End of training

README.md

@@ -1,12 +1,27 @@
 ---
+language:
+- en
 base_model: Hartunka/tiny_bert_rand_100_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_rand_100_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: 0.620626525630594
 ---
 
 <!-- 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_rand_100_v1_mnli
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_rand_100_v1](https://huggingface.co/Hartunka/tiny_bert_rand_100_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_rand_100_v1](https://huggingface.co/Hartunka/tiny_bert_rand_100_v1) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9937
-- Accuracy: 0.6231
+- Loss: 0.8518
+- Accuracy: 0.6206
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6203769740193581,
+    "eval_accuracy_mm": 0.620626525630594,
+    "eval_loss": 0.8466889262199402,
+    "eval_loss_mm": 0.8518470525741577,
+    "eval_runtime": 3.415,
+    "eval_runtime_mm": 3.4689,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2874.054,
+    "eval_samples_per_second_mm": 2834.294,
+    "eval_steps_per_second": 11.42,
+    "eval_steps_per_second_mm": 11.243,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.7577745300850004,
+    "train_runtime": 2592.0889,
+    "train_samples": 392702,
+    "train_samples_per_second": 7575.01,
+    "train_steps_per_second": 29.59
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6203769740193581,
+    "eval_accuracy_mm": 0.620626525630594,
+    "eval_loss": 0.8466889262199402,
+    "eval_loss_mm": 0.8518470525741577,
+    "eval_runtime": 3.415,
+    "eval_runtime_mm": 3.4689,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2874.054,
+    "eval_samples_per_second_mm": 2834.294,
+    "eval_steps_per_second": 11.42,
+    "eval_steps_per_second_mm": 11.243
+}

logs/events.out.tfevents.1744835951.s_004_m

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f80905bedd13de36698ecde17a84c8aa8d8ec4d9d5e2cbc9336ea8cca85bf953
+size 363

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.7577745300850004,
+    "train_runtime": 2592.0889,
+    "train_samples": 392702,
+    "train_samples_per_second": 7575.01,
+    "train_steps_per_second": 29.59
+}

trainer_state.json

@@ -0,0 +1,190 @@
+{
+  "best_metric": 0.8466889262199402,
+  "best_model_checkpoint": "tiny_bert_rand_100_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": 1.0576350688934326,
+      "learning_rate": 4.9e-05,
+      "loss": 0.9939,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5477330616403464,
+      "eval_loss": 0.9340195059776306,
+      "eval_runtime": 3.4588,
+      "eval_samples_per_second": 2837.716,
+      "eval_steps_per_second": 11.276,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.441909909248352,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9106,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5792154865002547,
+      "eval_loss": 0.8887940645217896,
+      "eval_runtime": 3.3917,
+      "eval_samples_per_second": 2893.827,
+      "eval_steps_per_second": 11.499,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4226245880126953,
+      "learning_rate": 4.7e-05,
+      "loss": 0.8576,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6055017829852267,
+      "eval_loss": 0.8593698143959045,
+      "eval_runtime": 3.44,
+      "eval_samples_per_second": 2853.221,
+      "eval_steps_per_second": 11.337,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.7679719924926758,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.8116,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6134488028527764,
+      "eval_loss": 0.8515642881393433,
+      "eval_runtime": 3.3552,
+      "eval_samples_per_second": 2925.297,
+      "eval_steps_per_second": 11.624,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.6639595031738281,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7679,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6203769740193581,
+      "eval_loss": 0.8466889262199402,
+      "eval_runtime": 3.3919,
+      "eval_samples_per_second": 2893.668,
+      "eval_steps_per_second": 11.498,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.073934555053711,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.7263,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6275089149261335,
+      "eval_loss": 0.85952228307724,
+      "eval_runtime": 3.4134,
+      "eval_samples_per_second": 2875.409,
+      "eval_steps_per_second": 11.425,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.292513847351074,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6861,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6258787570045848,
+      "eval_loss": 0.8680656552314758,
+      "eval_runtime": 3.3969,
+      "eval_samples_per_second": 2889.423,
+      "eval_steps_per_second": 11.481,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.142648935317993,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6464,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6231278655119715,
+      "eval_loss": 0.905820369720459,
+      "eval_runtime": 3.474,
+      "eval_samples_per_second": 2825.266,
+      "eval_steps_per_second": 11.226,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.892969846725464,
+      "learning_rate": 4.1e-05,
+      "loss": 0.6076,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6246561385634233,
+      "eval_loss": 0.9309349060058594,
+      "eval_runtime": 3.4173,
+      "eval_samples_per_second": 2872.186,
+      "eval_steps_per_second": 11.413,
+      "step": 13806
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 3.507983684539795,
+      "learning_rate": 4e-05,
+      "loss": 0.5699,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6231278655119715,
+      "eval_loss": 0.993652880191803,
+      "eval_runtime": 3.3978,
+      "eval_samples_per_second": 2888.612,
+      "eval_steps_per_second": 11.478,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "step": 15340,
+      "total_flos": 1.029836100834816e+17,
+      "train_loss": 0.7577745300850004,
+      "train_runtime": 2592.0889,
+      "train_samples_per_second": 7575.01,
+      "train_steps_per_second": 29.59
+    }
+  ],
+  "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
+}