Model save

README.md

@@ -0,0 +1,79 @@
+---
+base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
+library_name: peft
+license: llama3.1
+tags:
+- trl
+- sft
+- generated_from_trainer
+model-index:
+- name: Llama-31-8B_task-1_60-samples_config-2_auto
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# Llama-31-8B_task-1_60-samples_config-2_auto
+
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.5471
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0001
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 16
+- total_train_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 50
+
+### Training results
+
+| Training Loss | Epoch   | Step | Validation Loss |
+|:-------------:|:-------:|:----:|:---------------:|
+| 1.1077        | 0.6957  | 2    | 1.2225          |
+| 1.1089        | 1.7391  | 5    | 1.1334          |
+| 0.954         | 2.7826  | 8    | 1.0066          |
+| 0.9473        | 3.8261  | 11   | 0.9689          |
+| 0.7929        | 4.8696  | 14   | 0.9281          |
+| 0.7943        | 5.9130  | 17   | 0.9352          |
+| 0.6696        | 6.9565  | 20   | 0.9296          |
+| 0.5976        | 8.0     | 23   | 0.9200          |
+| 0.5767        | 8.6957  | 25   | 0.9291          |
+| 0.4423        | 9.7391  | 28   | 0.9620          |
+| 0.4059        | 10.7826 | 31   | 1.0431          |
+| 0.2734        | 11.8261 | 34   | 1.1927          |
+| 0.2387        | 12.8696 | 37   | 1.3013          |
+| 0.1347        | 13.9130 | 40   | 1.4096          |
+| 0.0897        | 14.9565 | 43   | 1.5471          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.44.0
+- Pytorch 2.1.2+cu121
+- Datasets 2.20.0
+- Tokenizers 0.19.1

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6b4a6e7e2bbb9472415138e33ff36dea86369883629a0ccda3796073bbd5f0a7
+oid sha256:a2b5ddea26c51b0f21203b46502f06e233f153c90e29fbc6351fc795a03d9eb3
 size 167832240

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 14.956521739130435,
+    "total_flos": 1.1330036980252672e+16,
+    "train_loss": 0.6092795010569484,
+    "train_runtime": 834.6765,
+    "train_samples": 46,
+    "train_samples_per_second": 2.756,
+    "train_steps_per_second": 0.12
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 14.956521739130435,
+    "total_flos": 1.1330036980252672e+16,
+    "train_loss": 0.6092795010569484,
+    "train_runtime": 834.6765,
+    "train_samples": 46,
+    "train_samples_per_second": 2.756,
+    "train_steps_per_second": 0.12
+}

