End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- id
 license: apache-2.0
 base_model: LazarusNLP/IndoNanoT5-base
 tags:
@@ -18,11 +20,11 @@ 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: 0.7170
-- Rouge1: 72.5364
-- Rouge2: 65.2519
-- Rougel: 69.5637
-- Rougelsum: 71.6884
-- Gen Len: 98.9053
+- Rouge1: 72.8591
+- Rouge2: 65.7981
+- Rougel: 69.8759
+- Rougelsum: 72.0557
+- Gen Len: 99.276
 
 ## Model description
 

all_results.json

@@ -1,29 +1,29 @@
 {
     "epoch": 5.0,
-    "eval_gen_len": 99.97733333333333,
-    "eval_loss": 0.7148926258087158,
-    "eval_rouge1": 72.3288,
-    "eval_rouge2": 65.2271,
-    "eval_rougeL": 69.3889,
-    "eval_rougeLsum": 71.4606,
-    "eval_runtime": 2219.8488,
+    "eval_gen_len": 99.276,
+    "eval_loss": 0.7169591784477234,
+    "eval_rouge1": 72.8591,
+    "eval_rouge2": 65.7981,
+    "eval_rougeL": 69.8759,
+    "eval_rougeLsum": 72.0557,
+    "eval_runtime": 875.1128,
     "eval_samples": 750,
-    "eval_samples_per_second": 0.338,
-    "eval_steps_per_second": 0.042,
-    "predict_gen_len": 99.5571504518873,
-    "predict_loss": 0.7386454939842224,
-    "predict_rouge1": 71.6109,
-    "predict_rouge2": 64.506,
-    "predict_rougeL": 68.5757,
-    "predict_rougeLsum": 70.7506,
-    "predict_runtime": 11191.3022,
+    "eval_samples_per_second": 0.857,
+    "eval_steps_per_second": 0.027,
+    "predict_gen_len": 98.62679425837321,
+    "predict_loss": 0.7523512244224548,
+    "predict_rouge1": 71.2774,
+    "predict_rouge2": 64.092,
+    "predict_rougeL": 68.2612,
+    "predict_rougeLsum": 70.4147,
+    "predict_runtime": 4356.0579,
     "predict_samples": 3762,
-    "predict_samples_per_second": 0.336,
-    "predict_steps_per_second": 0.042,
+    "predict_samples_per_second": 0.864,
+    "predict_steps_per_second": 0.027,
     "total_flos": 4.883004680306688e+16,
-    "train_loss": 0.7530090023945772,
-    "train_runtime": 12824.4301,
+    "train_loss": 0.5878115769458994,
+    "train_runtime": 5026.0592,
     "train_samples": 14262,
-    "train_samples_per_second": 5.56,
-    "train_steps_per_second": 1.39
+    "train_samples_per_second": 14.188,
+    "train_steps_per_second": 0.887
 }

eval_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 5.0,
-    "eval_gen_len": 99.97733333333333,
-    "eval_loss": 0.7148926258087158,
-    "eval_rouge1": 72.3288,
-    "eval_rouge2": 65.2271,
-    "eval_rougeL": 69.3889,
-    "eval_rougeLsum": 71.4606,
-    "eval_runtime": 2219.8488,
+    "eval_gen_len": 99.276,
+    "eval_loss": 0.7169591784477234,
+    "eval_rouge1": 72.8591,
+    "eval_rouge2": 65.7981,
+    "eval_rougeL": 69.8759,
+    "eval_rougeLsum": 72.0557,
+    "eval_runtime": 875.1128,
     "eval_samples": 750,
-    "eval_samples_per_second": 0.338,
-    "eval_steps_per_second": 0.042
+    "eval_samples_per_second": 0.857,
+    "eval_steps_per_second": 0.027
 }

