Model save

README.md

@@ -0,0 +1,68 @@
+---
+base_model: AmberYifan/mistral-v0.1-7b-sft-ultrachat-safeRLHF
+library_name: transformers
+model_name: Mistral-7B-v0.1-Llama-2-7b-mix
+tags:
+- generated_from_trainer
+- trl
+- dpo
+licence: license
+---
+
+# Model Card for Mistral-7B-v0.1-Llama-2-7b-mix
+
+This model is a fine-tuned version of [AmberYifan/mistral-v0.1-7b-sft-ultrachat-safeRLHF](https://huggingface.co/AmberYifan/mistral-v0.1-7b-sft-ultrachat-safeRLHF).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="AmberYifan/Mistral-7B-v0.1-Llama-2-7b-mix", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/yifanwang/huggingface/runs/07vdcz74)
+
+This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
+
+### Framework versions
+
+- TRL: 0.12.2
+- Transformers: 4.46.3
+- Pytorch: 2.5.1+cu118
+- Datasets: 3.2.0
+- Tokenizers: 0.20.3
+
+## Citations
+
+Cite DPO as:
+
+```bibtex
+@inproceedings{rafailov2023direct,
+    title        = {{Direct Preference Optimization: Your Language Model is Secretly a Reward Model}},
+    author       = {Rafael Rafailov and Archit Sharma and Eric Mitchell and Christopher D. Manning and Stefano Ermon and Chelsea Finn},
+    year         = 2023,
+    booktitle    = {Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023},
+    url          = {http://papers.nips.cc/paper_files/paper/2023/hash/a85b405ed65c6477a4fe8302b5e06ce7-Abstract-Conference.html},
+    editor       = {Alice Oh and Tristan Naumann and Amir Globerson and Kate Saenko and Moritz Hardt and Sergey Levine},
+}
+```
+
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "total_flos": 0.0,
+    "train_loss": 0.23619326285787792,
+    "train_runtime": 4850.984,
+    "train_samples": 10000,
+    "train_samples_per_second": 6.184,
+    "train_steps_per_second": 0.194
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.46.3"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "total_flos": 0.0,
+    "train_loss": 0.23619326285787792,
+    "train_runtime": 4850.984,
+    "train_samples": 10000,
+    "train_samples_per_second": 6.184,
+    "train_steps_per_second": 0.194
+}

trainer_state.json

@@ -0,0 +1,1500 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 939,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003194888178913738,
+      "grad_norm": 63.76829168210548,
+      "learning_rate": 5.3191489361702125e-09,
+      "logits/chosen": -3.703125,
+      "logits/rejected": -3.859375,
+      "logps/chosen": -139.0,
+      "logps/rejected": -80.0,
+      "loss": 0.6914,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 92.94577098387109,
+      "learning_rate": 5.3191489361702123e-08,
+      "logits/chosen": -3.625,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -99.5,
+      "logps/rejected": -96.0,
+      "loss": 0.6939,
+      "rewards/accuracies": 0.1527777761220932,
+      "rewards/chosen": -0.0010528564453125,
+      "rewards/margins": -0.001739501953125,
+      "rewards/rejected": 0.0006866455078125,
+      "step": 10
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 70.79325392606067,
+      "learning_rate": 1.0638297872340425e-07,
+      "logits/chosen": -3.375,
+      "logits/rejected": -3.5625,
+      "logps/chosen": -96.0,
+      "logps/rejected": -105.0,
+      "loss": 0.6872,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": -0.0242919921875,
+      "rewards/margins": 0.017333984375,
+      "rewards/rejected": -0.04150390625,
+      "step": 20
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 63.190582339280105,
+      "learning_rate": 1.5957446808510638e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.453125,
+      "logps/chosen": -99.5,
+      "logps/rejected": -108.0,
+      "loss": 0.671,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": -0.08935546875,
+      "rewards/margins": 0.0361328125,
+      "rewards/rejected": -0.125,
+      "step": 30
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 70.46556001874696,
+      "learning_rate": 2.127659574468085e-07,
+      "logits/chosen": -3.578125,
+      "logits/rejected": -3.59375,
+      "logps/chosen": -110.0,
+      "logps/rejected": -95.5,
+      "loss": 0.6578,
+      "rewards/accuracies": 0.48750001192092896,
+      "rewards/chosen": -0.1669921875,
+      "rewards/margins": 0.0164794921875,
+      "rewards/rejected": -0.1826171875,
+      "step": 40
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 53.92661468897121,
+      "learning_rate": 2.659574468085106e-07,
+      "logits/chosen": -3.6875,
+      "logits/rejected": -3.640625,
