adding model fine-tuned on QA (SQAC)

all_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 3.0,
+    "eval_exact_match": 66.1480686695279,
+    "eval_f1": 83.11268652174739,
+    "eval_samples": 2082,
+    "train_loss": 0.6812070615029748,
+    "train_runtime": 1785.0892,
+    "train_samples": 16598,
+    "train_samples_per_second": 27.894,
+    "train_steps_per_second": 1.744
+}

config.json

@@ -0,0 +1,34 @@
+{
+  "_name_or_path": "CenIA/albert_xxlarge_spanish",
+  "architectures": [
+    "AlbertForQuestionAnswering"
+  ],
+  "attention_probs_dropout_prob": 0,
+  "bos_token_id": 2,
+  "classifier_dropout_prob": 0.1,
+  "down_scale_factor": 1,
+  "embedding_size": 128,
+  "eos_token_id": 3,
+  "gap_size": 0,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0,
+  "hidden_size": 4096,
+  "initializer_range": 0.01,
+  "inner_group_num": 1,
+  "intermediate_size": 16384,
+  "layer_norm_eps": 1e-12,
+  "layers_to_keep": [],
+  "max_position_embeddings": 512,
+  "model_type": "albert",
+  "net_structure_type": 0,
+  "num_attention_heads": 64,
+  "num_hidden_groups": 1,
+  "num_hidden_layers": 12,
+  "num_memory_blocks": 0,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.15.0",
+  "type_vocab_size": 2,
+  "vocab_size": 31000
+}

eval_results.json

@@ -0,0 +1,6 @@
+{
+    "epoch": 3.0,
+    "eval_exact_match": 66.1480686695279,
+    "eval_f1": 83.11268652174739,
+    "eval_samples": 2082
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a99168ba2347656f358ce13c2b41754dd15722efc5c26f095dc99d8ea0887efd
+size 823816087

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "<unk>", "sep_token": "[SEP]", "pad_token": "<pad>", "cls_token": "[CLS]", "mask_token": "[MASK]"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": true, "remove_space": true, "keep_accents": true, "bos_token": "[CLS]", "eos_token": "[SEP]", "unk_token": "<unk>", "sep_token": "[SEP]", "pad_token": "<pad>", "cls_token": "[CLS]", "mask_token": {"content": "[MASK]", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "CenIA/albert_xxlarge_spanish", "tokenizer_class": "AlbertTokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.6812070615029748,
+    "train_runtime": 1785.0892,
+    "train_samples": 16598,
+    "train_samples_per_second": 27.894,
+    "train_steps_per_second": 1.744
+}

trainer_state.json

@@ -0,0 +1,151 @@
+{
+  "best_metric": 83.11268652174739,
+  "best_model_checkpoint": "/home/fbravo/data/all_results/sqac/albeto_xxlarge/epochs_3_bs_16_lr_5e-6/checkpoint-2000",
+  "epoch": 3.0,
+  "global_step": 3114,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.19,
+      "eval_exact_match": 50.21459227467811,
+      "eval_f1": 72.15606343917293,
+      "step": 200
+    },
+    {
+      "epoch": 0.39,
+      "eval_exact_match": 57.8862660944206,
+      "eval_f1": 77.27610294693481,
+      "step": 400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.205202312138729e-06,
+      "loss": 1.7198,
+      "step": 500
+    },
+    {
+      "epoch": 0.58,
+      "eval_exact_match": 61.26609442060086,
+      "eval_f1": 79.60226388707036,
+      "step": 600
+    },
+    {
+      "epoch": 0.77,
+      "eval_exact_match": 60.56866952789699,
+      "eval_f1": 80.21878308712405,
+      "step": 800
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.402376364804111e-06,
+      "loss": 0.9328,
+      "step": 1000
+    },
+    {
+      "epoch": 0.96,
+      "eval_exact_match": 64.10944206008584,
+      "eval_f1": 81.45706053754708,
+      "step": 1000
+    },
+    {
+      "epoch": 1.16,
+      "eval_exact_match": 64.431330472103,
+      "eval_f1": 82.10448987052378,
+      "step": 1200
+    },
+    {
+      "epoch": 1.35,
+      "eval_exact_match": 65.93347639484979,
+      "eval_f1": 82.46788441647152,
+      "step": 1400
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.599550417469493e-06,
+      "loss": 0.5496,
+      "step": 1500
+    },
+    {
+      "epoch": 1.54,
+      "eval_exact_match": 65.50429184549357,
+      "eval_f1": 82.69463199305457,
+      "step": 1600
+    },
+    {
+      "epoch": 1.73,
+      "eval_exact_match": 65.50429184549357,
+      "eval_f1": 82.47924185410717,
+      "step": 1800
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.798330122029544e-06,
+      "loss": 0.5137,
+      "step": 2000
+    },
+    {
+      "epoch": 1.93,
+      "eval_exact_match": 66.1480686695279,
+      "eval_f1": 83.11268652174739,
+      "step": 2000
+    },
+    {
+      "epoch": 2.12,
+      "eval_exact_match": 66.25536480686695,
+      "eval_f1": 82.81247091268429,
+      "step": 2200
+    },
+    {
+      "epoch": 2.31,
+      "eval_exact_match": 65.98712446351931,
+      "eval_f1": 82.52306189501985,
+      "step": 2400
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.955041746949263e-07,
+      "loss": 0.2552,
+      "step": 2500
+    },
+    {
+      "epoch": 2.5,
+      "eval_exact_match": 66.04077253218884,
+      "eval_f1": 82.98092243697893,
+      "step": 2600
+    },
+    {
+      "epoch": 2.7,
+      "eval_exact_match": 66.46995708154506,
+      "eval_f1": 82.55808756053281,
+      "step": 2800
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9267822736030828e-07,
+      "loss": 0.2252,
+      "step": 3000
+    },
+    {
+      "epoch": 2.89,
+      "eval_exact_match": 66.25536480686695,
+      "eval_f1": 82.8710129087105,
+      "step": 3000
+    },
+    {
+      "epoch": 3.0,
+      "step": 3114,
+      "total_flos": 2.835542958722253e+16,
+      "train_loss": 0.6812070615029748,
+      "train_runtime": 1785.0892,
+      "train_samples_per_second": 27.894,
+      "train_steps_per_second": 1.744
+    }
+  ],
+  "max_steps": 3114,
+  "num_train_epochs": 3,
+  "total_flos": 2.835542958722253e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:670db718f3477752e72914ecb8798d506f10a8d58a6329132e724ea4ef0a31c3
+size 2991