trainer_state.json

@@ -0,0 +1,325 @@
+{
+  "best_metric": 0.9199500679969788,
+  "best_model_checkpoint": "data/Llama-31-8B_task-1_60-samples_config-2_auto/checkpoint-23",
+  "epoch": 14.956521739130435,
+  "eval_steps": 500,
+  "global_step": 43,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.8502998352050781,
+      "learning_rate": 1e-05,
+      "loss": 1.1831,
+      "step": 1
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.7293677926063538,
+      "learning_rate": 2e-05,
+      "loss": 1.1077,
+      "step": 2
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "eval_loss": 1.222472071647644,
+      "eval_runtime": 5.0432,
+      "eval_samples_per_second": 2.379,
+      "eval_steps_per_second": 2.379,
+      "step": 2
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.8206527829170227,
+      "learning_rate": 4e-05,
+      "loss": 1.1089,
+      "step": 4
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "eval_loss": 1.133432149887085,
+      "eval_runtime": 5.0503,
+      "eval_samples_per_second": 2.376,
+      "eval_steps_per_second": 2.376,
+      "step": 5
+    },
+    {
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.6687086820602417,
+      "learning_rate": 6e-05,
+      "loss": 1.1241,
+      "step": 6
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.3683761656284332,
+      "learning_rate": 8e-05,
+      "loss": 0.954,
+      "step": 8
+    },
+    {
+      "epoch": 2.782608695652174,
+      "eval_loss": 1.0065932273864746,
+      "eval_runtime": 5.0556,
+      "eval_samples_per_second": 2.374,
+      "eval_steps_per_second": 2.374,
+      "step": 8
+    },
+    {
+      "epoch": 3.4782608695652173,
+      "grad_norm": 0.6216813921928406,
+      "learning_rate": 0.0001,
+      "loss": 0.9473,
+      "step": 10
+    },
+    {
+      "epoch": 3.8260869565217392,
+      "eval_loss": 0.9689114689826965,
+      "eval_runtime": 5.0503,
+      "eval_samples_per_second": 2.376,
+      "eval_steps_per_second": 2.376,
+      "step": 11
+    },
+    {
+      "epoch": 4.173913043478261,
+      "grad_norm": 0.6966493725776672,
+      "learning_rate": 9.987820251299122e-05,
+      "loss": 0.9514,
+      "step": 12
+    },
+    {
+      "epoch": 4.869565217391305,
+      "grad_norm": 0.46004796028137207,
+      "learning_rate": 9.951340343707852e-05,
+      "loss": 0.7929,
+      "step": 14
+    },
+    {
+      "epoch": 4.869565217391305,
+      "eval_loss": 0.9280648231506348,
+      "eval_runtime": 5.0284,
+      "eval_samples_per_second": 2.386,
+      "eval_steps_per_second": 2.386,
+      "step": 14
+    },
+    {
+      "epoch": 5.565217391304348,
+      "grad_norm": 0.3988930583000183,
+      "learning_rate": 9.890738003669029e-05,
+      "loss": 0.7943,
+      "step": 16
+    },
+    {
+      "epoch": 5.913043478260869,
+      "eval_loss": 0.9351689219474792,
+      "eval_runtime": 5.0318,
+      "eval_samples_per_second": 2.385,
+      "eval_steps_per_second": 2.385,
+      "step": 17
+    },
+    {
+      "epoch": 6.260869565217392,
+      "grad_norm": 0.45152226090431213,
+      "learning_rate": 9.806308479691595e-05,
+      "loss": 0.7135,
+      "step": 18
+    },
+    {
+      "epoch": 6.956521739130435,
+      "grad_norm": 0.4465623199939728,
+      "learning_rate": 9.698463103929542e-05,
+      "loss": 0.6696,
+      "step": 20
+    },
+    {
+      "epoch": 6.956521739130435,
+      "eval_loss": 0.9295620322227478,
+      "eval_runtime": 5.0285,
+      "eval_samples_per_second": 2.386,
+      "eval_steps_per_second": 2.386,
+      "step": 20
+    },
+    {
+      "epoch": 7.6521739130434785,
+      "grad_norm": 0.41258516907691956,
+      "learning_rate": 9.567727288213005e-05,
+      "loss": 0.5976,
+      "step": 22
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.9199500679969788,
+      "eval_runtime": 5.0244,
+      "eval_samples_per_second": 2.388,
+      "eval_steps_per_second": 2.388,
+      "step": 23
+    },
+    {
+      "epoch": 8.347826086956522,
+      "grad_norm": 0.4886399507522583,
+      "learning_rate": 9.414737964294636e-05,
+      "loss": 0.5767,
+      "step": 24
+    },
+    {
+      "epoch": 8.695652173913043,
+      "eval_loss": 0.9290821552276611,
+      "eval_runtime": 5.0292,
+      "eval_samples_per_second": 2.386,
+      "eval_steps_per_second": 2.386,
+      "step": 25
+    },
+    {
+      "epoch": 9.043478260869565,
+      "grad_norm": 0.5453073978424072,
+      "learning_rate": 9.24024048078213e-05,
+      "loss": 0.5473,
+      "step": 26
+    },
+    {
+      "epoch": 9.73913043478261,
+      "grad_norm": 0.5217234492301941,
+      "learning_rate": 9.045084971874738e-05,
+      "loss": 0.4423,
+      "step": 28
+    },
+    {
+      "epoch": 9.73913043478261,
+      "eval_loss": 0.9619866013526917,
+      "eval_runtime": 5.0273,
+      "eval_samples_per_second": 2.387,
+      "eval_steps_per_second": 2.387,
+      "step": 28
+    },
+    {
+      "epoch": 10.434782608695652,
+      "grad_norm": 0.49829018115997314,
+      "learning_rate": 8.83022221559489e-05,
+      "loss": 0.4059,
+      "step": 30
+    },
+    {
+      "epoch": 10.782608695652174,
+      "eval_loss": 1.0431071519851685,
+      "eval_runtime": 5.0284,
+      "eval_samples_per_second": 2.386,
+      "eval_steps_per_second": 2.386,
+      "step": 31
+    },
+    {
+      "epoch": 11.130434782608695,
+      "grad_norm": 0.641713559627533,
+      "learning_rate": 8.596699001693255e-05,
+      "loss": 0.3642,
+      "step": 32
+    },
+    {
+      "epoch": 11.826086956521738,
+      "grad_norm": 0.609914243221283,
+      "learning_rate": 8.345653031794292e-05,
+      "loss": 0.2734,
+      "step": 34
+    },
+    {
+      "epoch": 11.826086956521738,
+      "eval_loss": 1.1927047967910767,
+      "eval_runtime": 5.0322,
+      "eval_samples_per_second": 2.385,
+      "eval_steps_per_second": 2.385,
+      "step": 34
+    },
+    {
+      "epoch": 12.521739130434783,
+      "grad_norm": 0.5773984789848328,
+      "learning_rate": 8.07830737662829e-05,
+      "loss": 0.2387,
+      "step": 36
+    },
+    {
+      "epoch": 12.869565217391305,
+      "eval_loss": 1.3013005256652832,
+      "eval_runtime": 5.0321,
+      "eval_samples_per_second": 2.385,
+      "eval_steps_per_second": 2.385,
+      "step": 37
+    },
+    {
+      "epoch": 13.217391304347826,
+      "grad_norm": 0.75190269947052,
+      "learning_rate": 7.795964517353735e-05,
+      "loss": 0.1781,
+      "step": 38
+    },
+    {
+      "epoch": 13.91304347826087,
+      "grad_norm": 0.7618876099586487,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.1347,
+      "step": 40
+    },
+    {
+      "epoch": 13.91304347826087,
+      "eval_loss": 1.4095535278320312,
+      "eval_runtime": 5.0317,
+      "eval_samples_per_second": 2.385,
+      "eval_steps_per_second": 2.385,
+      "step": 40
+    },
+    {
+      "epoch": 14.608695652173914,
+      "grad_norm": 0.5433855652809143,
+      "learning_rate": 7.191855733945387e-05,
+      "loss": 0.0897,
+      "step": 42
+    },
+    {
+      "epoch": 14.956521739130435,
+      "eval_loss": 1.5471285581588745,
+      "eval_runtime": 5.0278,
+      "eval_samples_per_second": 2.387,
+      "eval_steps_per_second": 2.387,
+      "step": 43
+    },
+    {
+      "epoch": 14.956521739130435,
+      "step": 43,
+      "total_flos": 1.1330036980252672e+16,
+      "train_loss": 0.6092795010569484,
+      "train_runtime": 834.6765,
+      "train_samples_per_second": 2.756,
+      "train_steps_per_second": 0.12
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 25,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 7,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1330036980252672e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}