fintune bibert on iwslt14 ende with mlm task

README.md

@@ -1,3 +1,58 @@
 ---
-license: openrail
+tags:
+- generated_from_trainer
+metrics:
+- accuracy
+model-index:
+- name: bibert-ende
+  results: []
 ---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# bibert-ende
+
+This model is a fine-tuned version of [jhu-clsp/bibert-ende](https://huggingface.co/jhu-clsp/bibert-ende) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.8713
+- Accuracy: 0.6310
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-05
+- train_batch_size: 32
+- eval_batch_size: 32
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 128
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 3.0
+- mixed_precision_training: Native AMP
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.28.0.dev0
+- Pytorch 1.12.0+cu113
+- Datasets 2.10.1
+- Tokenizers 0.13.2

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.631046900996146,
+    "eval_loss": 1.871326208114624,
+    "eval_runtime": 14.0466,
+    "eval_samples": 14566,
+    "eval_samples_per_second": 1036.979,
+    "eval_steps_per_second": 32.463,
+    "perplexity": 6.496906939043341,
+    "train_loss": 1.378044397486057,
+    "train_runtime": 6661.4989,
+    "train_samples": 640956,
+    "train_samples_per_second": 288.654,
+    "train_steps_per_second": 2.255
+}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "jhu-clsp/bibert-ende",
+  "architectures": [
+    "RobertaForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 2,
+  "classifier_dropout": null,
+  "eos_token_id": 3,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 513,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.0.dev0",
+  "type_vocab_size": 2,
+  "unk_token_id": 1,
+  "use_cache": true,
+  "vocab_size": 52000
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.631046900996146,
+    "eval_loss": 1.871326208114624,
+    "eval_runtime": 14.0466,
+    "eval_samples": 14566,
+    "eval_samples_per_second": 1036.979,
+    "eval_steps_per_second": 32.463,
+    "perplexity": 6.496906939043341
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:916b8491f6341336eecde8e2bb78bf0db51ba47d063a9b65e05d8f3af50d982f
+size 504198571

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1,17 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "tokenizer_file": null,
+  "unk_token": "[UNK]"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 1.378044397486057,
+    "train_runtime": 6661.4989,
+    "train_samples": 640956,
+    "train_samples_per_second": 288.654,
+    "train_steps_per_second": 2.255
+}

trainer_state.json

@@ -0,0 +1,205 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999700449326011,
+  "global_step": 15021,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8338992077757805e-05,
+      "loss": 2.2639,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.667465548232475e-05,
+      "loss": 1.7816,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.501031888689169e-05,
+      "loss": 1.6597,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.334598229145863e-05,
+      "loss": 1.5932,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.168164569602557e-05,
+      "loss": 1.5107,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0017309100592506e-05,
+      "loss": 1.4919,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8352972505159445e-05,
+      "loss": 1.4724,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6688635909726384e-05,
+      "loss": 1.43,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.502429931429332e-05,
+      "loss": 1.4131,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.336329139205113e-05,
+      "loss": 1.3983,
+      "step": 5000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.169895479661807e-05,
+      "loss": 1.3639,
+      "step": 5500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.003461820118501e-05,
+      "loss": 1.3449,
+      "step": 6000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8370281605751946e-05,
+      "loss": 1.3259,
+      "step": 6500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6709273683509755e-05,
+      "loss": 1.3239,
+      "step": 7000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.504493708807669e-05,
+      "loss": 1.3131,
+      "step": 7500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3380600492643633e-05,
+      "loss": 1.3052,
+      "step": 8000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1716263897210572e-05,
+      "loss": 1.2947,
+      "step": 8500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0058584648159245e-05,
+      "loss": 1.2927,
+      "step": 9000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8394248052726184e-05,
+      "loss": 1.2736,
+      "step": 9500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6729911457293123e-05,
+      "loss": 1.265,
+      "step": 10000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5065574861860063e-05,
+      "loss": 1.2467,
+      "step": 10500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3401238266427002e-05,
+      "loss": 1.2387,
+      "step": 11000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1736901670993943e-05,
+      "loss": 1.2283,
+      "step": 11500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0072565075560882e-05,
+      "loss": 1.2271,
+      "step": 12000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.408228480127821e-06,
+      "loss": 1.2382,
+      "step": 12500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.7438918846947615e-06,
+      "loss": 1.2197,
+      "step": 13000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.0795552892617005e-06,
+      "loss": 1.217,
+      "step": 13500
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.4152186938286403e-06,
+      "loss": 1.2096,
+      "step": 14000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7508820983955797e-06,
+      "loss": 1.2048,
+      "step": 14500
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.654550296251915e-08,
+      "loss": 1.2008,
+      "step": 15000
+    },
+    {
+      "epoch": 3.0,
+      "step": 15021,
+      "total_flos": 1.1987319801621504e+17,
+      "train_loss": 1.378044397486057,
+      "train_runtime": 6661.4989,
+      "train_samples_per_second": 288.654,
+      "train_steps_per_second": 2.255
+    }
+  ],
+  "max_steps": 15021,
+  "num_train_epochs": 3,
+  "total_flos": 1.1987319801621504e+17,
+  "trial_name": null,
+  "trial_params": null
+}