Model save

README.md

@@ -0,0 +1,58 @@
+---
+license: llama3.1
+base_model: meta-llama/Llama-3.1-8B
+tags:
+- generated_from_trainer
+model-index:
+- name: Llama-3.1-8B-gen-dpo-2k
+  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-3.1-8B-gen-dpo-2k
+
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on an unknown dataset.
+
+## 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: 5e-07
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
+- total_eval_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.43.3
+- Pytorch 2.2.2+cu121
+- Datasets 2.20.0
+- Tokenizers 0.19.1

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.992,
+    "total_flos": 0.0,
+    "train_loss": 0.7627417502864715,
+    "train_runtime": 798.2471,
+    "train_samples": 2000,
+    "train_samples_per_second": 2.505,
+    "train_steps_per_second": 0.078
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "eos_token_id": 128001,
+  "temperature": 0.6,
+  "top_p": 0.9,
+  "transformers_version": "4.43.3"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.992,
+    "total_flos": 0.0,
+    "train_loss": 0.7627417502864715,
+    "train_runtime": 798.2471,
+    "train_samples": 2000,
+    "train_samples_per_second": 2.505,
+    "train_steps_per_second": 0.078
+}

trainer_state.json

@@ -0,0 +1,147 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.992,
+  "eval_steps": 200,
+  "global_step": 62,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.016,
+      "grad_norm": 406.82268355522706,
+      "learning_rate": 7.142857142857142e-08,
+      "logits/generated": -1.7205866575241089,
+      "logits/real": -1.5578112602233887,
+      "logps/generated": -286.5030822753906,
+      "logps/real": -215.1640625,
+      "loss": 0.9129,
+      "rewards/accuracies": 0.0,
+      "rewards/generated": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/real": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 255.30926008233882,
+      "learning_rate": 4.727272727272727e-07,
+      "logits/generated": -1.643042802810669,
+      "logits/real": -1.5913658142089844,
+      "logps/generated": -247.2132110595703,
+      "logps/real": -229.58377075195312,
+      "loss": 0.9075,
+      "rewards/accuracies": 0.5138888955116272,
+      "rewards/generated": 0.033381447196006775,
+      "rewards/margins": 0.06060503050684929,
+      "rewards/real": 0.09398648142814636,
+      "step": 10
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 126.87007618777977,
+      "learning_rate": 3.818181818181818e-07,
+      "logits/generated": -1.6706269979476929,
+      "logits/real": -1.6231067180633545,
+      "logps/generated": -255.0316619873047,
+      "logps/real": -233.18115234375,
+      "loss": 0.8444,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/generated": -0.507367730140686,
+      "rewards/margins": 0.4785459637641907,
+      "rewards/real": -0.028821701183915138,
+      "step": 20
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 58.59915104165316,
+      "learning_rate": 2.909090909090909e-07,
+      "logits/generated": -1.6554796695709229,
+      "logits/real": -1.5990644693374634,
+      "logps/generated": -259.2175598144531,
+      "logps/real": -229.8797607421875,
+      "loss": 0.7261,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/generated": -0.31560593843460083,
+      "rewards/margins": 0.8427440524101257,
+      "rewards/real": 0.5271381139755249,
+      "step": 30
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 69.00503449133376,
+      "learning_rate": 2e-07,
+      "logits/generated": -1.6960480213165283,
+      "logits/real": -1.6210428476333618,
+      "logps/generated": -269.42431640625,
+      "logps/real": -236.22348022460938,
+      "loss": 0.6977,
+      "rewards/accuracies": 0.824999988079071,
+      "rewards/generated": -0.6807600259780884,
+      "rewards/margins": 1.2224102020263672,
+      "rewards/real": 0.5416500568389893,
+      "step": 40
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 57.46050240577938,
+      "learning_rate": 1.0909090909090908e-07,
+      "logits/generated": -1.7112897634506226,
+      "logits/real": -1.6359144449234009,
+      "logps/generated": -271.5468444824219,
+      "logps/real": -238.4114532470703,
+      "loss": 0.6996,
+      "rewards/accuracies": 0.7749999761581421,
+      "rewards/generated": -1.030948519706726,
+      "rewards/margins": 1.2689697742462158,
+      "rewards/real": 0.2380211055278778,
+      "step": 50
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 104.69079883597416,
+      "learning_rate": 1.818181818181818e-08,
+      "logits/generated": -1.7024548053741455,
+      "logits/real": -1.6506109237670898,
+      "logps/generated": -265.20843505859375,
+      "logps/real": -237.0784454345703,
+      "loss": 0.7147,
+      "rewards/accuracies": 0.824999988079071,
+      "rewards/generated": -0.9356173276901245,
+      "rewards/margins": 1.1806285381317139,
+      "rewards/real": 0.24501121044158936,
+      "step": 60
+    },
+    {
+      "epoch": 0.992,
+      "step": 62,
+      "total_flos": 0.0,
+      "train_loss": 0.7627417502864715,
+      "train_runtime": 798.2471,
+      "train_samples_per_second": 2.505,
+      "train_steps_per_second": 0.078
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 62,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}