End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Meta-Llama-3.1-8B
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: seed_math_nvidia_math
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # seed_math_nvidia_math
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the mlfoundations-dev/seed_math_nvidia_math dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.4341
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.9451476793248945,
+    "eval_loss": 0.43405216932296753,
+    "eval_runtime": 32.4269,
+    "eval_samples_per_second": 24.64,
+    "eval_steps_per_second": 0.401,
+    "total_flos": 145518860697600.0,
+    "train_loss": 0.45809418853672074,
+    "train_runtime": 5393.8667,
+    "train_samples_per_second": 8.434,
+    "train_steps_per_second": 0.016
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.9451476793248945,
+    "eval_loss": 0.43405216932296753,
+    "eval_runtime": 32.4269,
+    "eval_samples_per_second": 24.64,
+    "eval_steps_per_second": 0.401
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.9451476793248945,
+    "total_flos": 145518860697600.0,
+    "train_loss": 0.45809418853672074,
+    "train_runtime": 5393.8667,
+    "train_samples_per_second": 8.434,
+    "train_steps_per_second": 0.016
+}

trainer_state.json

@@ -0,0 +1,122 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9451476793248945,
+  "eval_steps": 500,
+  "global_step": 87,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.33755274261603374,
+      "grad_norm": 4.635889257727947,
+      "learning_rate": 5e-06,
+      "loss": 0.5992,
+      "step": 10
+    },
+    {
+      "epoch": 0.6751054852320675,
+      "grad_norm": 1.5293826490731008,
+      "learning_rate": 5e-06,
+      "loss": 0.5008,
+      "step": 20
+    },
+    {
+      "epoch": 0.9789029535864979,
+      "eval_loss": 0.47183147072792053,
+      "eval_runtime": 32.83,
+      "eval_samples_per_second": 24.337,
+      "eval_steps_per_second": 0.396,
+      "step": 29
+    },
+    {
+      "epoch": 1.0168776371308017,
+      "grad_norm": 0.9046996051523172,
+      "learning_rate": 5e-06,
+      "loss": 0.4944,
+      "step": 30
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.7204275278694038,
+      "learning_rate": 5e-06,
+      "loss": 0.4423,
+      "step": 40
+    },
+    {
+      "epoch": 1.6919831223628692,
+      "grad_norm": 0.5065677729290418,
+      "learning_rate": 5e-06,
+      "loss": 0.4328,
+      "step": 50
+    },
+    {
+      "epoch": 1.9957805907172996,
+      "eval_loss": 0.44264522194862366,
+      "eval_runtime": 32.535,
+      "eval_samples_per_second": 24.558,
+      "eval_steps_per_second": 0.4,
+      "step": 59
+    },
+    {
+      "epoch": 2.0337552742616034,
+      "grad_norm": 0.7871918626662016,
+      "learning_rate": 5e-06,
+      "loss": 0.4452,
+      "step": 60
+    },
+    {
+      "epoch": 2.371308016877637,
+      "grad_norm": 0.5264547710314222,
+      "learning_rate": 5e-06,
+      "loss": 0.4058,
+      "step": 70
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.9173800973197113,
+      "learning_rate": 5e-06,
+      "loss": 0.3936,
+      "step": 80
+    },
+    {
+      "epoch": 2.9451476793248945,
+      "eval_loss": 0.43405216932296753,
+      "eval_runtime": 32.0216,
+      "eval_samples_per_second": 24.952,
+      "eval_steps_per_second": 0.406,
+      "step": 87
+    },
+    {
+      "epoch": 2.9451476793248945,
+      "step": 87,
+      "total_flos": 145518860697600.0,
+      "train_loss": 0.45809418853672074,
+      "train_runtime": 5393.8667,
+      "train_samples_per_second": 8.434,
+      "train_steps_per_second": 0.016
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 87,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 145518860697600.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}