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-kasa-alpaca-11-v2
+  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-kasa-alpaca-11-v2
+
+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: 1.8537
+
+## 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: 16
+- eval_batch_size: 16
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 256
+- total_eval_batch_size: 128
+- 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 |
+|:-------------:|:------:|:----:|:---------------:|
+| 1.9403        | 0.9927 | 68   | 1.8537          |
+
+
+### Framework versions
+
+- PEFT 0.13.1.dev0
+- 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": 0.9927007299270073,
+    "total_flos": 4.015265797185208e+17,
+    "train_loss": 2.086618286721847,
+    "train_runtime": 356.9899,
+    "train_samples": 51241,
+    "train_samples_per_second": 49.105,
+    "train_steps_per_second": 0.19
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9927007299270073,
+    "total_flos": 4.015265797185208e+17,
+    "train_loss": 2.086618286721847,
+    "train_runtime": 356.9899,
+    "train_samples": 51241,
+    "train_samples_per_second": 49.105,
+    "train_steps_per_second": 0.19
+}

trainer_state.json

@@ -0,0 +1,148 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9927007299270073,
+  "eval_steps": 500,
+  "global_step": 68,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.014598540145985401,
+      "grad_norm": 2.1202776432037354,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 2.5137,
+      "step": 1
+    },
+    {
+      "epoch": 0.072992700729927,
+      "grad_norm": 2.048658847808838,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 2.5005,
+      "step": 5
+    },
+    {
+      "epoch": 0.145985401459854,
+      "grad_norm": 0.44883406162261963,
+      "learning_rate": 0.00019880878960910772,
+      "loss": 2.3956,
+      "step": 10
+    },
+    {
+      "epoch": 0.21897810218978103,
+      "grad_norm": 0.5522064566612244,
+      "learning_rate": 0.0001916316904487005,
+      "loss": 2.2636,
+      "step": 15
+    },
+    {
+      "epoch": 0.291970802919708,
+      "grad_norm": 0.44437357783317566,
+      "learning_rate": 0.00017841198065767107,
+      "loss": 2.1699,
+      "step": 20
+    },
+    {
+      "epoch": 0.36496350364963503,
+      "grad_norm": 0.4230095446109772,
+      "learning_rate": 0.00016002142805483685,
+      "loss": 2.0816,
+      "step": 25
+    },
+    {
+      "epoch": 0.43795620437956206,
+      "grad_norm": 0.29534509778022766,
+      "learning_rate": 0.00013767278936351854,
+      "loss": 2.0278,
+      "step": 30
+    },
+    {
+      "epoch": 0.5109489051094891,
+      "grad_norm": 0.20641866326332092,
+      "learning_rate": 0.00011283983551465511,
+      "loss": 1.9927,
+      "step": 35
+    },
+    {
+      "epoch": 0.583941605839416,
+      "grad_norm": 0.18540987372398376,
+      "learning_rate": 8.71601644853449e-05,
+      "loss": 1.975,
+      "step": 40
+    },
+    {
+      "epoch": 0.656934306569343,
+      "grad_norm": 0.20340602099895477,
+      "learning_rate": 6.232721063648148e-05,
+      "loss": 1.9558,
+      "step": 45
+    },
+    {
+      "epoch": 0.7299270072992701,
+      "grad_norm": 0.15005244314670563,
+      "learning_rate": 3.997857194516319e-05,
+      "loss": 1.9683,
+      "step": 50
+    },
+    {
+      "epoch": 0.8029197080291971,
+      "grad_norm": 0.1379683017730713,
+      "learning_rate": 2.1588019342328968e-05,
+      "loss": 1.9412,
+      "step": 55
+    },
+    {
+      "epoch": 0.8759124087591241,
+      "grad_norm": 0.12855419516563416,
+      "learning_rate": 8.368309551299536e-06,
+      "loss": 1.9671,
+      "step": 60
+    },
+    {
+      "epoch": 0.948905109489051,
+      "grad_norm": 0.11833244562149048,
+      "learning_rate": 1.1912103908922945e-06,
+      "loss": 1.9403,
+      "step": 65
+    },
+    {
+      "epoch": 0.9927007299270073,
+      "eval_loss": 1.8536934852600098,
+      "eval_runtime": 1.7343,
+      "eval_samples_per_second": 100.331,
+      "eval_steps_per_second": 1.153,
+      "step": 68
+    },
+    {
+      "epoch": 0.9927007299270073,
+      "step": 68,
+      "total_flos": 4.015265797185208e+17,
+      "train_loss": 2.086618286721847,
+      "train_runtime": 356.9899,
+      "train_samples_per_second": 49.105,
+      "train_steps_per_second": 0.19
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 68,
+  "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": 4.015265797185208e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}