End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Llama-3.1-8B
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: llama3_mammoth_dcft_ablation_50k
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3_mammoth_dcft_ablation_50k
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the mlfoundations-dev/wia_dcft_convert_to_shargpt dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0596
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.768,
+    "eval_loss": 0.05957087501883507,
+    "eval_runtime": 11.1821,
+    "eval_samples_per_second": 18.869,
+    "eval_steps_per_second": 1.252,
+    "total_flos": 6.510485588200653e+16,
+    "train_loss": 0.9969074328740438,
+    "train_runtime": 1818.6489,
+    "train_samples_per_second": 6.585,
+    "train_steps_per_second": 0.012
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.768,
+    "eval_loss": 0.05957087501883507,
+    "eval_runtime": 11.1821,
+    "eval_samples_per_second": 18.869,
+    "eval_steps_per_second": 1.252
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.768,
+    "total_flos": 6.510485588200653e+16,
+    "train_loss": 0.9969074328740438,
+    "train_runtime": 1818.6489,
+    "train_samples_per_second": 6.585,
+    "train_steps_per_second": 0.012
+}

trainer_state.json

@@ -0,0 +1,80 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.768,
+  "eval_steps": 500,
+  "global_step": 21,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.896,
+      "eval_loss": 0.06502191722393036,
+      "eval_runtime": 11.7116,
+      "eval_samples_per_second": 18.016,
+      "eval_steps_per_second": 1.195,
+      "step": 7
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.4279218679900545,
+      "learning_rate": 5e-06,
+      "loss": 1.079,
+      "step": 10
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 0.06135111674666405,
+      "eval_runtime": 11.3453,
+      "eval_samples_per_second": 18.598,
+      "eval_steps_per_second": 1.234,
+      "step": 15
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.7902273183527557,
+      "learning_rate": 5e-06,
+      "loss": 0.9271,
+      "step": 20
+    },
+    {
+      "epoch": 2.768,
+      "eval_loss": 0.05957087501883507,
+      "eval_runtime": 11.0153,
+      "eval_samples_per_second": 19.155,
+      "eval_steps_per_second": 1.271,
+      "step": 21
+    },
+    {
+      "epoch": 2.768,
+      "step": 21,
+      "total_flos": 6.510485588200653e+16,
+      "train_loss": 0.9969074328740438,
+      "train_runtime": 1818.6489,
+      "train_samples_per_second": 6.585,
+      "train_steps_per_second": 0.012
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 21,
+  "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": 6.510485588200653e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}