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-alpaca-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-alpaca-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: 14.0161
+
+## 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 |
+|:-------------:|:------:|:----:|:---------------:|
+| 13.938        | 0.9945 | 91   | 14.0161         |
+
+
+### Framework versions
+
+- PEFT 0.13.3.dev0
+- Transformers 4.46.2
+- Pytorch 2.3.1+cu121
+- Datasets 3.1.0
+- Tokenizers 0.20.3

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.994535519125683,
+    "total_flos": 4.03002565094015e+17,
+    "train_loss": 13.96694589971186,
+    "train_runtime": 1243.0718,
+    "train_samples": 51241,
+    "train_samples_per_second": 14.102,
+    "train_steps_per_second": 0.073
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.994535519125683,
+    "total_flos": 4.03002565094015e+17,
+    "train_loss": 13.96694589971186,
+    "train_runtime": 1243.0718,
+    "train_samples": 51241,
+    "train_samples_per_second": 14.102,
+    "train_steps_per_second": 0.073
+}

trainer_state.json

@@ -0,0 +1,183 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.994535519125683,
+  "eval_steps": 500,
+  "global_step": 91,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01092896174863388,
+      "grad_norm": 1.7711571016959918e+18,
+      "learning_rate": 2e-05,
+      "loss": 14.0238,
+      "step": 1
+    },
+    {
+      "epoch": 0.0546448087431694,
+      "grad_norm": 3.320609228523045e+18,
+      "learning_rate": 0.0001,
+      "loss": 13.9905,
+      "step": 5
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 1.1468547292982673e+19,
+      "learning_rate": 0.0002,
+      "loss": 14.0267,
+      "step": 10
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 3.9691701809399726e+18,
+      "learning_rate": 0.00019812553106273847,
+      "loss": 14.0288,
+      "step": 15
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 1.0431252630176006e+19,
+      "learning_rate": 0.00019257239692688907,
+      "loss": 14.0141,
+      "step": 20
+    },
+    {
+      "epoch": 0.273224043715847,
+      "grad_norm": 1.0054431303639499e+19,
+      "learning_rate": 0.00018354878114129367,
+      "loss": 13.9837,
+      "step": 25
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 4.974251626953966e+18,
+      "learning_rate": 0.00017139297345578994,
+      "loss": 13.9929,
+      "step": 30
+    },
+    {
+      "epoch": 0.3825136612021858,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00015656068754865387,
+      "loss": 13.9978,
+      "step": 35
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0001396079766039157,
+      "loss": 13.9489,
+      "step": 40
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 1.6553702809539707e+19,
+      "learning_rate": 0.0001211703872229411,
+      "loss": 13.9487,
+      "step": 45
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00010193913317718244,
+      "loss": 13.9381,
+      "step": 50
+    },
+    {
+      "epoch": 0.6010928961748634,
+      "grad_norm": 4.656146420363952e+18,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 13.9454,
+      "step": 55
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 8.26754403884073e+18,
+      "learning_rate": 6.398222751952899e-05,
+      "loss": 13.9116,
+      "step": 60
+    },
+    {
+      "epoch": 0.7103825136612022,
+      "grad_norm": Infinity,
+      "learning_rate": 4.66795567198309e-05,
+      "loss": 13.9312,
+      "step": 65
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 3.3132386523262484e+18,
+      "learning_rate": 3.137583621312665e-05,
+      "loss": 13.9726,
+      "step": 70
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 3.950367432593375e+18,
+      "learning_rate": 1.864479297370325e-05,
+      "loss": 13.9497,
+      "step": 75
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 3.439258352910074e+18,
+      "learning_rate": 8.963705903385345e-06,
+      "loss": 13.9259,
+      "step": 80
+    },
+    {
+      "epoch": 0.9289617486338798,
+      "grad_norm": Infinity,
+      "learning_rate": 2.6955129420176196e-06,
+      "loss": 13.9617,
+      "step": 85
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": Infinity,
+      "learning_rate": 7.520474957699586e-08,
+      "loss": 13.938,
+      "step": 90
+    },
+    {
+      "epoch": 0.994535519125683,
+      "eval_loss": 14.016096115112305,
+      "eval_runtime": 2.0876,
+      "eval_samples_per_second": 83.349,
+      "eval_steps_per_second": 0.958,
+      "step": 91
+    },
+    {
+      "epoch": 0.994535519125683,
+      "step": 91,
+      "total_flos": 4.03002565094015e+17,
+      "train_loss": 13.96694589971186,
+      "train_runtime": 1243.0718,
+      "train_samples_per_second": 14.102,
+      "train_steps_per_second": 0.073
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 91,
+  "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": 4.03002565094015e+17,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}