predict_results.json

@@ -1,12 +1,12 @@
 {
-    "predict_gen_len": 99.5571504518873,
-    "predict_loss": 0.7386454939842224,
-    "predict_rouge1": 71.6109,
-    "predict_rouge2": 64.506,
-    "predict_rougeL": 68.5757,
-    "predict_rougeLsum": 70.7506,
-    "predict_runtime": 11191.3022,
+    "predict_gen_len": 98.62679425837321,
+    "predict_loss": 0.7523512244224548,
+    "predict_rouge1": 71.2774,
+    "predict_rouge2": 64.092,
+    "predict_rougeL": 68.2612,
+    "predict_rougeLsum": 70.4147,
+    "predict_runtime": 4356.0579,
     "predict_samples": 3762,
-    "predict_samples_per_second": 0.336,
-    "predict_steps_per_second": 0.042
+    "predict_samples_per_second": 0.864,
+    "predict_steps_per_second": 0.027
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 5.0,
     "total_flos": 4.883004680306688e+16,
-    "train_loss": 0.7530090023945772,
-    "train_runtime": 12824.4301,
+    "train_loss": 0.5878115769458994,
+    "train_runtime": 5026.0592,
     "train_samples": 14262,
-    "train_samples_per_second": 5.56,
-    "train_steps_per_second": 1.39
+    "train_samples_per_second": 14.188,
+    "train_steps_per_second": 0.887
 }

