End of training

README.md

@@ -4,6 +4,7 @@ library_name: peft
 license: other
 tags:
 - llama-factory
+- lora
 - generated_from_trainer
 model-index:
 - name: Ministral-8B-Instruct-2410-sft-1000
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Ministral-8B-Instruct-2410-sft-1000
 
-This model is a fine-tuned version of [mistralai/Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410) on an unknown dataset.
+This model is a fine-tuned version of [mistralai/Ministral-8B-Instruct-2410](https://huggingface.co/mistralai/Ministral-8B-Instruct-2410) on the bct_non_cot_sft_1000 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1897
+- Loss: 0.0730
 
 ## Model description
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 9.955555555555556,
+    "eval_loss": 0.07303701341152191,
+    "eval_runtime": 2.3011,
+    "eval_samples_per_second": 43.457,
+    "eval_steps_per_second": 10.864,
+    "total_flos": 5.590602537998746e+16,
+    "train_loss": 0.10414764762723021,
+    "train_runtime": 762.3102,
+    "train_samples_per_second": 11.806,
+    "train_steps_per_second": 0.367
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 9.955555555555556,
+    "eval_loss": 0.07303701341152191,
+    "eval_runtime": 2.3011,
+    "eval_samples_per_second": 43.457,
+    "eval_steps_per_second": 10.864
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.955555555555556,
+    "total_flos": 5.590602537998746e+16,
+    "train_loss": 0.10414764762723021,
+    "train_runtime": 762.3102,
+    "train_samples_per_second": 11.806,
+    "train_steps_per_second": 0.367
+}