+      "logps/chosen": -95.0,
+      "logps/rejected": -106.5,
+      "loss": 0.6153,
+      "rewards/accuracies": 0.574999988079071,
+      "rewards/chosen": -0.251953125,
+      "rewards/margins": 0.1376953125,
+      "rewards/rejected": -0.388671875,
+      "step": 50
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 61.31819398041802,
+      "learning_rate": 3.1914893617021275e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.328125,
+      "logps/chosen": -101.0,
+      "logps/rejected": -107.0,
+      "loss": 0.7335,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": -0.283203125,
+      "rewards/margins": 0.478515625,
+      "rewards/rejected": -0.76171875,
+      "step": 60
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 55.44332583758203,
+      "learning_rate": 3.7234042553191484e-07,
+      "logits/chosen": -3.421875,
+      "logits/rejected": -3.546875,
+      "logps/chosen": -103.0,
+      "logps/rejected": -112.0,
+      "loss": 0.6863,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": -0.1220703125,
+      "rewards/margins": 0.6328125,
+      "rewards/rejected": -0.75390625,
+      "step": 70
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 51.149001898192076,
+      "learning_rate": 4.25531914893617e-07,
+      "logits/chosen": -3.546875,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -110.5,
+      "logps/rejected": -110.5,
+      "loss": 0.5805,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": -0.0986328125,
+      "rewards/margins": 0.62890625,
+      "rewards/rejected": -0.7265625,
+      "step": 80
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 55.06999142181094,
+      "learning_rate": 4.787234042553192e-07,
+      "logits/chosen": -3.671875,
+      "logits/rejected": -3.59375,
+      "logps/chosen": -107.0,
+      "logps/rejected": -116.0,
+      "loss": 0.5765,
+      "rewards/accuracies": 0.7124999761581421,
+      "rewards/chosen": -0.1494140625,
+      "rewards/margins": 0.7734375,
+      "rewards/rejected": -0.921875,
+      "step": 90
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 33.54407800330637,
+      "learning_rate": 4.964497041420119e-07,
+      "logits/chosen": -3.546875,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -105.0,
+      "logps/rejected": -112.5,
+      "loss": 0.5228,
+      "rewards/accuracies": 0.824999988079071,
+      "rewards/chosen": -0.1162109375,
+      "rewards/margins": 0.61328125,
+      "rewards/rejected": -0.7265625,
+      "step": 100
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 71.5169022650209,
+      "learning_rate": 4.905325443786982e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.515625,
+      "logps/chosen": -101.5,
+      "logps/rejected": -123.0,
+      "loss": 0.5408,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": -0.1337890625,
+      "rewards/margins": 0.8515625,
+      "rewards/rejected": -0.984375,
+      "step": 110
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 65.33074960619584,
+      "learning_rate": 4.846153846153846e-07,
+      "logits/chosen": -3.546875,
+      "logits/rejected": -3.5625,
+      "logps/chosen": -100.5,
+      "logps/rejected": -139.0,
+      "loss": 0.5381,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/chosen": 0.021484375,
+      "rewards/margins": 1.671875,
+      "rewards/rejected": -1.6484375,
+      "step": 120
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 49.04091358646178,
+      "learning_rate": 4.78698224852071e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.5,
+      "logps/chosen": -95.0,
+      "logps/rejected": -116.5,
+      "loss": 0.5387,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": -0.10791015625,
+      "rewards/margins": 0.99609375,
+      "rewards/rejected": -1.1015625,
+      "step": 130
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 49.48725125304273,
+      "learning_rate": 4.727810650887574e-07,
+      "logits/chosen": -3.671875,
+      "logits/rejected": -3.515625,
+      "logps/chosen": -101.0,
+      "logps/rejected": -121.0,
+      "loss": 0.4801,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": -0.0260009765625,
+      "rewards/margins": 1.609375,
+      "rewards/rejected": -1.6328125,
+      "step": 140
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 88.63757443724184,
+      "learning_rate": 4.668639053254438e-07,
+      "logits/chosen": -3.671875,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -104.5,
+      "logps/rejected": -107.5,
