Training in progress, step 45000, checkpoint

checkpoint-45000/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

checkpoint-45000/config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "demdecuong/vihealthbert-base-word",
+  "architectures": [
+    "RobertaForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "finetuning_task": "word-level",
+  "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": 258,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_hidden_states": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "tokenizer_class": "PhobertTokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.40.2",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 64001
+}

checkpoint-45000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:335a27b46ad4d7ed7a27a8e1e2eabec97937266036b74297507161923c7964c0
+size 540279452

checkpoint-45000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bffb6bdd99be0975f96f64f471ab72492d25e32c8ff700f4dd176ad366c58878
+size 1080630853

checkpoint-45000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93eb177aacaf5e88ea83e84acade070baf68fdf3ee9d6e2333ea4d351eeccf50
+size 14575

checkpoint-45000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da8f0623d7d3ce2de798703bd240096931cc9a22433b3ce1dd1c05d08533ccf5
+size 627

checkpoint-45000/special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

checkpoint-45000/tokenizer_config.json

@@ -0,0 +1,54 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}

checkpoint-45000/trainer_state.json

@@ -0,0 +1,172 @@
+{
+  "best_metric": 0.681425908452811,
+  "best_model_checkpoint": "output/pretraining/vihealthbert-w_unsup-SynPD/lr3e-5_wr0.1_wd0.0/checkpoint-40000",
+  "epoch": 7.754609684645873,
+  "eval_steps": 5000,
+  "global_step": 45000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017232465965879716,
+      "grad_norm": 41.827266693115234,
+      "learning_rate": 5.169739789763915e-09,
+      "loss": 13.7242,
+      "step": 1
+    },
+    {
+      "epoch": 0.8616232982939859,
+      "grad_norm": 5.6151957511901855,
+      "learning_rate": 2.584869894881958e-05,
+      "loss": 7.0234,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8616232982939859,
+      "eval_accuracy": 0.5576033171446898,
+      "eval_loss": 2.590928077697754,
+      "eval_runtime": 37.589,
+      "eval_samples_per_second": 260.023,
+      "eval_steps_per_second": 16.255,
+      "step": 5000
+    },
+    {
+      "epoch": 1.7232465965879717,
+      "grad_norm": 4.410743713378906,
+      "learning_rate": 2.758917801137343e-05,
+      "loss": 5.2736,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7232465965879717,
+      "eval_accuracy": 0.596153624823046,
+      "eval_loss": 2.188957452774048,
+      "eval_runtime": 37.6297,
+      "eval_samples_per_second": 259.741,
+      "eval_steps_per_second": 16.237,
+      "step": 10000
+    },
+    {
+      "epoch": 2.5848698948819577,
+      "grad_norm": 4.754213809967041,
+      "learning_rate": 2.4717100350393475e-05,
+      "loss": 4.9126,
+      "step": 15000
+    },
+    {
+      "epoch": 2.5848698948819577,
+      "eval_accuracy": 0.6381037526075694,
+      "eval_loss": 1.909491777420044,
+      "eval_runtime": 36.8556,
+      "eval_samples_per_second": 265.197,
+      "eval_steps_per_second": 16.578,
+      "step": 15000
+    },
+    {
+      "epoch": 3.4464931931759435,
+      "grad_norm": 4.2047200202941895,
+      "learning_rate": 2.184502268941352e-05,
+      "loss": 4.791,
+      "step": 20000
+    },
+    {
+      "epoch": 3.4464931931759435,
+      "eval_accuracy": 0.6468589272593681,
+      "eval_loss": 1.8286069631576538,
+      "eval_runtime": 35.9342,
+      "eval_samples_per_second": 271.997,
+      "eval_steps_per_second": 17.003,
+      "step": 20000
+    },
+    {
+      "epoch": 4.30811649146993,
+      "grad_norm": 4.629519462585449,
+      "learning_rate": 1.897294502843357e-05,
+      "loss": 4.6538,
+      "step": 25000
+    },
+    {
+      "epoch": 4.30811649146993,
+      "eval_accuracy": 0.6644416567883901,
+      "eval_loss": 1.714424967765808,
+      "eval_runtime": 35.6122,
+      "eval_samples_per_second": 274.456,
+      "eval_steps_per_second": 17.157,
+      "step": 25000
+    },
+    {
+      "epoch": 5.1697397897639155,
+      "grad_norm": 3.993448257446289,
+      "learning_rate": 1.6100867367453616e-05,
+      "loss": 4.5846,
+      "step": 30000
+    },
+    {
+      "epoch": 5.1697397897639155,
+      "eval_accuracy": 0.6703549228453575,
+      "eval_loss": 1.6779303550720215,
+      "eval_runtime": 36.1572,
+      "eval_samples_per_second": 270.32,
+      "eval_steps_per_second": 16.898,
+      "step": 30000
+    },
+    {
+      "epoch": 6.031363088057901,
+      "grad_norm": 4.6959357261657715,
+      "learning_rate": 1.3228789706473663e-05,
+      "loss": 4.5568,
+      "step": 35000
+    },
+    {
+      "epoch": 6.031363088057901,
+      "eval_accuracy": 0.6765665494261385,
+      "eval_loss": 1.6362268924713135,
+      "eval_runtime": 36.2906,
+      "eval_samples_per_second": 269.326,
+      "eval_steps_per_second": 16.836,
+      "step": 35000
+    },
+    {
+      "epoch": 6.892986386351887,
+      "grad_norm": 4.4082441329956055,
+      "learning_rate": 1.035671204549371e-05,
+      "loss": 4.5079,
+      "step": 40000
+    },
+    {
+      "epoch": 6.892986386351887,
+      "eval_accuracy": 0.681425908452811,
+      "eval_loss": 1.6008453369140625,
+      "eval_runtime": 36.1284,
+      "eval_samples_per_second": 270.535,
+      "eval_steps_per_second": 16.912,
+      "step": 40000
+    },
+    {
+      "epoch": 7.754609684645873,
+      "grad_norm": 5.832913398742676,
+      "learning_rate": 7.484634384513757e-06,
+      "loss": 4.469,
+      "step": 45000
+    },
+    {
+      "epoch": 7.754609684645873,
+      "eval_accuracy": 0.6805218690729352,
+      "eval_loss": 1.6063588857650757,
+      "eval_runtime": 36.0752,
+      "eval_samples_per_second": 270.934,
+      "eval_steps_per_second": 16.937,
+      "step": 45000
+    }
+  ],
+  "logging_steps": 5000,
+  "max_steps": 58030,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 5000,
+  "total_flos": 0.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-45000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2690ff9d34fa5fa3d550f183e187fe6edb89806dc8d5d196cb79437ad79b8f1
+size 4859