trainer_state.json

@@ -0,0 +1,278 @@
+{
+  "best_metric": 0.07303701341152191,
+  "best_model_checkpoint": "saves/Ministral-8B-Instruct-2410/sft-1000/train/checkpoint-50",
+  "epoch": 9.955555555555556,
+  "eval_steps": 50,
+  "global_step": 280,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 5.454965591430664,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 2.0272,
+      "step": 10
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.068017601966858,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.4176,
+      "step": 20
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 1.4859881401062012,
+      "learning_rate": 9.998445910004082e-05,
+      "loss": 0.0971,
+      "step": 30
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 0.8133116364479065,
+      "learning_rate": 9.944154131125642e-05,
+      "loss": 0.0592,
+      "step": 40
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.5301327109336853,
+      "learning_rate": 9.81312123475006e-05,
+      "loss": 0.0691,
+      "step": 50
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "eval_loss": 0.07303701341152191,
+      "eval_runtime": 2.2373,
+      "eval_samples_per_second": 44.696,
+      "eval_steps_per_second": 11.174,
+      "step": 50
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.3665470778942108,
+      "learning_rate": 9.607381059352038e-05,
+      "loss": 0.0708,
+      "step": 60
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 0.30413269996643066,
+      "learning_rate": 9.330127018922194e-05,
+      "loss": 0.0358,
+      "step": 70
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 0.5749717354774475,
+      "learning_rate": 8.985662536114613e-05,
+      "loss": 0.0317,
+      "step": 80
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.29326656460762024,
+      "learning_rate": 8.579334246298593e-05,
+      "loss": 0.0285,
+      "step": 90
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 0.2154272347688675,
+      "learning_rate": 8.117449009293668e-05,
+      "loss": 0.0185,
+      "step": 100
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "eval_loss": 0.0987556055188179,
+      "eval_runtime": 2.2603,
+      "eval_samples_per_second": 44.242,
+      "eval_steps_per_second": 11.061,
+      "step": 100
+    },
+    {
+      "epoch": 3.911111111111111,
+      "grad_norm": 0.3583829998970032,
+      "learning_rate": 7.60717601689749e-05,
+      "loss": 0.0172,
+      "step": 110
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 0.04914231598377228,
+      "learning_rate": 7.056435515653059e-05,
+      "loss": 0.0063,
+      "step": 120
+    },
+    {
+      "epoch": 4.622222222222222,
+      "grad_norm": 0.16161295771598816,
+      "learning_rate": 6.473775872054521e-05,
+      "loss": 0.0087,
+      "step": 130
+    },
+    {
+      "epoch": 4.977777777777778,
+      "grad_norm": 0.3194357454776764,
+      "learning_rate": 5.868240888334653e-05,
+      "loss": 0.0099,
+      "step": 140
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.10840409994125366,
+      "learning_rate": 5.249229428303486e-05,
+      "loss": 0.0054,
+      "step": 150
+    },
+    {
+      "epoch": 5.333333333333333,
+      "eval_loss": 0.14543628692626953,
+      "eval_runtime": 2.2648,
+      "eval_samples_per_second": 44.153,
+      "eval_steps_per_second": 11.038,
+      "step": 150
+    },
+    {
+      "epoch": 5.688888888888889,
+      "grad_norm": 0.16890129446983337,
+      "learning_rate": 4.626349532067879e-05,
+      "loss": 0.0032,
+      "step": 160
+    },
+    {
+      "epoch": 6.044444444444444,
+      "grad_norm": 0.14680124819278717,
+      "learning_rate": 4.0092692840030134e-05,
+      "loss": 0.0041,
+      "step": 170
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.049324482679367065,
+      "learning_rate": 3.4075667487415785e-05,
+      "loss": 0.0018,
+      "step": 180
+    },
+    {
+      "epoch": 6.7555555555555555,
+      "grad_norm": 0.11419277638196945,
+      "learning_rate": 2.8305813044122097e-05,
+      "loss": 0.0013,
+      "step": 190
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 0.009863507002592087,
+      "learning_rate": 2.2872686806712035e-05,
+      "loss": 0.0002,
+      "step": 200
+    },
+    {
+      "epoch": 7.111111111111111,
+      "eval_loss": 0.18249186873435974,
+      "eval_runtime": 2.2589,
+      "eval_samples_per_second": 44.27,
+      "eval_steps_per_second": 11.067,
+      "step": 200
+    },
+    {
+      "epoch": 7.466666666666667,
+      "grad_norm": 0.004697670228779316,
+      "learning_rate": 1.7860619515673033e-05,
+      "loss": 0.0001,
+      "step": 210
+    },
+    {
+      "epoch": 7.822222222222222,
+      "grad_norm": 0.012849841266870499,
+      "learning_rate": 1.3347406408508695e-05,
+      "loss": 0.0007,
+      "step": 220
+    },
+    {
+      "epoch": 8.177777777777777,
+      "grad_norm": 0.013538352213799953,
+      "learning_rate": 9.403099714207175e-06,
+      "loss": 0.0003,
+      "step": 230
+    },
+    {
+      "epoch": 8.533333333333333,
+      "grad_norm": 0.009252621792256832,
+      "learning_rate": 6.088921331488568e-06,
+      "loss": 0.0001,
+      "step": 240
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.0053879632614552975,
+      "learning_rate": 3.4563125677897932e-06,
+      "loss": 0.0005,
+      "step": 250
+    },
+    {
+      "epoch": 8.88888888888889,
+      "eval_loss": 0.18974514305591583,
+      "eval_runtime": 2.2394,
+      "eval_samples_per_second": 44.654,
+      "eval_steps_per_second": 11.163,
+      "step": 250
+    },
+    {
+      "epoch": 9.244444444444444,
+      "grad_norm": 0.0849616751074791,
+      "learning_rate": 1.5461356885461075e-06,
+      "loss": 0.0002,
+      "step": 260
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.0013617346994578838,
+      "learning_rate": 3.8803966999139684e-07,
+      "loss": 0.0003,
+      "step": 270
+    },
+    {
+      "epoch": 9.955555555555556,
+      "grad_norm": 0.010355186648666859,
+      "learning_rate": 0.0,
+      "loss": 0.0002,
+      "step": 280
+    },
+    {
+      "epoch": 9.955555555555556,
+      "step": 280,
+      "total_flos": 5.590602537998746e+16,
+      "train_loss": 0.10414764762723021,
+      "train_runtime": 762.3102,
+      "train_samples_per_second": 11.806,
+      "train_steps_per_second": 0.367
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 280,
+  "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": 5.590602537998746e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}