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: llama3-1_8b_dolphin
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3-1_8b_dolphin
 
-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/dolphin dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5015
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.5014501214027405,
+    "eval_runtime": 6.4572,
+    "eval_samples_per_second": 264.664,
+    "eval_steps_per_second": 1.084,
+    "total_flos": 321607151124480.0,
+    "train_loss": 0.5088142100721598,
+    "train_runtime": 1474.9936,
+    "train_samples_per_second": 66.021,
+    "train_steps_per_second": 0.13
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.5014501214027405,
+    "eval_runtime": 6.4572,
+    "eval_samples_per_second": 264.664,
+    "eval_steps_per_second": 1.084
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 321607151124480.0,
+    "train_loss": 0.5088142100721598,
+    "train_runtime": 1474.9936,
+    "train_samples_per_second": 66.021,
+    "train_steps_per_second": 0.13
+}

trainer_state.json

@@ -0,0 +1,199 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 192,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15625,
+      "grad_norm": 3.410148832715147,
+      "learning_rate": 5e-06,
+      "loss": 0.7255,
+      "step": 10
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 1.9658070319836614,
+      "learning_rate": 5e-06,
+      "loss": 0.6224,
+      "step": 20
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 2.0869328497874875,
+      "learning_rate": 5e-06,
+      "loss": 0.5848,
+      "step": 30
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 1.2748507240351463,
+      "learning_rate": 5e-06,
+      "loss": 0.5638,
+      "step": 40
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.8013309231484568,
+      "learning_rate": 5e-06,
+      "loss": 0.5486,
+      "step": 50
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.8121748600070269,
+      "learning_rate": 5e-06,
+      "loss": 0.5351,
+      "step": 60
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.5265702605247498,
+      "eval_runtime": 6.4953,
+      "eval_samples_per_second": 263.115,
+      "eval_steps_per_second": 1.078,
+      "step": 64
+    },
+    {
+      "epoch": 1.09375,
+      "grad_norm": 0.6230488954615538,
+      "learning_rate": 5e-06,
+      "loss": 0.5117,
+      "step": 70
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6381342635296321,
+      "learning_rate": 5e-06,
+      "loss": 0.4927,
+      "step": 80
+    },
+    {
+      "epoch": 1.40625,
+      "grad_norm": 0.5885474331678158,
+      "learning_rate": 5e-06,
+      "loss": 0.488,
+      "step": 90
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.517571001381851,
+      "learning_rate": 5e-06,
+      "loss": 0.4873,
+      "step": 100
+    },
+    {
+      "epoch": 1.71875,
+      "grad_norm": 0.4914014029041883,
+      "learning_rate": 5e-06,
+      "loss": 0.4868,
+      "step": 110
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.5996088431576516,
+      "learning_rate": 5e-06,
+      "loss": 0.483,
+      "step": 120
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.5022982358932495,
+      "eval_runtime": 6.4313,
+      "eval_samples_per_second": 265.73,
+      "eval_steps_per_second": 1.088,
+      "step": 128
+    },
+    {
+      "epoch": 2.03125,
+      "grad_norm": 0.9648708540633031,
+      "learning_rate": 5e-06,
+      "loss": 0.4743,
+      "step": 130
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 0.641069504623041,
+      "learning_rate": 5e-06,
+      "loss": 0.4464,
+      "step": 140
+    },
+    {
+      "epoch": 2.34375,
+      "grad_norm": 0.5277740771835514,
+      "learning_rate": 5e-06,
+      "loss": 0.4484,
+      "step": 150
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.5645101097649855,
+      "learning_rate": 5e-06,
+      "loss": 0.4458,
+      "step": 160
+    },
+    {
+      "epoch": 2.65625,
+      "grad_norm": 0.6429640058172255,
+      "learning_rate": 5e-06,
+      "loss": 0.4485,
+      "step": 170
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 0.4995911800780122,
+      "learning_rate": 5e-06,
+      "loss": 0.4471,
+      "step": 180
+    },
+    {
+      "epoch": 2.96875,
+      "grad_norm": 0.6882039357462453,
+      "learning_rate": 5e-06,
+      "loss": 0.4461,
+      "step": 190
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.5014501214027405,
+      "eval_runtime": 6.2961,
+      "eval_samples_per_second": 271.439,
+      "eval_steps_per_second": 1.112,
+      "step": 192
+    },
+    {
+      "epoch": 3.0,
+      "step": 192,
+      "total_flos": 321607151124480.0,
+      "train_loss": 0.5088142100721598,
+      "train_runtime": 1474.9936,
+      "train_samples_per_second": 66.021,
+      "train_steps_per_second": 0.13
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 192,
+  "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": 321607151124480.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}