adding model finetuned on PAWS-X

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.809499979019165,
+    "eval_loss": 0.42281684279441833,
+    "eval_runtime": 0.5205,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 3842.632,
+    "eval_steps_per_second": 240.165,
+    "train_loss": 0.24578420763806358,
+    "train_runtime": 1895.6753,
+    "train_samples": 49401,
+    "train_samples_per_second": 104.239,
+    "train_steps_per_second": 6.516
+}

config.json

@@ -0,0 +1,34 @@
+{
+  "_name_or_path": "CenIA/albert_tiny_spanish",
+  "architectures": [
+    "AlbertForSequenceClassification"
+  ],
+  "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": 312,
+  "initializer_range": 0.02,
+  "inner_group_num": 1,
+  "intermediate_size": 1248,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "albert",
+  "net_structure_type": 0,
+  "num_attention_heads": 12,
+  "num_hidden_groups": 1,
+  "num_hidden_layers": 4,
+  "num_memory_blocks": 0,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.12.5",
+  "type_vocab_size": 2,
+  "vocab_size": 31000
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.809499979019165,
+    "eval_loss": 0.42281684279441833,
+    "eval_runtime": 0.5205,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 3842.632,
+    "eval_steps_per_second": 240.165
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8e72f236fd70a343cc55abe6e3b8f1de6073b85f9de8c82c4b24fd96b8b59a9
+size 21393937

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_tiny_spanish", "tokenizer_class": "AlbertTokenizer"}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.24578420763806358,
+    "train_runtime": 1895.6753,
+    "train_samples": 49401,
+    "train_samples_per_second": 104.239,
+    "train_steps_per_second": 6.516
+}

trainer_state.json

@@ -0,0 +1,538 @@
+{
+  "best_metric": 0.42281684279441833,
+  "best_model_checkpoint": "/data/jcanete/all_results/pawsx/albeto_tiny/epochs_4_bs_16_lr_5e-5/checkpoint-3000",
+  "epoch": 4.0,
+  "global_step": 12352,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.1,
+      "eval_accuracy": 0.5764999985694885,
+      "eval_loss": 0.6765033006668091,
+      "eval_runtime": 0.5222,
+      "eval_samples_per_second": 3830.009,
+      "eval_steps_per_second": 239.376,
+      "step": 300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.798413212435233e-05,
+      "loss": 0.6642,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "eval_accuracy": 0.7055000066757202,
+      "eval_loss": 0.5674529671669006,
+      "eval_runtime": 1.3054,
+      "eval_samples_per_second": 1532.102,
+      "eval_steps_per_second": 95.756,
+      "step": 600
+    },
+    {
+      "epoch": 0.29,
+      "eval_accuracy": 0.737500011920929,
+      "eval_loss": 0.5434138178825378,
+      "eval_runtime": 0.5173,
+      "eval_samples_per_second": 3866.263,
+      "eval_steps_per_second": 241.641,
+      "step": 900
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.596421632124353e-05,
+      "loss": 0.5097,
+      "step": 1000
+    },
+    {
+      "epoch": 0.39,
+      "eval_accuracy": 0.7609999775886536,
+      "eval_loss": 0.523171603679657,
+      "eval_runtime": 1.4935,
+      "eval_samples_per_second": 1339.147,
+      "eval_steps_per_second": 83.697,
+      "step": 1200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.394025259067357e-05,
+      "loss": 0.4473,
+      "step": 1500
+    },
+    {
+      "epoch": 0.49,
+      "eval_accuracy": 0.7864999771118164,
+      "eval_loss": 0.46667101979255676,
+      "eval_runtime": 0.5272,
+      "eval_samples_per_second": 3793.714,
+      "eval_steps_per_second": 237.107,
+      "step": 1500
+    },
+    {
+      "epoch": 0.58,
+      "eval_accuracy": 0.8009999990463257,
+      "eval_loss": 0.4371644854545593,
+      "eval_runtime": 0.5192,
+      "eval_samples_per_second": 3852.095,
+      "eval_steps_per_second": 240.756,
+      "step": 1800
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.191628886010363e-05,