trainer_state.json

@@ -1,130 +1,130 @@
 {
-  "best_metric": 71.7161,
-  "best_model_checkpoint": "bin/indosum-base-0/checkpoint-17830",
+  "best_metric": 72.5364,
+  "best_model_checkpoint": "bin/indosum-base-0/checkpoint-4460",
   "epoch": 5.0,
   "eval_steps": 500,
-  "global_step": 17830,
+  "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": 2.044853687286377,
+      "grad_norm": 1.1983163356781006,
       "learning_rate": 0.0008,
-      "loss": 1.5249,
-      "step": 3566
+      "loss": 1.2132,
+      "step": 892
     },
     {
       "epoch": 1.0,
-      "eval_gen_len": 103.82933333333334,
-      "eval_loss": 0.8857953548431396,
-      "eval_rouge1": 65.9377,
-      "eval_rouge2": 58.0172,
-      "eval_rougeL": 63.1,
-      "eval_rougeLsum": 65.0124,
-      "eval_runtime": 640.0795,
-      "eval_samples_per_second": 1.172,
-      "eval_steps_per_second": 0.147,
-      "step": 3566
+      "eval_gen_len": 94.092,
+      "eval_loss": 0.7741625905036926,
+      "eval_rouge1": 67.4414,
+      "eval_rouge2": 59.7409,
+      "eval_rougeL": 64.517,
+      "eval_rougeLsum": 66.4918,
+      "eval_runtime": 191.8213,
+      "eval_samples_per_second": 3.91,
+      "eval_steps_per_second": 0.125,
+      "step": 892
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.6789307594299316,
+      "grad_norm": 1.1038604974746704,
       "learning_rate": 0.0006,
-      "loss": 0.8202,
-      "step": 7132
+      "loss": 0.686,
+      "step": 1784
     },
     {
       "epoch": 2.0,
-      "eval_gen_len": 101.31733333333334,
-      "eval_loss": 0.7492297291755676,
-      "eval_rouge1": 69.7595,
-      "eval_rouge2": 62.2139,
-      "eval_rougeL": 66.6576,
-      "eval_rougeLsum": 68.9014,
-      "eval_runtime": 623.9837,
-      "eval_samples_per_second": 1.202,
-      "eval_steps_per_second": 0.151,
-      "step": 7132
+      "eval_gen_len": 100.29333333333334,
+      "eval_loss": 0.6673194766044617,
+      "eval_rouge1": 70.2138,
+      "eval_rouge2": 62.8202,
+      "eval_rougeL": 67.1553,
+      "eval_rougeLsum": 69.3063,
+      "eval_runtime": 191.8169,
+      "eval_samples_per_second": 3.91,
+      "eval_steps_per_second": 0.125,
+      "step": 1784
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.6821009516716003,
+      "grad_norm": 0.5924516916275024,
       "learning_rate": 0.0004,
-      "loss": 0.6232,
-      "step": 10698
+      "loss": 0.491,
+      "step": 2676
     },
     {
       "epoch": 3.0,
-      "eval_gen_len": 97.70933333333333,
-      "eval_loss": 0.6953277587890625,
-      "eval_rouge1": 70.7507,
-      "eval_rouge2": 63.4429,
-      "eval_rougeL": 67.7582,
-      "eval_rougeLsum": 69.8526,
-      "eval_runtime": 605.7402,
-      "eval_samples_per_second": 1.238,
-      "eval_steps_per_second": 0.155,
-      "step": 10698
+      "eval_gen_len": 100.944,
+      "eval_loss": 0.6274383664131165,
+      "eval_rouge1": 71.2142,
+      "eval_rouge2": 63.9943,
+      "eval_rougeL": 68.2722,
+      "eval_rougeLsum": 70.2971,
+      "eval_runtime": 193.0753,
+      "eval_samples_per_second": 3.884,
+      "eval_steps_per_second": 0.124,
+      "step": 2676
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.5773214101791382,
+      "grad_norm": 0.4989166855812073,
       "learning_rate": 0.0002,
-      "loss": 0.4728,
-      "step": 14264
+      "loss": 0.343,
+      "step": 3568
     },
     {
       "epoch": 4.0,
-      "eval_gen_len": 97.78933333333333,
-      "eval_loss": 0.6717060804367065,
-      "eval_rouge1": 70.8339,
-      "eval_rouge2": 63.5583,
-      "eval_rougeL": 67.7546,
-      "eval_rougeLsum": 69.8695,
-      "eval_runtime": 607.3846,
-      "eval_samples_per_second": 1.235,
-      "eval_steps_per_second": 0.155,
-      "step": 14264
+      "eval_gen_len": 98.82266666666666,
+      "eval_loss": 0.6469114422798157,
+      "eval_rouge1": 71.7114,
+      "eval_rouge2": 64.489,
+      "eval_rougeL": 68.7214,
+      "eval_rougeLsum": 70.7949,
+      "eval_runtime": 192.2303,
+      "eval_samples_per_second": 3.902,
+      "eval_steps_per_second": 0.125,
+      "step": 3568
     },
     {
       "epoch": 5.0,
-      "grad_norm": 1.1052696704864502,
+      "grad_norm": 0.6214805245399475,
       "learning_rate": 0.0,
-      "loss": 0.3238,
-      "step": 17830
+      "loss": 0.2059,
+      "step": 4460
     },
     {
       "epoch": 5.0,
-      "eval_gen_len": 99.53733333333334,
-      "eval_loss": 0.7148926258087158,
-      "eval_rouge1": 71.7161,
-      "eval_rouge2": 64.386,
-      "eval_rougeL": 68.5898,
-      "eval_rougeLsum": 70.7721,
-      "eval_runtime": 612.7477,
-      "eval_samples_per_second": 1.224,
-      "eval_steps_per_second": 0.153,
-      "step": 17830
+      "eval_gen_len": 98.90533333333333,
+      "eval_loss": 0.7169591784477234,
+      "eval_rouge1": 72.5364,
+      "eval_rouge2": 65.2519,
+      "eval_rougeL": 69.5637,
+      "eval_rougeLsum": 71.6884,
+      "eval_runtime": 195.8892,
+      "eval_samples_per_second": 3.829,
+      "eval_steps_per_second": 0.123,
+      "step": 4460
     },
     {
       "epoch": 5.0,
-      "step": 17830,
+      "step": 4460,
       "total_flos": 4.883004680306688e+16,
-      "train_loss": 0.7530090023945772,
-      "train_runtime": 12824.4301,
-      "train_samples_per_second": 5.56,
-      "train_steps_per_second": 1.39
+      "train_loss": 0.5878115769458994,
+      "train_runtime": 5026.0592,
+      "train_samples_per_second": 14.188,
+      "train_steps_per_second": 0.887
     }
   ],
   "logging_steps": 500,
-  "max_steps": 17830,
+  "max_steps": 4460,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
   "total_flos": 4.883004680306688e+16,
-  "train_batch_size": 4,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null
 }