adding model fine-tuned on QA (SQAC)

all_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 4.0,
+    "eval_exact_match": 63.197424892703864,
+    "eval_f1": 80.27730664434178,
+    "eval_samples": 2113,
+    "train_loss": 0.6520571581522624,
+    "train_runtime": 5114.6805,
+    "train_samples": 16790,
+    "train_samples_per_second": 13.131,
+    "train_steps_per_second": 0.821
+}

config.json

@@ -0,0 +1,27 @@
+{
+  "_name_or_path": "dccuchile/bert-base-spanish-wwm-cased",
+  "architectures": [
+    "BertForQuestionAnswering"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "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-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.15.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 31002
+}

eval_results.json

@@ -0,0 +1,6 @@
+{
+    "epoch": 4.0,
+    "eval_exact_match": 63.197424892703864,
+    "eval_f1": 80.27730664434178,
+    "eval_samples": 2113
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0794f1ea48b92b81648c01e5b097fd9efb248156dc75d2315b6dafb78a1581bb
+size 437130673

special_tokens_map.json

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

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": false, "name_or_path": "dccuchile/bert-base-spanish-wwm-cased", "do_basic_tokenize": true, "never_split": null, "model_max_length": 512, "special_tokens_map_file": "/data/jcanete/cache/9848a00af462c42dfb4ec88ef438fbab5256330f7f6f50badc48d277f9367d49.f982506b52498d4adb4bd491f593dc92b2ef6be61bfdbe9d30f53f963f9f5b66", "tokenizer_class": "BertTokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.6520571581522624,
+    "train_runtime": 5114.6805,
+    "train_samples": 16790,
+    "train_samples_per_second": 13.131,
+    "train_steps_per_second": 0.821
+}

trainer_state.json

@@ -0,0 +1,199 @@
+{
+  "best_metric": 80.27730664434178,
+  "best_model_checkpoint": "/data/jcanete/all_results/sqac/beto_cased/epochs_4_bs_16_lr_3e-5/checkpoint-4200",
+  "epoch": 4.0,
+  "global_step": 4200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.19,
+      "eval_exact_match": 47.96137339055794,
+      "eval_f1": 68.6556986249518,
+      "step": 200
+    },
+    {
+      "epoch": 0.38,
+      "eval_exact_match": 55.09656652360515,
+      "eval_f1": 73.8086956650881,
+      "step": 400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6442857142857144e-05,
+      "loss": 1.8427,
+      "step": 500
+    },
+    {
+      "epoch": 0.57,
+      "eval_exact_match": 55.847639484978544,
+      "eval_f1": 74.94671429876136,
+      "step": 600
+    },
+    {
+      "epoch": 0.76,
+      "eval_exact_match": 57.67167381974249,
+      "eval_f1": 76.41878566636721,
+      "step": 800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.287142857142857e-05,
+      "loss": 1.1475,
+      "step": 1000
+    },
+    {
+      "epoch": 0.95,
+      "eval_exact_match": 58.208154506437765,
+      "eval_f1": 77.16784556628494,
+      "step": 1000
+    },
+    {
+      "epoch": 1.14,
+      "eval_exact_match": 60.30042918454936,
+      "eval_f1": 78.11145340783891,
+      "step": 1200
+    },
+    {
+      "epoch": 1.33,
+      "eval_exact_match": 60.89055793991416,
+      "eval_f1": 78.59391938565095,
+      "step": 1400
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.9299999999999998e-05,
+      "loss": 0.685,
+      "step": 1500
+    },
+    {
+      "epoch": 1.52,
+      "eval_exact_match": 61.96351931330472,
+      "eval_f1": 79.0103467396917,
+      "step": 1600
+    },
+    {
+      "epoch": 1.71,
+      "eval_exact_match": 60.622317596566525,
+      "eval_f1": 78.40353020710157,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.572857142857143e-05,
+      "loss": 0.6531,
+      "step": 2000
+    },
+    {
+      "epoch": 1.9,
+      "eval_exact_match": 61.5343347639485,
+      "eval_f1": 79.60513374898902,
+      "step": 2000
+    },
+    {
+      "epoch": 2.1,
+      "eval_exact_match": 61.802575107296136,
+      "eval_f1": 79.18672103994875,
+      "step": 2200
+    },
+    {
+      "epoch": 2.29,
+      "eval_exact_match": 62.66094420600859,
+      "eval_f1": 80.0520536342908,
+      "step": 2400
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.2157142857142857e-05,
+      "loss": 0.38,
+      "step": 2500
+    },
+    {
+      "epoch": 2.48,
+      "eval_exact_match": 62.92918454935622,
+      "eval_f1": 79.925554389572,
+      "step": 2600
+    },
+    {
+      "epoch": 2.67,
+      "eval_exact_match": 62.55364806866953,
+      "eval_f1": 79.90621692100268,
+      "step": 2800
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.592857142857142e-06,
+      "loss": 0.3225,
+      "step": 3000
+    },
+    {
+      "epoch": 2.86,
+      "eval_exact_match": 62.76824034334764,
+      "eval_f1": 79.51326094896278,
+      "step": 3000
+    },
+    {
+      "epoch": 3.05,
+      "eval_exact_match": 62.1244635193133,
+      "eval_f1": 79.6112386410588,
+      "step": 3200
+    },
+    {
+      "epoch": 3.24,
+      "eval_exact_match": 62.82188841201717,
+      "eval_f1": 80.13849915024434,
+      "step": 3400
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5.0285714285714285e-06,
+      "loss": 0.2161,
+      "step": 3500
+    },
+    {
+      "epoch": 3.43,
+      "eval_exact_match": 63.09012875536481,
+      "eval_f1": 80.22707827276506,
+      "step": 3600
+    },
+    {
+      "epoch": 3.62,
+      "eval_exact_match": 63.14377682403433,
+      "eval_f1": 80.25570792364071,
+      "step": 3800
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.4571428571428571e-06,
+      "loss": 0.1661,
+      "step": 4000
+    },
+    {
+      "epoch": 3.81,
+      "eval_exact_match": 63.03648068669528,
+      "eval_f1": 80.23874566810112,
+      "step": 4000
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 63.197424892703864,
+      "eval_f1": 80.27730664434178,
+      "step": 4200
+    },
+    {
+      "epoch": 4.0,
+      "step": 4200,
+      "total_flos": 1.721582261087904e+16,
+      "train_loss": 0.6520571581522624,
+      "train_runtime": 5114.6805,
+      "train_samples_per_second": 13.131,
+      "train_steps_per_second": 0.821
+    }
+  ],
+  "max_steps": 4200,
+  "num_train_epochs": 4,
+  "total_flos": 1.721582261087904e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae07346d785d0aa0c69405a6679cdba02a48ff274f0b5bc97783596befab23c6
+size 2991