+      "loss": 0.5525,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": -0.0185546875,
+      "rewards/margins": 0.88671875,
+      "rewards/rejected": -0.90625,
+      "step": 150
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 65.41898185778713,
+      "learning_rate": 4.6094674556213014e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.375,
+      "logps/chosen": -102.5,
+      "logps/rejected": -110.0,
+      "loss": 0.4824,
+      "rewards/accuracies": 0.737500011920929,
+      "rewards/chosen": -0.20703125,
+      "rewards/margins": 1.0390625,
+      "rewards/rejected": -1.2421875,
+      "step": 160
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 64.3627134555641,
+      "learning_rate": 4.5502958579881655e-07,
+      "logits/chosen": -3.671875,
+      "logits/rejected": -3.5625,
+      "logps/chosen": -95.5,
+      "logps/rejected": -114.0,
+      "loss": 0.4837,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/chosen": 0.04736328125,
+      "rewards/margins": 1.3984375,
+      "rewards/rejected": -1.3515625,
+      "step": 170
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 50.45380597678468,
+      "learning_rate": 4.491124260355029e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.453125,
+      "logps/chosen": -100.5,
+      "logps/rejected": -114.5,
+      "loss": 0.4732,
+      "rewards/accuracies": 0.8125,
+      "rewards/chosen": -0.031005859375,
+      "rewards/margins": 1.1328125,
+      "rewards/rejected": -1.1640625,
+      "step": 180
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 47.054283179898476,
+      "learning_rate": 4.4319526627218936e-07,
+      "logits/chosen": -3.59375,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -103.5,
+      "logps/rejected": -112.0,
+      "loss": 0.4777,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": 0.08154296875,
+      "rewards/margins": 1.203125,
+      "rewards/rejected": -1.125,
+      "step": 190
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 60.50447069634289,
+      "learning_rate": 4.372781065088757e-07,
+      "logits/chosen": -3.625,
+      "logits/rejected": -3.46875,
+      "logps/chosen": -100.0,
+      "logps/rejected": -112.0,
+      "loss": 0.4606,
+      "rewards/accuracies": 0.7749999761581421,
+      "rewards/chosen": 0.162109375,
+      "rewards/margins": 1.375,
+      "rewards/rejected": -1.2109375,
+      "step": 200
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 42.756358795311726,
+      "learning_rate": 4.313609467455621e-07,
+      "logits/chosen": -3.59375,
+      "logits/rejected": -3.46875,
+      "logps/chosen": -92.5,
+      "logps/rejected": -119.0,
+      "loss": 0.3956,
+      "rewards/accuracies": 0.7875000238418579,
+      "rewards/chosen": -0.058837890625,
+      "rewards/margins": 1.7578125,
+      "rewards/rejected": -1.8203125,
+      "step": 210
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 37.12579027519649,
+      "learning_rate": 4.2544378698224847e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.484375,
+      "logps/chosen": -101.5,
+      "logps/rejected": -112.0,
+      "loss": 0.4621,
+      "rewards/accuracies": 0.7875000238418579,
+      "rewards/chosen": 0.06884765625,
+      "rewards/margins": 1.3046875,
+      "rewards/rejected": -1.234375,
+      "step": 220
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 34.37003470756731,
+      "learning_rate": 4.195266272189349e-07,
+      "logits/chosen": -3.40625,
+      "logits/rejected": -3.421875,
+      "logps/chosen": -102.0,
+      "logps/rejected": -119.0,
+      "loss": 0.4478,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": -0.06640625,
+      "rewards/margins": 1.59375,
+      "rewards/rejected": -1.6640625,
+      "step": 230
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 39.975701583088714,
+      "learning_rate": 4.1360946745562133e-07,
+      "logits/chosen": -3.484375,
+      "logits/rejected": -3.4375,
+      "logps/chosen": -95.5,
+      "logps/rejected": -123.0,
+      "loss": 0.4043,
+      "rewards/accuracies": 0.762499988079071,
+      "rewards/chosen": 0.0322265625,
+      "rewards/margins": 1.5546875,
+      "rewards/rejected": -1.5234375,
+      "step": 240
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 40.31896174589532,
+      "learning_rate": 4.076923076923077e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.484375,
+      "logps/chosen": -107.0,
+      "logps/rejected": -122.0,
+      "loss": 0.4742,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.54296875,
