End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.2
 base_model: meta-llama/Llama-3.2-3B-Instruct
 tags:
 - llama-factory
+- lora
 - generated_from_trainer
 model-index:
 - name: Llama-3.2-3B-Instruct_sft_sg_values
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.2-3B-Instruct_sft_sg_values
 
-This model is a fine-tuned version of [meta-llama/Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) on the sft_sg_values dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2791
+- Loss: 0.2716
 
 ## Model description
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 0.9988785046728972,
+    "eval_loss": 0.27161329984664917,
+    "eval_runtime": 8.2317,
+    "eval_samples_per_second": 60.741,
+    "eval_steps_per_second": 15.185,
+    "total_flos": 3.4491556954112e+16,
+    "train_loss": 0.4039554995690991,
+    "train_runtime": 629.4568,
+    "train_samples_per_second": 16.999,
+    "train_steps_per_second": 1.061
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 0.9988785046728972,
+    "eval_loss": 0.27161329984664917,
+    "eval_runtime": 8.2317,
+    "eval_samples_per_second": 60.741,
+    "eval_steps_per_second": 15.185
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9988785046728972,
+    "total_flos": 3.4491556954112e+16,
+    "train_loss": 0.4039554995690991,
+    "train_runtime": 629.4568,
+    "train_samples_per_second": 16.999,
+    "train_steps_per_second": 1.061
+}

trainer_state.json

@@ -0,0 +1,192 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9988785046728972,
+  "eval_steps": 250,
+  "global_step": 668,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0508411214953271,
+      "grad_norm": 6.148442268371582,
+      "learning_rate": 4.925373134328357e-07,
+      "loss": 0.8298,
+      "step": 34
+    },
+    {
+      "epoch": 0.1016822429906542,
+      "grad_norm": 5.179326057434082,
+      "learning_rate": 1e-06,
+      "loss": 0.7879,
+      "step": 68
+    },
+    {
+      "epoch": 0.1525233644859813,
+      "grad_norm": 8.394994735717773,
+      "learning_rate": 9.434276206322796e-07,
+      "loss": 0.7715,
+      "step": 102
+    },
+    {
+      "epoch": 0.2033644859813084,
+      "grad_norm": 5.82094669342041,
+      "learning_rate": 8.86855241264559e-07,
+      "loss": 0.6331,
+      "step": 136
+    },
+    {
+      "epoch": 0.2542056074766355,
+      "grad_norm": 4.441252708435059,
+      "learning_rate": 8.302828618968386e-07,
+      "loss": 0.5245,
+      "step": 170
+    },
+    {
+      "epoch": 0.3050467289719626,
+      "grad_norm": 2.468092441558838,
+      "learning_rate": 7.737104825291181e-07,
+      "loss": 0.4262,
+      "step": 204
+    },
+    {
+      "epoch": 0.3558878504672897,
+      "grad_norm": 2.0495502948760986,
+      "learning_rate": 7.171381031613977e-07,
+      "loss": 0.3937,
+      "step": 238
+    },
+    {
+      "epoch": 0.37383177570093457,
+      "eval_loss": 0.35631871223449707,
+      "eval_runtime": 8.2626,
+      "eval_samples_per_second": 60.514,
+      "eval_steps_per_second": 15.128,
+      "step": 250
+    },
+    {
+      "epoch": 0.4067289719626168,
+      "grad_norm": 2.0684010982513428,
+      "learning_rate": 6.605657237936771e-07,
+      "loss": 0.3419,
+      "step": 272
+    },
+    {
+      "epoch": 0.4575700934579439,
+      "grad_norm": 1.443009853363037,
+      "learning_rate": 6.039933444259567e-07,
+      "loss": 0.3097,
+      "step": 306
+    },
+    {
+      "epoch": 0.508411214953271,
+      "grad_norm": 2.7391750812530518,
+      "learning_rate": 5.474209650582362e-07,
+      "loss": 0.3268,
+      "step": 340
+    },
+    {
+      "epoch": 0.5592523364485982,
+      "grad_norm": 1.2236157655715942,
+      "learning_rate": 4.908485856905157e-07,
+      "loss": 0.2848,
+      "step": 374
+    },
+    {
+      "epoch": 0.6100934579439252,
+      "grad_norm": 2.509227991104126,
+      "learning_rate": 4.342762063227953e-07,
+      "loss": 0.2915,
+      "step": 408
+    },
+    {
+      "epoch": 0.6609345794392524,
+      "grad_norm": 1.413485050201416,
+      "learning_rate": 3.7770382695507487e-07,
+      "loss": 0.2943,
+      "step": 442
+    },
+    {
+      "epoch": 0.7117757009345794,
+      "grad_norm": 1.3338403701782227,
+      "learning_rate": 3.211314475873544e-07,
+      "loss": 0.2652,
+      "step": 476
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "eval_loss": 0.2791045904159546,
+      "eval_runtime": 8.2621,
+      "eval_samples_per_second": 60.517,
+      "eval_steps_per_second": 15.129,
+      "step": 500
+    },
+    {
+      "epoch": 0.7626168224299066,
+      "grad_norm": 1.2423838376998901,
+      "learning_rate": 2.6455906821963394e-07,
+      "loss": 0.2638,
+      "step": 510
+    },
+    {
+      "epoch": 0.8134579439252336,
+      "grad_norm": 1.2810912132263184,
+      "learning_rate": 2.0798668885191347e-07,
+      "loss": 0.255,
+      "step": 544
+    },
+    {
+      "epoch": 0.8642990654205608,
+      "grad_norm": 0.9057357907295227,
+      "learning_rate": 1.51414309484193e-07,
+      "loss": 0.2611,
+      "step": 578
+    },
+    {
+      "epoch": 0.9151401869158878,
+      "grad_norm": 0.9654864072799683,
+      "learning_rate": 9.484193011647254e-08,
+      "loss": 0.2443,
+      "step": 612
+    },
+    {
+      "epoch": 0.965981308411215,
+      "grad_norm": 2.114741802215576,
+      "learning_rate": 3.826955074875208e-08,
+      "loss": 0.2622,
+      "step": 646
+    },
+    {
+      "epoch": 0.9988785046728972,
+      "step": 668,
+      "total_flos": 3.4491556954112e+16,
+      "train_loss": 0.4039554995690991,
+      "train_runtime": 629.4568,
+      "train_samples_per_second": 16.999,
+      "train_steps_per_second": 1.061
+    }
+  ],
+  "logging_steps": 34,
+  "max_steps": 668,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 3.4491556954112e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}