First version of the bertweet-base-multi-mami model and tokenizer

added_tokens.json

@@ -0,0 +1 @@
+{"<mask>": 64000}

config.json

@@ -0,0 +1,44 @@
+{
+  "_name_or_path": "vinai/bertweet-base",
+  "architectures": [
+    "RobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2",
+    "3": "LABEL_3",
+    "4": "LABEL_4"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "misogynous": 0,
+    "objectification": 3,
+    "shaming": 1,
+    "stereotype": 2,
+    "violence": 4
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 130,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "problem_type": "multi_label_classification",
+  "tokenizer_class": "BertweetTokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.12.5",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 64001
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7223766efff8fe830336541ae02e96beba37d594716fbcde42efaf7558ea4b6b
+size 539697581

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b19671e619c18d584d730f375e39269329073b67ca11c0f8071069413f5dabd
+size 14503

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>"}

test_results.txt

@@ -0,0 +1 @@
+{"test_loss": 1.3586835861206055, "test_accuracy": 0.4555, "test_f1_macro_avg": 0.5502965174823465, "test_f1_macro_weighted": 0.6266297015316841, "test_f1_samples": 0.3475138888888889, "test_runtime": 6.3833, "test_samples_per_second": 313.319, "test_steps_per_second": 39.165}

tokenizer_config.json

@@ -0,0 +1 @@
+{"normalization": false, "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "model_max_length": 128, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "vinai/bertweet-base", "tokenizer_class": "BertweetTokenizer"}

trainer_state.json

@@ -0,0 +1,52 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "global_step": 800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 400
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.515,
+      "eval_f1_macro_avg": 0.5601429722987724,
+      "eval_f1_macro_weighted": 0.6377567484003918,
+      "eval_f1_samples": 0.32541716269841264,
+      "eval_loss": 1.1126549243927002,
+      "eval_runtime": 4.9376,
+      "eval_samples_per_second": 324.041,
+      "eval_steps_per_second": 20.253,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 800
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.464375,
+      "eval_f1_macro_avg": 0.5785179362668098,
+      "eval_f1_macro_weighted": 0.649904719860711,
+      "eval_f1_samples": 0.35723908730158727,
+      "eval_loss": 1.280062198638916,
+      "eval_runtime": 4.9492,
+      "eval_samples_per_second": 323.283,
+      "eval_steps_per_second": 20.205,
+      "step": 800
+    }
+  ],
+  "max_steps": 2400,
+  "num_train_epochs": 6,
+  "total_flos": 841978055884800.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71d56f5b5a892196d7183bb6b8c6f2c81c662824283d890e64496f99e38ccc7a
+size 2799

val_results.txt

@@ -0,0 +1 @@
+{"val_loss": 1.280062198638916, "val_accuracy": 0.464375, "val_f1_macro_avg": 0.5785179362668098, "val_f1_macro_weighted": 0.649904719860711, "val_f1_samples": 0.35723908730158727, "val_runtime": 5.4173, "val_samples_per_second": 295.349, "val_steps_per_second": 36.919}