End of training

README.md

@@ -14,7 +14,7 @@ model-index:
       name: Causal Language Modeling
       type: text-generation
     dataset:
-      name: wikitext
+      name: wikitext wikitext-2-raw-v1
       type: wikitext
       config: wikitext-2-raw-v1
       split: validation
@@ -30,7 +30,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # clm
 
-This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the wikitext dataset.
+This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the wikitext wikitext-2-raw-v1 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 3.4802
 - Accuracy: 0.3719

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 0.99,
+    "eval_accuracy": 0.37187601824698596,
+    "eval_loss": 3.4801909923553467,
+    "eval_runtime": 104.7921,
+    "eval_samples": 240,
+    "eval_samples_per_second": 2.29,
+    "eval_steps_per_second": 0.286,
+    "perplexity": 32.46592222670883,
+    "train_loss": 3.7061044375101724,
+    "train_runtime": 2296.6495,
+    "train_samples": 2318,
+    "train_samples_per_second": 1.009,
+    "train_steps_per_second": 0.031
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 0.99,
+    "eval_accuracy": 0.37187601824698596,
+    "eval_loss": 3.4801909923553467,
+    "eval_runtime": 104.7921,
+    "eval_samples": 240,
+    "eval_samples_per_second": 2.29,
+    "eval_steps_per_second": 0.286,
+    "perplexity": 32.46592222670883
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.99,
+    "train_loss": 3.7061044375101724,
+    "train_runtime": 2296.6495,
+    "train_samples": 2318,
+    "train_samples_per_second": 1.009,
+    "train_steps_per_second": 0.031
+}

trainer_state.json

@@ -0,0 +1,79 @@
+{
+  "best_metric": 0.37187601824698596,
+  "best_model_checkpoint": ".\\output\\checkpoint-72",
+  "epoch": 0.993103448275862,
+  "eval_steps": 500,
+  "global_step": 72,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.90625e-05,
+      "loss": 3.9372,
+      "step": 10
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.4375e-05,
+      "loss": 3.76,
+      "step": 20
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.96875e-05,
+      "loss": 3.6869,
+      "step": 30
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5e-05,
+      "loss": 3.6689,
+      "step": 40
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.03125e-05,
+      "loss": 3.6483,
+      "step": 50
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.625e-06,
+      "loss": 3.6523,
+      "step": 60
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.375e-07,
+      "loss": 3.6158,
+      "step": 70
+    },
+    {
+      "epoch": 0.99,
+      "eval_accuracy": 0.37187601824698596,
+      "eval_loss": 3.4801909923553467,
+      "eval_runtime": 101.4158,
+      "eval_samples_per_second": 2.366,
+      "eval_steps_per_second": 0.296,
+      "step": 72
+    },
+    {
+      "epoch": 0.99,
+      "step": 72,
+      "total_flos": 602027713363968.0,
+      "train_loss": 3.7061044375101724,
+      "train_runtime": 2296.6495,
+      "train_samples_per_second": 1.009,
+      "train_steps_per_second": 0.031
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 72,
+  "num_train_epochs": 1,
+  "save_steps": 100.0,
+  "total_flos": 602027713363968.0,
+  "trial_name": null,
+  "trial_params": null
+}