End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- id
 license: apache-2.0
 base_model: LazarusNLP/IndoNanoT5-base
 tags:
@@ -17,12 +19,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [LazarusNLP/IndoNanoT5-base](https://huggingface.co/LazarusNLP/IndoNanoT5-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5157
-- Rouge1: 72.5683
-- Rouge2: 65.5815
-- Rougel: 69.6312
-- Rougelsum: 71.7918
-- Gen Len: 102.8131
+- Loss: 0.5316
+- Rouge1: 72.4894
+- Rouge2: 65.4392
+- Rougel: 69.5679
+- Rougelsum: 71.7185
+- Gen Len: 103.4539
 
 ## Model description
 

all_results.json

@@ -0,0 +1,29 @@
+{
+    "epoch": 5.0,
+    "eval_gen_len": 103.4539385847797,
+    "eval_loss": 0.5315777063369751,
+    "eval_rouge1": 72.4894,
+    "eval_rouge2": 65.4392,
+    "eval_rougeL": 69.5679,
+    "eval_rougeLsum": 71.7185,
+    "eval_runtime": 982.5591,
+    "eval_samples": 749,
+    "eval_samples_per_second": 0.762,
+    "eval_steps_per_second": 0.024,
+    "predict_gen_len": 102.60738968633706,
+    "predict_loss": 0.5196795463562012,
+    "predict_rouge1": 72.3688,
+    "predict_rouge2": 65.5214,
+    "predict_rougeL": 69.4706,
+    "predict_rougeLsum": 71.5857,
+    "predict_runtime": 4940.3619,
+    "predict_samples": 3762,
+    "predict_samples_per_second": 0.761,
+    "predict_steps_per_second": 0.024,
+    "total_flos": 4.90272982695936e+16,
+    "train_loss": 0.6274513963092069,
+    "train_runtime": 4828.0728,
+    "train_samples": 14263,
+    "train_samples_per_second": 14.771,
+    "train_steps_per_second": 0.924
+}

eval_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 5.0,
+    "eval_gen_len": 103.4539385847797,
+    "eval_loss": 0.5315777063369751,
+    "eval_rouge1": 72.4894,
+    "eval_rouge2": 65.4392,
+    "eval_rougeL": 69.5679,
+    "eval_rougeLsum": 71.7185,
+    "eval_runtime": 982.5591,
+    "eval_samples": 749,
+    "eval_samples_per_second": 0.762,
+    "eval_steps_per_second": 0.024
+}

predict_results.json

@@ -0,0 +1,12 @@
+{
+    "predict_gen_len": 102.60738968633706,
+    "predict_loss": 0.5196795463562012,
+    "predict_rouge1": 72.3688,
+    "predict_rouge2": 65.5214,
+    "predict_rougeL": 69.4706,
+    "predict_rougeLsum": 71.5857,
+    "predict_runtime": 4940.3619,
+    "predict_samples": 3762,
+    "predict_samples_per_second": 0.761,
+    "predict_steps_per_second": 0.024
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 5.0,
+    "total_flos": 4.90272982695936e+16,
+    "train_loss": 0.6274513963092069,
+    "train_runtime": 4828.0728,
+    "train_samples": 14263,
+    "train_samples_per_second": 14.771,
+    "train_steps_per_second": 0.924
+}

trainer_state.json

@@ -0,0 +1,130 @@
+{
+  "best_metric": 72.5836,
+  "best_model_checkpoint": "bin/indosum-lora-1/checkpoint-2676",
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 4460,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8521901369094849,
+      "learning_rate": 0.0008,
+      "loss": 0.8284,
+      "step": 892
+    },
+    {
+      "epoch": 1.0,
+      "eval_gen_len": 103.92656875834446,
+      "eval_loss": 0.5653803944587708,
+      "eval_rouge1": 69.049,
+      "eval_rouge2": 61.4592,
+      "eval_rougeL": 65.911,
+      "eval_rougeLsum": 68.1282,
+      "eval_runtime": 217.8265,
+      "eval_samples_per_second": 3.439,
+      "eval_steps_per_second": 0.11,
+      "step": 892
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7248573303222656,
+      "learning_rate": 0.0006,
+      "loss": 0.6258,
+      "step": 1784
+    },
+    {
+      "epoch": 2.0,
+      "eval_gen_len": 100.14819759679573,
+      "eval_loss": 0.564172089099884,
+      "eval_rouge1": 70.9515,
+      "eval_rouge2": 63.7019,
+      "eval_rougeL": 67.8649,
+      "eval_rougeLsum": 70.1412,
+      "eval_runtime": 218.3362,
+      "eval_samples_per_second": 3.43,
+      "eval_steps_per_second": 0.11,
+      "step": 1784
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6117345690727234,
+      "learning_rate": 0.0004,
+      "loss": 0.5833,
+      "step": 2676
+    },
+    {
+      "epoch": 3.0,
+      "eval_gen_len": 103.02803738317758,
+      "eval_loss": 0.5315777063369751,
+      "eval_rouge1": 72.5836,
+      "eval_rouge2": 65.5267,
+      "eval_rougeL": 69.5676,
+      "eval_rougeLsum": 71.788,
+      "eval_runtime": 219.1965,
+      "eval_samples_per_second": 3.417,
+      "eval_steps_per_second": 0.109,
+      "step": 2676
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.5995155572891235,
+      "learning_rate": 0.0002,
+      "loss": 0.5589,
+      "step": 3568
+    },
+    {
+      "epoch": 4.0,
+      "eval_gen_len": 104.12817089452604,
+      "eval_loss": 0.519862174987793,
+      "eval_rouge1": 72.1782,
+      "eval_rouge2": 65.1594,
+      "eval_rougeL": 69.2161,
+      "eval_rougeLsum": 71.3916,
+      "eval_runtime": 217.5291,
+      "eval_samples_per_second": 3.443,
+      "eval_steps_per_second": 0.11,
+      "step": 3568
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.6459794044494629,
+      "learning_rate": 0.0,
+      "loss": 0.5409,
+      "step": 4460
+    },
+    {
+      "epoch": 5.0,
+      "eval_gen_len": 102.81308411214954,
+      "eval_loss": 0.5157368779182434,
+      "eval_rouge1": 72.5683,
+      "eval_rouge2": 65.5815,
+      "eval_rougeL": 69.6312,
+      "eval_rougeLsum": 71.7918,
+      "eval_runtime": 220.2635,
+      "eval_samples_per_second": 3.4,
+      "eval_steps_per_second": 0.109,
+      "step": 4460
+    },
+    {
+      "epoch": 5.0,
+      "step": 4460,
+      "total_flos": 4.90272982695936e+16,
+      "train_loss": 0.6274513963092069,
+      "train_runtime": 4828.0728,
+      "train_samples_per_second": 14.771,
+      "train_steps_per_second": 0.924
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 4460,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 4.90272982695936e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}