Model save

README.md

@@ -0,0 +1,69 @@
+---
+library_name: peft
+license: llama3
+base_model: meta-llama/Meta-Llama-3-8B
+tags:
+- trl
+- sft
+- generated_from_trainer
+datasets:
+- generator
+model-index:
+- name: llama3-8b-pissa-summarization-11-v1
+  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. -->
+
+# llama3-8b-pissa-summarization-11-v1
+
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.3924
+
+## 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: 0.0002
+- train_batch_size: 12
+- eval_batch_size: 12
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 192
+- total_eval_batch_size: 96
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 1.632         | 0.9966 | 148  | 2.3924          |
+
+
+### Framework versions
+
+- PEFT 0.13.2
+- Transformers 4.46.3
+- Pytorch 2.3.1+cu121
+- Datasets 3.1.0
+- Tokenizers 0.20.3

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9966329966329966,
+    "total_flos": 6.55432743055786e+17,
+    "train_loss": 1.752148318935085,
+    "train_runtime": 587.2689,
+    "train_samples": 129221,
+    "train_samples_per_second": 48.443,
+    "train_steps_per_second": 0.252
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9966329966329966,
+    "total_flos": 6.55432743055786e+17,
+    "train_loss": 1.752148318935085,
+    "train_runtime": 587.2689,
+    "train_samples": 129221,
+    "train_samples_per_second": 48.443,
+    "train_steps_per_second": 0.252
+}

