End of training

README.md

@@ -4,6 +4,7 @@ library_name: peft
 license: llama3.1
 tags:
 - llama-factory
+- lora
 - generated_from_trainer
 model-index:
 - name: Llama-3.1-8B-Instruct-SFT-400
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.1-8B-Instruct-SFT-400
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on the bct_non_cot_sft_400 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0645
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 9.777777777777779,
+    "eval_loss": 0.06450273841619492,
+    "eval_runtime": 1.0476,
+    "eval_samples_per_second": 38.182,
+    "eval_steps_per_second": 19.091,
+    "total_flos": 1.959093130965811e+16,
+    "train_loss": 0.4326907604932785,
+    "train_runtime": 310.7933,
+    "train_samples_per_second": 11.583,
+    "train_steps_per_second": 0.708
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 9.777777777777779,
+    "eval_loss": 0.06450273841619492,
+    "eval_runtime": 1.0476,
+    "eval_samples_per_second": 38.182,
+    "eval_steps_per_second": 19.091
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.777777777777779,
+    "total_flos": 1.959093130965811e+16,
+    "train_loss": 0.4326907604932785,
+    "train_runtime": 310.7933,
+    "train_samples_per_second": 11.583,
+    "train_steps_per_second": 0.708
+}

trainer_state.json

@@ -0,0 +1,228 @@
+{
+  "best_metric": 0.06450273841619492,
+  "best_model_checkpoint": "saves/Llama-3.1-8B-Instruct/lora/sft-400/checkpoint-200",
+  "epoch": 9.777777777777779,
+  "eval_steps": 50,
+  "global_step": 220,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 7.734859943389893,
+      "learning_rate": 2.0454545454545457e-06,
+      "loss": 1.6805,
+      "step": 10
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 5.1195478439331055,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 1.6677,
+      "step": 20
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 3.948561668395996,
+      "learning_rate": 4.9845961611531356e-06,
+      "loss": 1.3842,
+      "step": 30
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 3.0817477703094482,
+      "learning_rate": 4.9096053963998555e-06,
+      "loss": 1.0403,
+      "step": 40
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 3.429473638534546,
+      "learning_rate": 4.7740799883862966e-06,
+      "loss": 0.8183,
+      "step": 50
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "eval_loss": 0.6115408539772034,
+      "eval_runtime": 1.0308,
+      "eval_samples_per_second": 38.806,
+      "eval_steps_per_second": 19.403,
+      "step": 50
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 3.2133073806762695,
+      "learning_rate": 4.5814246365869285e-06,
+      "loss": 0.633,
+      "step": 60
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 2.653892755508423,
+      "learning_rate": 4.336479271643833e-06,
+      "loss": 0.4486,
+      "step": 70
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 3.072075605392456,
+      "learning_rate": 4.045397465551513e-06,
+      "loss": 0.3127,
+      "step": 80
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.2421813011169434,
+      "learning_rate": 3.715491840251172e-06,
+      "loss": 0.2,
+      "step": 90
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 1.154951810836792,
+      "learning_rate": 3.3550503583141726e-06,
+      "loss": 0.1498,
+      "step": 100
+    },
+    {
+      "epoch": 4.444444444444445,
+      "eval_loss": 0.08401396125555038,
+      "eval_runtime": 1.0303,
+      "eval_samples_per_second": 38.823,
+      "eval_steps_per_second": 19.411,
+      "step": 100
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 1.2458988428115845,
+      "learning_rate": 2.973128110901026e-06,
+      "loss": 0.1251,
+      "step": 110
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.47104161977767944,
+      "learning_rate": 2.57931983374517e-06,
+      "loss": 0.1033,
+      "step": 120
+    },
+    {
+      "epoch": 5.777777777777778,
+      "grad_norm": 1.8045340776443481,
+      "learning_rate": 2.183518866065627e-06,
+      "loss": 0.102,
+      "step": 130
+    },
+    {
+      "epoch": 6.222222222222222,
+      "grad_norm": 0.7171550989151001,
+      "learning_rate": 1.7956686078964257e-06,
+      "loss": 0.0972,
+      "step": 140
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 1.1168746948242188,
+      "learning_rate": 1.425512719777071e-06,
+      "loss": 0.0829,
+      "step": 150
+    },
+    {
+      "epoch": 6.666666666666667,
+      "eval_loss": 0.06506412476301193,
+      "eval_runtime": 1.0294,
+      "eval_samples_per_second": 38.857,
+      "eval_steps_per_second": 19.429,
+      "step": 150
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 1.1370668411254883,
+      "learning_rate": 1.0823503403430736e-06,
+      "loss": 0.111,
+      "step": 160
+    },
+    {
+      "epoch": 7.555555555555555,
+      "grad_norm": 0.8633684515953064,
+      "learning_rate": 7.748024712947205e-07,
+      "loss": 0.1118,
+      "step": 170
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.568353533744812,
+      "learning_rate": 5.105953986729196e-07,
+      "loss": 0.0766,
+      "step": 180
+    },
+    {
+      "epoch": 8.444444444444445,
+      "grad_norm": 0.6475286483764648,
+      "learning_rate": 2.963665913810451e-07,
+      "loss": 0.0952,
+      "step": 190
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.6339282393455505,
+      "learning_rate": 1.3749795321332887e-07,
+      "loss": 0.0952,
+      "step": 200
+    },
+    {
+      "epoch": 8.88888888888889,
+      "eval_loss": 0.06450273841619492,
+      "eval_runtime": 1.0279,
+      "eval_samples_per_second": 38.913,
+      "eval_steps_per_second": 19.457,
+      "step": 200
+    },
+    {
+      "epoch": 9.333333333333334,
+      "grad_norm": 0.8839389681816101,
+      "learning_rate": 3.798061746947995e-08,
+      "loss": 0.0727,
+      "step": 210
+    },
+    {
+      "epoch": 9.777777777777779,
+      "grad_norm": 1.410608172416687,
+      "learning_rate": 3.146808153123293e-10,
+      "loss": 0.1111,
+      "step": 220
+    },
+    {
+      "epoch": 9.777777777777779,
+      "step": 220,
+      "total_flos": 1.959093130965811e+16,
+      "train_loss": 0.4326907604932785,
+      "train_runtime": 310.7933,
+      "train_samples_per_second": 11.583,
+      "train_steps_per_second": 0.708
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 220,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.959093130965811e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}