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-milora-coding-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-milora-coding-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: 1.6430
+
+## 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 |
+|:-------------:|:-----:|:----:|:---------------:|
+| 1.1598        | 1.0   | 137  | 1.6430          |
+
+
+### Framework versions
+
+- PEFT 0.13.3.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": 1.0,
+    "total_flos": 8.08957492854784e+17,
+    "train_loss": 1.2680163078934608,
+    "train_runtime": 662.1797,
+    "train_samples": 116368,
+    "train_samples_per_second": 52.795,
+    "train_steps_per_second": 0.207
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.0,
+    "total_flos": 8.08957492854784e+17,
+    "train_loss": 1.2680163078934608,
+    "train_runtime": 662.1797,
+    "train_samples": 116368,
+    "train_samples_per_second": 52.795,
+    "train_steps_per_second": 0.207
+}

trainer_state.json

@@ -0,0 +1,246 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 137,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0072992700729927005,
+      "grad_norm": 2.037527561187744,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 1.7627,
+      "step": 1
+    },
+    {
+      "epoch": 0.0364963503649635,
+      "grad_norm": 1.0240328311920166,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 1.729,
+      "step": 5
+    },
+    {
+      "epoch": 0.072992700729927,
+      "grad_norm": 0.4906023442745209,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 1.686,
+      "step": 10
+    },
+    {
+      "epoch": 0.10948905109489052,
+      "grad_norm": 0.5248916745185852,
+      "learning_rate": 0.00019996738360808565,
+      "loss": 1.624,
+      "step": 15
+    },
+    {
+      "epoch": 0.145985401459854,
+      "grad_norm": 0.5368553400039673,
+      "learning_rate": 0.00019882804237803488,
+      "loss": 1.516,
+      "step": 20
+    },
+    {
+      "epoch": 0.18248175182481752,
+      "grad_norm": 0.2822141945362091,
+      "learning_rate": 0.00019607909582962477,
+      "loss": 1.4264,
+      "step": 25
+    },
+    {
+      "epoch": 0.21897810218978103,
+      "grad_norm": 0.2575882375240326,
+      "learning_rate": 0.0001917653158603628,
+      "loss": 1.3472,
+      "step": 30
+    },
+    {
+      "epoch": 0.25547445255474455,
+      "grad_norm": 0.19714128971099854,
+      "learning_rate": 0.00018595696069872013,
+      "loss": 1.2921,
+      "step": 35
+    },
+    {
+      "epoch": 0.291970802919708,
+      "grad_norm": 0.12405715137720108,
+      "learning_rate": 0.00017874863061334657,
+      "loss": 1.2586,
+      "step": 40
+    },
+    {
+      "epoch": 0.3284671532846715,
+      "grad_norm": 0.09485290199518204,
+      "learning_rate": 0.00017025772716520323,
+      "loss": 1.2474,
+      "step": 45
+    },
+    {
+      "epoch": 0.36496350364963503,
+      "grad_norm": 0.08865249156951904,
+      "learning_rate": 0.0001606225410966638,
+      "loss": 1.2304,
+      "step": 50
+    },
+    {
+      "epoch": 0.40145985401459855,
+      "grad_norm": 0.08730119466781616,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2155,
+      "step": 55
+    },
+    {
+      "epoch": 0.43795620437956206,
+      "grad_norm": 0.07352937012910843,
+      "learning_rate": 0.0001385631124488136,
+      "loss": 1.2032,
+      "step": 60
+    },
+    {
+      "epoch": 0.4744525547445255,
+      "grad_norm": 0.06901456415653229,
+      "learning_rate": 0.0001264981502196662,
+      "loss": 1.1911,
+      "step": 65
+    },
+    {
+      "epoch": 0.5109489051094891,
+      "grad_norm": 0.06873830407857895,
+      "learning_rate": 0.00011400161449686293,
+      "loss": 1.1886,
+      "step": 70
+    },
+    {
+      "epoch": 0.5474452554744526,
+      "grad_norm": 0.07086185365915298,
+      "learning_rate": 0.00010127703547159739,
+      "loss": 1.1737,
+      "step": 75
+    },
+    {
+      "epoch": 0.583941605839416,
+      "grad_norm": 0.06949092447757721,
+      "learning_rate": 8.853165746015997e-05,
+      "loss": 1.1693,
+      "step": 80
+    },
+    {
+      "epoch": 0.6204379562043796,
+      "grad_norm": 0.06831766664981842,
+      "learning_rate": 7.597306353045393e-05,
+      "loss": 1.1595,
+      "step": 85
+    },
+    {
+      "epoch": 0.656934306569343,
+      "grad_norm": 0.06490089744329453,
+      "learning_rate": 6.380579461128819e-05,
+      "loss": 1.1741,
+      "step": 90
+    },
+    {
+      "epoch": 0.6934306569343066,
+      "grad_norm": 0.06254352629184723,
+      "learning_rate": 5.222801814877369e-05,
+      "loss": 1.1567,
+      "step": 95
+    },
+    {
+      "epoch": 0.7299270072992701,
+      "grad_norm": 0.06689339876174927,
+      "learning_rate": 4.142830056718052e-05,
+      "loss": 1.1625,
+      "step": 100
+    },
+    {
+      "epoch": 0.7664233576642335,
+      "grad_norm": 0.06685953587293625,
+      "learning_rate": 3.158253610095697e-05,
+      "loss": 1.1569,
+      "step": 105
+    },
+    {
+      "epoch": 0.8029197080291971,
+      "grad_norm": 0.06890476495027542,
+      "learning_rate": 2.2851082017805703e-05,
+      "loss": 1.153,
+      "step": 110
+    },
+    {
+      "epoch": 0.8394160583941606,
+      "grad_norm": 0.06309118121862411,
+      "learning_rate": 1.5376146891235598e-05,
+      "loss": 1.1583,
+      "step": 115
+    },
+    {
+      "epoch": 0.8759124087591241,
+      "grad_norm": 0.06340091675519943,
+      "learning_rate": 9.279474459608805e-06,
+      "loss": 1.1639,
+      "step": 120
+    },
+    {
+      "epoch": 0.9124087591240876,
+      "grad_norm": 0.06254470348358154,
+      "learning_rate": 4.660360794506946e-06,
+      "loss": 1.1547,
+      "step": 125
+    },
+    {
+      "epoch": 0.948905109489051,
+      "grad_norm": 0.06014714017510414,
+      "learning_rate": 1.5940370726542863e-06,
+      "loss": 1.161,
+      "step": 130
+    },
+    {
+      "epoch": 0.9854014598540146,
+      "grad_norm": 0.0610809363424778,
+      "learning_rate": 1.3044429107700318e-07,
+      "loss": 1.1598,
+      "step": 135
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.6429708003997803,
+      "eval_runtime": 0.7863,
+      "eval_samples_per_second": 11.446,
+      "eval_steps_per_second": 1.272,
+      "step": 137
+    },
+    {
+      "epoch": 1.0,
+      "step": 137,
+      "total_flos": 8.08957492854784e+17,
+      "train_loss": 1.2680163078934608,
+      "train_runtime": 662.1797,
+      "train_samples_per_second": 52.795,
+      "train_steps_per_second": 0.207
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 137,
+  "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": 8.08957492854784e+17,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}