trainer_state.json

@@ -0,0 +1,260 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9966329966329966,
+  "eval_steps": 500,
+  "global_step": 148,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.006734006734006734,
+      "grad_norm": 74.241455078125,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 2.5135,
+      "step": 1
+    },
+    {
+      "epoch": 0.03367003367003367,
+      "grad_norm": 3.1438040733337402,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 2.4726,
+      "step": 5
+    },
+    {
+      "epoch": 0.06734006734006734,
+      "grad_norm": 1.2921487092971802,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 2.3003,
+      "step": 10
+    },
+    {
+      "epoch": 0.10101010101010101,
+      "grad_norm": 1.0113569498062134,
+      "learning_rate": 0.0002,
+      "loss": 2.1321,
+      "step": 15
+    },
+    {
+      "epoch": 0.13468013468013468,
+      "grad_norm": 2.183227777481079,
+      "learning_rate": 0.00019930337092856243,
+      "loss": 1.9945,
+      "step": 20
+    },
+    {
+      "epoch": 0.16835016835016836,
+      "grad_norm": 0.47641709446907043,
+      "learning_rate": 0.00019722318955551306,
+      "loss": 1.8885,
+      "step": 25
+    },
+    {
+      "epoch": 0.20202020202020202,
+      "grad_norm": 0.7186397910118103,
+      "learning_rate": 0.00019378843817721854,
+      "loss": 1.8266,
+      "step": 30
+    },
+    {
+      "epoch": 0.2356902356902357,
+      "grad_norm": 0.37949636578559875,
+      "learning_rate": 0.00018904697174694447,
+      "loss": 1.7831,
+      "step": 35
+    },
+    {
+      "epoch": 0.26936026936026936,
+      "grad_norm": 0.5123534202575684,
+      "learning_rate": 0.0001830648511318223,
+      "loss": 1.7567,
+      "step": 40
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 0.5447937250137329,
+      "learning_rate": 0.00017592542271443887,
+      "loss": 1.7298,
+      "step": 45
+    },
+    {
+      "epoch": 0.3367003367003367,
+      "grad_norm": 0.202178493142128,
+      "learning_rate": 0.00016772815716257412,
+      "loss": 1.7114,
+      "step": 50
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 0.3912014663219452,
+      "learning_rate": 0.00015858726354602248,
+      "loss": 1.7067,
+      "step": 55
+    },
+    {
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.2787935137748718,
+      "learning_rate": 0.00014863009810942815,
+      "loss": 1.6797,
+      "step": 60
+    },
+    {
+      "epoch": 0.4377104377104377,
+      "grad_norm": 0.1719539761543274,
+      "learning_rate": 0.000137995389871036,
+      "loss": 1.6721,
+      "step": 65
+    },
+    {
+      "epoch": 0.4713804713804714,
+      "grad_norm": 0.2619878053665161,
+      "learning_rate": 0.0001268313077693485,
+      "loss": 1.6664,
+      "step": 70
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "grad_norm": 0.268819659948349,
+      "learning_rate": 0.0001152933962873246,
+      "loss": 1.6685,
+      "step": 75
+    },
+    {
+      "epoch": 0.5387205387205387,
+      "grad_norm": 0.1786731630563736,
+      "learning_rate": 0.00010354240831620541,
+      "loss": 1.6523,
+      "step": 80
+    },
+    {
+      "epoch": 0.5723905723905723,
+      "grad_norm": 0.1858537644147873,
+      "learning_rate": 9.174206545276677e-05,
+      "loss": 1.6465,
+      "step": 85
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 0.25154948234558105,
+      "learning_rate": 8.005677693484077e-05,
+      "loss": 1.6344,
+      "step": 90
+    },
+    {
+      "epoch": 0.6397306397306397,
+      "grad_norm": 0.1697424203157425,
+      "learning_rate": 6.864934899622191e-05,
+      "loss": 1.6472,
+      "step": 95
+    },
+    {
+      "epoch": 0.6734006734006734,
+      "grad_norm": 0.1832387000322342,
+      "learning_rate": 5.767871655555751e-05,
+      "loss": 1.6354,
+      "step": 100
+    },
+    {
+      "epoch": 0.7070707070707071,
+      "grad_norm": 0.1768219918012619,
+      "learning_rate": 4.729772884265212e-05,
+      "loss": 1.6274,
+      "step": 105
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.17322488129138947,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 1.6235,
+      "step": 110
+    },
+    {
+      "epoch": 0.7744107744107744,
+      "grad_norm": 0.1708805114030838,
+      "learning_rate": 2.8872993029040508e-05,
+      "loss": 1.6349,
+      "step": 115
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.1731610745191574,
+      "learning_rate": 2.1085949060360654e-05,
+      "loss": 1.6322,
+      "step": 120
+    },
+    {
+      "epoch": 0.8417508417508418,
+      "grad_norm": 0.1952390819787979,
+      "learning_rate": 1.439838153227e-05,
+      "loss": 1.6391,
+      "step": 125
+    },
+    {
+      "epoch": 0.8754208754208754,
+      "grad_norm": 0.1885603815317154,
+      "learning_rate": 8.903465523913957e-06,
+      "loss": 1.6368,
+      "step": 130
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.19654668867588043,
+      "learning_rate": 4.6777594000230855e-06,
+      "loss": 1.6307,
+      "step": 135
+    },
+    {
+      "epoch": 0.9427609427609428,
+      "grad_norm": 0.17118915915489197,
+      "learning_rate": 1.7801381552624563e-06,
+      "loss": 1.6287,
+      "step": 140
+    },
+    {
+      "epoch": 0.9764309764309764,
+      "grad_norm": 0.17712068557739258,
+      "learning_rate": 2.509731335744281e-07,
+      "loss": 1.632,
+      "step": 145
+    },
+    {
+      "epoch": 0.9966329966329966,
+      "eval_loss": 2.3924484252929688,
+      "eval_runtime": 0.6217,
+      "eval_samples_per_second": 33.78,
+      "eval_steps_per_second": 1.609,
+      "step": 148
+    },
+    {
+      "epoch": 0.9966329966329966,
+      "step": 148,
+      "total_flos": 6.55432743055786e+17,
+      "train_loss": 1.752148318935085,
+      "train_runtime": 587.2689,
+      "train_samples_per_second": 48.443,
+      "train_steps_per_second": 0.252
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 148,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.55432743055786e+17,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}