+      "loss": 0.3934,
+      "step": 2000
+    },
+    {
+      "epoch": 0.68,
+      "eval_accuracy": 0.7894999980926514,
+      "eval_loss": 0.4638134837150574,
+      "eval_runtime": 0.5279,
+      "eval_samples_per_second": 3788.294,
+      "eval_steps_per_second": 236.768,
+      "step": 2100
+    },
+    {
+      "epoch": 0.78,
+      "eval_accuracy": 0.7885000109672546,
+      "eval_loss": 0.4593009650707245,
+      "eval_runtime": 0.5306,
+      "eval_samples_per_second": 3769.222,
+      "eval_steps_per_second": 235.576,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.989637305699482e-05,
+      "loss": 0.3659,
+      "step": 2500
+    },
+    {
+      "epoch": 0.87,
+      "eval_accuracy": 0.7960000038146973,
+      "eval_loss": 0.4648894965648651,
+      "eval_runtime": 0.5101,
+      "eval_samples_per_second": 3920.44,
+      "eval_steps_per_second": 245.027,
+      "step": 2700
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.787240932642487e-05,
+      "loss": 0.3408,
+      "step": 3000
+    },
+    {
+      "epoch": 0.97,
+      "eval_accuracy": 0.809499979019165,
+      "eval_loss": 0.42281684279441833,
+      "eval_runtime": 1.3845,
+      "eval_samples_per_second": 1444.57,
+      "eval_steps_per_second": 90.286,
+      "step": 3000
+    },
+    {
+      "epoch": 1.07,
+      "eval_accuracy": 0.8059999942779541,
+      "eval_loss": 0.4953405261039734,
+      "eval_runtime": 0.5733,
+      "eval_samples_per_second": 3488.535,
+      "eval_steps_per_second": 218.033,
+      "step": 3300
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5848445595854926e-05,
+      "loss": 0.2812,
+      "step": 3500
+    },
+    {
+      "epoch": 1.17,
+      "eval_accuracy": 0.7994999885559082,
+      "eval_loss": 0.5381343364715576,
+      "eval_runtime": 0.5151,
+      "eval_samples_per_second": 3882.955,
+      "eval_steps_per_second": 242.685,
+      "step": 3600
+    },
+    {
+      "epoch": 1.26,
+      "eval_accuracy": 0.8054999709129333,
+      "eval_loss": 0.4457036256790161,
+      "eval_runtime": 0.5211,
+      "eval_samples_per_second": 3838.159,
+      "eval_steps_per_second": 239.885,
+      "step": 3900
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.382448186528497e-05,
+      "loss": 0.2545,
+      "step": 4000
+    },
+    {
+      "epoch": 1.36,
+      "eval_accuracy": 0.8149999976158142,
+      "eval_loss": 0.4693449139595032,
+      "eval_runtime": 0.5216,
+      "eval_samples_per_second": 3834.156,
+      "eval_steps_per_second": 239.635,
+      "step": 4200
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.180051813471503e-05,
+      "loss": 0.2588,
+      "step": 4500
+    },
+    {
+      "epoch": 1.46,
+      "eval_accuracy": 0.8220000267028809,
+      "eval_loss": 0.48552772402763367,
+      "eval_runtime": 2.4236,
+      "eval_samples_per_second": 825.207,
+      "eval_steps_per_second": 51.575,
+      "step": 4500
+    },
+    {
+      "epoch": 1.55,
+      "eval_accuracy": 0.8184999823570251,
+      "eval_loss": 0.48397132754325867,
+      "eval_runtime": 0.5143,
+      "eval_samples_per_second": 3888.874,
+      "eval_steps_per_second": 243.055,
+      "step": 4800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.9780602331606216e-05,
+      "loss": 0.2436,
+      "step": 5000
+    },
+    {
+      "epoch": 1.65,
+      "eval_accuracy": 0.8080000281333923,
+      "eval_loss": 0.5025840997695923,
+      "eval_runtime": 0.5261,
+      "eval_samples_per_second": 3801.816,
+      "eval_steps_per_second": 237.613,
+      "step": 5100
+    },
+    {
+      "epoch": 1.75,
+      "eval_accuracy": 0.8140000104904175,
+      "eval_loss": 0.4753943681716919,
+      "eval_runtime": 0.5312,
+      "eval_samples_per_second": 3765.313,
+      "eval_steps_per_second": 235.332,
+      "step": 5400
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7756638601036272e-05,
+      "loss": 0.2414,
+      "step": 5500
