c1f4ec63907848ef6c461754000226da
 size 535721146

run-2/checkpoint-72/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c47b999f373d02779993f31b8f70bc6d254ebaa78c7cb6aceff9aa39b2c0fd53
+oid sha256:4686679b5492922b658249d2e94e6f498b978fdd3a7bff4a48680a5cd459c4af
 size 14308

run-2/checkpoint-72/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:deab0b4d6a1efc5852cfd2b718d2f3d57a531cfe61c04096b0c43021400997ea
+oid sha256:bde2e00598c06b03eeecccf4acec830a76c4f4fee8368c10b212531b829e48b1
 size 1064

run-2/checkpoint-72/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.774068032504475,
+  "best_metric": 0.4052420655597578,
   "best_model_checkpoint": "distilbert-base-uncased-finetuned-stsb/run-2/checkpoint-72",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,37 +10,37 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 2.4969136714935303,
-      "eval_pearson": 0.32223422122729983,
-      "eval_runtime": 0.9052,
-      "eval_samples_per_second": 1657.071,
-      "eval_spearmanr": 0.32647520010278785,
-      "eval_steps_per_second": 103.843,
+      "eval_loss": 2.3542871475219727,
+      "eval_pearson": 0.22535612418664588,
+      "eval_runtime": 0.7946,
+      "eval_samples_per_second": 1887.683,
+      "eval_spearmanr": 0.2439162232725169,
+      "eval_steps_per_second": 118.295,
       "step": 36
     },
     {
       "epoch": 2.0,
-      "eval_loss": 1.1902464628219604,
-      "eval_pearson": 0.774068032504475,
-      "eval_runtime": 0.9073,
-      "eval_samples_per_second": 1653.256,
-      "eval_spearmanr": 0.7685396079860547,
-      "eval_steps_per_second": 103.604,
+      "eval_loss": 2.184481382369995,
+      "eval_pearson": 0.4052420655597578,
+      "eval_runtime": 0.8115,
+      "eval_samples_per_second": 1848.458,
+      "eval_spearmanr": 0.4375008049819036,
+      "eval_steps_per_second": 115.837,
       "step": 72
     }
   ],
   "logging_steps": 500,
-  "max_steps": 180,
+  "max_steps": 72,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "total_flos": 0,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": {
-    "learning_rate": 2.5099285971844188e-05,
-    "num_train_epochs": 5,
+    "learning_rate": 1.2299344669970108e-05,
+    "num_train_epochs": 2,
     "per_device_train_batch_size": 16,
-    "seed": 8
+    "seed": 20
   }
 }

run-2/checkpoint-72/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12643df0525ba3f2bf5dee245df6a74c3697e1dda56998a9dd617f86119470b4
+oid sha256:51b688100c0cb9752c8b835323e5f372a383d39c54fd09c2eb9daa32a890cbaf
 size 4920

run-3/checkpoint-144/config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "distilbert-base-uncased",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "LABEL_0"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "LABEL_0": 0
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "pad_token_id": 0,
+  "problem_type": "regression",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.38.2",
+  "vocab_size": 30522
+}

run-3/checkpoint-144/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32f2b6344dbeefedd77c7295f5f43e3c60235bc524019549e354574302929153
+size 267829484

run-3/checkpoint-144/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dce180cf84c457e26dbe5dc71e2cf60b9fb73806b95e73e9436ad0a1db34b15c
+size 535721146

run-3/checkpoint-144/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dea8c3947bb8f5869ac25773574fc8e420bd3520749b6b82f0981d4e5526f48
+size 14244

run-3/checkpoint-144/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38b6a327e093676a8bb58cd5a3fbc152f4e0b61b1785248363a0da24141a066b
+size 1064

run-3/checkpoint-144/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

run-3/checkpoint-144/tokenizer_config.json

@@ -0,0 +1,55 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "DistilBertTokenizer",
+  "unk_token": "[UNK]"
+}

run-3/checkpoint-144/trainer_state.json

@@ -0,0 +1,36 @@
+{
+  "best_metric": 0.7678320454774757,
+  "best_model_checkpoint": "distilbert-base-uncased-finetuned-stsb/run-3/checkpoint-144",
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 144,
+  "is_hyper_param_search": true,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9698190689086914,
+      "eval_pearson": 0.7678320454774757,
+      "eval_runtime": 0.8035,
+      "eval_samples_per_second": 1866.77,
+      "eval_spearmanr": 0.712509933248186,
+      "eval_steps_per_second": 116.984,
+      "step": 144
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 720,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": {
+    "learning_rate": 5.4134010471541346e-05,
+    "num_train_epochs": 5,
+    "per_device_train_batch_size": 4,
+    "seed": 4
+  }
+}

run-3/checkpoint-144/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a89dbe21e169d083920168fe1c35f86c15d94a0c3b9ca2615dabe1e0930b053d
+size 4920

runs/Mar06_19-02-43_8a0d46b883ab/events.out.tfevents.1709752114.8a0d46b883ab.2624.5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a00b46844c0a4135284537aaa87757525be395c2447fa52e6789880c36267f1
+size 7046

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:51b688100c0cb9752c8b835323e5f372a383d39c54fd09c2eb9daa32a890cbaf
+oid sha256:a89dbe21e169d083920168fe1c35f86c15d94a0c3b9ca2615dabe1e0930b053d
 size 4920