Model save

README.md

@@ -0,0 +1,58 @@
+---
+license: apache-2.0
+base_model: mistralai/Mistral-7B-v0.1
+tags:
+- generated_from_trainer
+model-index:
+- name: Mistral-7B-v0.1-gen-dpo-2k-hhrlhf
+  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. -->
+
+# Mistral-7B-v0.1-gen-dpo-2k-hhrlhf
+
+This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) 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": 0.6760619878768921,
+    "train_runtime": 756.2787,
+    "train_samples": 1999,
+    "train_samples_per_second": 2.643,
+    "train_steps_per_second": 0.082
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.43.3"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.992,
+    "total_flos": 0.0,
+    "train_loss": 0.6760619878768921,
+    "train_runtime": 756.2787,
+    "train_samples": 1999,
+    "train_samples_per_second": 2.643,
+    "train_steps_per_second": 0.082
+}

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": 183.69720508046524,
+      "learning_rate": 7.142857142857142e-08,
+      "logits/generated": -2.991734027862549,
+      "logits/real": -2.994403839111328,
+      "logps/generated": -240.1473388671875,
+      "logps/real": -227.88580322265625,
+      "loss": 0.9228,
+      "rewards/accuracies": 0.0,
+      "rewards/generated": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/real": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 209.203628848012,
+      "learning_rate": 4.727272727272727e-07,
+      "logits/generated": -2.9239561557769775,
+      "logits/real": -2.85876727104187,
+      "logps/generated": -240.1071014404297,
+      "logps/real": -223.96583557128906,
+      "loss": 0.8953,
+      "rewards/accuracies": 0.5277777910232544,
+      "rewards/generated": -0.47323077917099,
+      "rewards/margins": 0.1997741013765335,
+      "rewards/real": -0.2734566926956177,
+      "step": 10
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 165.44555944998004,
+      "learning_rate": 3.818181818181818e-07,
+      "logits/generated": -2.8631365299224854,
+      "logits/real": -2.816193103790283,
+      "logps/generated": -243.2711944580078,
+      "logps/real": -222.0016632080078,
+      "loss": 0.769,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/generated": -0.8804818987846375,
+      "rewards/margins": 1.2681009769439697,
+      "rewards/real": 0.3876190185546875,
+      "step": 20
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 161.48327588452977,
+      "learning_rate": 2.909090909090909e-07,
+      "logits/generated": -2.8604977130889893,
+      "logits/real": -2.8982977867126465,
+      "logps/generated": -249.11526489257812,
+      "logps/real": -224.89797973632812,
+      "loss": 0.6513,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/generated": -0.7799578309059143,
+      "rewards/margins": 1.13313889503479,
+      "rewards/real": 0.35318106412887573,
+      "step": 30
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 116.3767583719071,
+      "learning_rate": 2e-07,
+      "logits/generated": -2.920691728591919,
+      "logits/real": -2.928046464920044,
+      "logps/generated": -260.450927734375,
+      "logps/real": -223.0548095703125,
+      "loss": 0.5266,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/generated": -2.5974373817443848,
+      "rewards/margins": 2.437154769897461,
+      "rewards/real": -0.16028310358524323,
+      "step": 40
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 128.94846944059813,
+      "learning_rate": 1.0909090909090908e-07,
+      "logits/generated": -2.9112114906311035,
+      "logits/real": -2.9749252796173096,
+      "logps/generated": -251.634521484375,
+      "logps/real": -235.67990112304688,
+      "loss": 0.6166,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/generated": -1.891053557395935,
+      "rewards/margins": 1.692144751548767,
+      "rewards/real": -0.19890853762626648,
+      "step": 50
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 94.97279622620897,
+      "learning_rate": 1.818181818181818e-08,
+      "logits/generated": -2.898266553878784,
+      "logits/real": -2.936591625213623,
+      "logps/generated": -252.472900390625,
+      "logps/real": -228.364013671875,
+      "loss": 0.5937,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/generated": -1.678950309753418,
+      "rewards/margins": 1.5119153261184692,
+      "rewards/real": -0.16703513264656067,
+      "step": 60
+    },
+    {
+      "epoch": 0.992,
+      "step": 62,
+      "total_flos": 0.0,
+      "train_loss": 0.6760619878768921,
+      "train_runtime": 756.2787,
+      "train_samples_per_second": 2.643,
+      "train_steps_per_second": 0.082
+    }
+  ],
+  "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
+}