+    },
+    {
+      "epoch": 1.85,
+      "eval_accuracy": 0.8034999966621399,
+      "eval_loss": 0.48492932319641113,
+      "eval_runtime": 0.5256,
+      "eval_samples_per_second": 3805.048,
+      "eval_steps_per_second": 237.815,
+      "step": 5700
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.573267487046632e-05,
+      "loss": 0.2527,
+      "step": 6000
+    },
+    {
+      "epoch": 1.94,
+      "eval_accuracy": 0.8100000023841858,
+      "eval_loss": 0.48092204332351685,
+      "eval_runtime": 0.52,
+      "eval_samples_per_second": 3846.385,
+      "eval_steps_per_second": 240.399,
+      "step": 6000
+    },
+    {
+      "epoch": 2.04,
+      "eval_accuracy": 0.8134999871253967,
+      "eval_loss": 0.5660321116447449,
+      "eval_runtime": 0.5164,
+      "eval_samples_per_second": 3873.13,
+      "eval_steps_per_second": 242.071,
+      "step": 6300
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.3708711139896374e-05,
+      "loss": 0.1818,
+      "step": 6500
+    },
+    {
+      "epoch": 2.14,
+      "eval_accuracy": 0.8180000185966492,
+      "eval_loss": 0.6352373361587524,
+      "eval_runtime": 0.5235,
+      "eval_samples_per_second": 3820.269,
+      "eval_steps_per_second": 238.767,
+      "step": 6600
+    },
+    {
+      "epoch": 2.23,
+      "eval_accuracy": 0.8209999799728394,
+      "eval_loss": 0.6510393619537354,
+      "eval_runtime": 0.5162,
+      "eval_samples_per_second": 3874.656,
+      "eval_steps_per_second": 242.166,
+      "step": 6900
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.1692843264248704e-05,
+      "loss": 0.1776,
+      "step": 7000
+    },
+    {
+      "epoch": 2.33,
+      "eval_accuracy": 0.8065000176429749,
+      "eval_loss": 0.684609591960907,
+      "eval_runtime": 0.5072,
+      "eval_samples_per_second": 3943.283,
+      "eval_steps_per_second": 246.455,
+      "step": 7200
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.9668879533678756e-05,
+      "loss": 0.1754,
+      "step": 7500
+    },
+    {
+      "epoch": 2.43,
+      "eval_accuracy": 0.8100000023841858,
+      "eval_loss": 0.6463525295257568,
+      "eval_runtime": 1.4644,
+      "eval_samples_per_second": 1365.751,
+      "eval_steps_per_second": 85.359,
+      "step": 7500
+    },
+    {
+      "epoch": 2.53,
+      "eval_accuracy": 0.8125,
+      "eval_loss": 0.6290258169174194,
+      "eval_runtime": 3.6012,
+      "eval_samples_per_second": 555.376,
+      "eval_steps_per_second": 34.711,
+      "step": 7800
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.764491580310881e-05,
+      "loss": 0.1763,
+      "step": 8000
+    },
+    {
+      "epoch": 2.62,
+      "eval_accuracy": 0.8144999742507935,
+      "eval_loss": 0.6613443493843079,
+      "eval_runtime": 2.9055,
+      "eval_samples_per_second": 688.341,
+      "eval_steps_per_second": 43.021,
+      "step": 8100
+    },
+    {
+      "epoch": 2.72,
+      "eval_accuracy": 0.8224999904632568,
+      "eval_loss": 0.6761817336082458,
+      "eval_runtime": 1.1073,
+      "eval_samples_per_second": 1806.22,
+      "eval_steps_per_second": 112.889,
+      "step": 8400
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.5620952072538862e-05,
+      "loss": 0.1853,
+      "step": 8500
+    },
+    {
+      "epoch": 2.82,
+      "eval_accuracy": 0.8119999766349792,
+      "eval_loss": 0.6429938077926636,
+      "eval_runtime": 0.5254,
+      "eval_samples_per_second": 3806.417,
+      "eval_steps_per_second": 237.901,
+      "step": 8700
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.3596988341968913e-05,
+      "loss": 0.1804,
+      "step": 9000
+    },
+    {
+      "epoch": 2.91,
+      "eval_accuracy": 0.8255000114440918,
+      "eval_loss": 0.6622639298439026,
+      "eval_runtime": 1.2718,
+      "eval_samples_per_second": 1572.63,
+      "eval_steps_per_second": 98.289,
+      "step": 9000
+    },
+    {
+      "epoch": 3.01,
+      "eval_accuracy": 0.8245000243186951,
