End of training

README.md

@@ -1,12 +1,27 @@
 ---
+language:
+- en
 base_model: Hartunka/tiny_bert_km_50_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_km_50_v1_mnli
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MNLI
+      type: glue
+      args: mnli
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.6413751017087063
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -14,10 +29,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tiny_bert_km_50_v1_mnli
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_km_50_v1](https://huggingface.co/Hartunka/tiny_bert_km_50_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_km_50_v1](https://huggingface.co/Hartunka/tiny_bert_km_50_v1) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9343
-- Accuracy: 0.6449
+- Loss: 0.8205
+- Accuracy: 0.6414
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6298522669383596,
+    "eval_accuracy_mm": 0.6413751017087063,
+    "eval_loss": 0.8321161270141602,
+    "eval_loss_mm": 0.8204724192619324,
+    "eval_runtime": 3.4322,
+    "eval_runtime_mm": 3.4931,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2859.651,
+    "eval_samples_per_second_mm": 2814.703,
+    "eval_steps_per_second": 11.363,
+    "eval_steps_per_second_mm": 11.165,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.751894306048576,
+    "train_runtime": 2625.2366,
+    "train_samples": 392702,
+    "train_samples_per_second": 7479.364,
+    "train_steps_per_second": 29.216
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6298522669383596,
+    "eval_accuracy_mm": 0.6413751017087063,
+    "eval_loss": 0.8321161270141602,
+    "eval_loss_mm": 0.8204724192619324,
+    "eval_runtime": 3.4322,
+    "eval_runtime_mm": 3.4931,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2859.651,
+    "eval_samples_per_second_mm": 2814.703,
+    "eval_steps_per_second": 11.363,
+    "eval_steps_per_second_mm": 11.165
+}

logs/events.out.tfevents.1744818134.s_004_m

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee906ca2a160189e6dd3563c0e134cee112c6d3f1bea721c1d2d9f923ce0018d
+size 363

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.751894306048576,
+    "train_runtime": 2625.2366,
+    "train_samples": 392702,
+    "train_samples_per_second": 7479.364,
+    "train_steps_per_second": 29.216
+}

trainer_state.json

@@ -0,0 +1,190 @@
+{
+  "best_metric": 0.8321161270141602,
+  "best_model_checkpoint": "tiny_bert_km_50_v1_mnli/checkpoint-7670",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 15340,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9514528512954712,
+      "learning_rate": 4.9e-05,
+      "loss": 1.0082,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5437595517065715,
+      "eval_loss": 0.942550778388977,
+      "eval_runtime": 3.5158,
+      "eval_samples_per_second": 2791.707,
+      "eval_steps_per_second": 11.093,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3989537954330444,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9147,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5845134997452879,
+      "eval_loss": 0.8889986276626587,
+      "eval_runtime": 3.4266,
+      "eval_samples_per_second": 2864.392,
+      "eval_steps_per_second": 11.382,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4162970781326294,
+      "learning_rate": 4.7e-05,
+      "loss": 0.8604,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.606316861946001,
+      "eval_loss": 0.8645786046981812,
+      "eval_runtime": 3.4993,
+      "eval_samples_per_second": 2804.851,
+      "eval_steps_per_second": 11.145,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.4847444295883179,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.8145,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.615588385124809,
+      "eval_loss": 0.8447086811065674,
+      "eval_runtime": 3.4418,
+      "eval_samples_per_second": 2851.689,
+      "eval_steps_per_second": 11.331,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.7801928520202637,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7652,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6298522669383596,
+      "eval_loss": 0.8321161270141602,
+      "eval_runtime": 3.451,
+      "eval_samples_per_second": 2844.112,
+      "eval_steps_per_second": 11.301,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.2125303745269775,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.7172,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6405501782985227,
+      "eval_loss": 0.84904545545578,
+      "eval_runtime": 3.4667,
+      "eval_samples_per_second": 2831.233,
+      "eval_steps_per_second": 11.25,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.1476242542266846,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6722,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6512480896586856,
+      "eval_loss": 0.8362392783164978,
+      "eval_runtime": 3.4933,
+      "eval_samples_per_second": 2809.683,
+      "eval_steps_per_second": 11.164,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.0134174823760986,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6289,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6459500764136525,
+      "eval_loss": 0.8634655475616455,
+      "eval_runtime": 3.4256,
+      "eval_samples_per_second": 2865.208,
+      "eval_steps_per_second": 11.385,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.7003607749938965,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5887,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6463576158940397,
+      "eval_loss": 0.9014362096786499,
+      "eval_runtime": 3.4293,
+      "eval_samples_per_second": 2862.09,
+      "eval_steps_per_second": 11.373,
+      "step": 13806
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 3.0219602584838867,
+      "learning_rate": 4e-05,
+      "loss": 0.5489,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6449312277126846,
+      "eval_loss": 0.9343048334121704,
+      "eval_runtime": 3.4187,
+      "eval_samples_per_second": 2870.96,
+      "eval_steps_per_second": 11.408,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "step": 15340,
+      "total_flos": 1.029836100834816e+17,
+      "train_loss": 0.751894306048576,
+      "train_runtime": 2625.2366,
+      "train_samples_per_second": 7479.364,
+      "train_steps_per_second": 29.216
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 76700,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "total_flos": 1.029836100834816e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}