+      "eval_loss": 0.6402567028999329,
+      "eval_runtime": 0.5322,
+      "eval_samples_per_second": 3757.904,
+      "eval_steps_per_second": 234.869,
+      "step": 9300
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.1577072538860104e-05,
+      "loss": 0.1321,
+      "step": 9500
+    },
+    {
+      "epoch": 3.11,
+      "eval_accuracy": 0.8184999823570251,
+      "eval_loss": 0.76594477891922,
+      "eval_runtime": 1.2705,
+      "eval_samples_per_second": 1574.241,
+      "eval_steps_per_second": 98.39,
+      "step": 9600
+    },
+    {
+      "epoch": 3.21,
+      "eval_accuracy": 0.8264999985694885,
+      "eval_loss": 0.7427929043769836,
+      "eval_runtime": 0.5386,
+      "eval_samples_per_second": 3713.17,
+      "eval_steps_per_second": 232.073,
+      "step": 9900
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.553108808290157e-06,
+      "loss": 0.1135,
+      "step": 10000
+    },
+    {
+      "epoch": 3.3,
+      "eval_accuracy": 0.8144999742507935,
+      "eval_loss": 0.7777068018913269,
+      "eval_runtime": 2.4153,
+      "eval_samples_per_second": 828.039,
+      "eval_steps_per_second": 51.752,
+      "step": 10200
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 7.529145077720208e-06,
+      "loss": 0.1182,
+      "step": 10500
+    },
+    {
+      "epoch": 3.4,
+      "eval_accuracy": 0.8199999928474426,
+      "eval_loss": 0.7448311448097229,
+      "eval_runtime": 0.5257,
+      "eval_samples_per_second": 3804.49,
+      "eval_steps_per_second": 237.781,
+      "step": 10500
+    },
+    {
+      "epoch": 3.5,
+      "eval_accuracy": 0.824999988079071,
+      "eval_loss": 0.7690622806549072,
+      "eval_runtime": 0.5391,
+      "eval_samples_per_second": 3709.781,
+      "eval_steps_per_second": 231.861,
+      "step": 10800
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.50518134715026e-06,
+      "loss": 0.1114,
+      "step": 11000
+    },
+    {
+      "epoch": 3.59,
+      "eval_accuracy": 0.8195000290870667,
+      "eval_loss": 0.7780025601387024,
+      "eval_runtime": 0.5331,
+      "eval_samples_per_second": 3751.575,
+      "eval_steps_per_second": 234.473,
+      "step": 11100
+    },
+    {
+      "epoch": 3.69,
+      "eval_accuracy": 0.8220000267028809,
+      "eval_loss": 0.7819697260856628,
+      "eval_runtime": 0.5267,
+      "eval_samples_per_second": 3797.154,
+      "eval_steps_per_second": 237.322,
+      "step": 11400
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.4852655440414507e-06,
+      "loss": 0.0992,
+      "step": 11500
+    },
+    {
+      "epoch": 3.79,
+      "eval_accuracy": 0.8215000033378601,
+      "eval_loss": 0.79557865858078,
+      "eval_runtime": 0.537,
+      "eval_samples_per_second": 3724.493,
+      "eval_steps_per_second": 232.781,
+      "step": 11700
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.4613018134715026e-06,
+      "loss": 0.1068,
+      "step": 12000
+    },
+    {
+      "epoch": 3.89,
+      "eval_accuracy": 0.8224999904632568,
+      "eval_loss": 0.793406069278717,
+      "eval_runtime": 0.5363,
+      "eval_samples_per_second": 3729.122,
+      "eval_steps_per_second": 233.07,
+      "step": 12000
+    },
+    {
+      "epoch": 3.98,
+      "eval_accuracy": 0.8215000033378601,
+      "eval_loss": 0.8001092672348022,
+      "eval_runtime": 0.5353,
+      "eval_samples_per_second": 3736.045,
+      "eval_steps_per_second": 233.503,
+      "step": 12300
+    },
+    {
+      "epoch": 4.0,
+      "step": 12352,
+      "total_flos": 144723222180000.0,
+      "train_loss": 0.24578420763806358,
+      "train_runtime": 1895.6753,
+      "train_samples_per_second": 104.239,
+      "train_steps_per_second": 6.516
+    }
+  ],
+  "max_steps": 12352,
+  "num_train_epochs": 4,
+  "total_flos": 144723222180000.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fedf9ecaa2d365a3b667011a6cc527dad72f1ad0726c031e1c0a91aa48e4499
+size 2863