+      "rewards/margins": 2.46875,
+      "rewards/rejected": -1.921875,
+      "step": 250
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 32.53695466017645,
+      "learning_rate": 4.017751479289941e-07,
+      "logits/chosen": -3.578125,
+      "logits/rejected": -3.5,
+      "logps/chosen": -81.0,
+      "logps/rejected": -135.0,
+      "loss": 0.3528,
+      "rewards/accuracies": 0.8500000238418579,
+      "rewards/chosen": 0.044921875,
+      "rewards/margins": 2.765625,
+      "rewards/rejected": -2.71875,
+      "step": 260
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 35.58565159605434,
+      "learning_rate": 3.9585798816568044e-07,
+      "logits/chosen": -3.625,
+      "logits/rejected": -3.46875,
+      "logps/chosen": -106.5,
+      "logps/rejected": -107.5,
+      "loss": 0.3774,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 0.318359375,
+      "rewards/margins": 1.8984375,
+      "rewards/rejected": -1.578125,
+      "step": 270
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 50.853618748566205,
+      "learning_rate": 3.8994082840236685e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.484375,
+      "logps/chosen": -99.5,
+      "logps/rejected": -119.0,
+      "loss": 0.4586,
+      "rewards/accuracies": 0.7124999761581421,
+      "rewards/chosen": 0.26953125,
+      "rewards/margins": 1.671875,
+      "rewards/rejected": -1.40625,
+      "step": 280
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 49.95783582965489,
+      "learning_rate": 3.840236686390532e-07,
+      "logits/chosen": -3.640625,
+      "logits/rejected": -3.453125,
+      "logps/chosen": -100.5,
+      "logps/rejected": -113.0,
+      "loss": 0.4365,
+      "rewards/accuracies": 0.7875000238418579,
+      "rewards/chosen": 0.37109375,
+      "rewards/margins": 1.8671875,
+      "rewards/rejected": -1.5,
+      "step": 290
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 26.77848071496543,
+      "learning_rate": 3.7810650887573966e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.59375,
+      "logps/chosen": -99.5,
+      "logps/rejected": -135.0,
+      "loss": 0.4772,
+      "rewards/accuracies": 0.8374999761581421,
+      "rewards/chosen": 0.25390625,
+      "rewards/margins": 2.046875,
+      "rewards/rejected": -1.7890625,
+      "step": 300
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 44.271261713875845,
+      "learning_rate": 3.72189349112426e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.484375,
+      "logps/chosen": -108.5,
+      "logps/rejected": -122.5,
+      "loss": 0.3872,
+      "rewards/accuracies": 0.800000011920929,
+      "rewards/chosen": 0.0625,
+      "rewards/margins": 1.625,
+      "rewards/rejected": -1.5703125,
+      "step": 310
+    },
+    {
+      "epoch": 1.0,
+      "eval_logits/chosen": -3.5625,
+      "eval_logits/rejected": -3.5,
+      "eval_logps/chosen": -98.5,
+      "eval_logps/rejected": -125.0,
+      "eval_loss": 0.5349218845367432,
+      "eval_rewards/accuracies": 0.7857142686843872,
+      "eval_rewards/chosen": -0.12060546875,
+      "eval_rewards/margins": 1.203125,
+      "eval_rewards/rejected": -1.3203125,
+      "eval_runtime": 12.3829,
+      "eval_samples_per_second": 16.151,
+      "eval_steps_per_second": 0.565,
+      "step": 313
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 33.843167460384336,
+      "learning_rate": 3.662721893491124e-07,
+      "logits/chosen": -3.5,
+      "logits/rejected": -3.328125,
+      "logps/chosen": -98.0,
+      "logps/rejected": -138.0,
+      "loss": 0.19,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 0.236328125,
+      "rewards/margins": 2.96875,
+      "rewards/rejected": -2.734375,
+      "step": 320
+    },
+    {
+      "epoch": 1.0543130990415335,
+      "grad_norm": 14.421132534916795,
+      "learning_rate": 3.603550295857988e-07,
+      "logits/chosen": -3.546875,
+      "logits/rejected": -3.4375,
+      "logps/chosen": -101.0,
+      "logps/rejected": -125.0,
+      "loss": 0.148,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.70703125,
+      "rewards/margins": 2.953125,
+      "rewards/rejected": -2.234375,
+      "step": 330
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 18.495628015771892,
+      "learning_rate": 3.544378698224852e-07,
+      "logits/chosen": -3.484375,
+      "logits/rejected": -3.4375,
+      "logps/chosen": -95.5,
+      "logps/rejected": -125.5,
+      "loss": 0.1841,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.71484375,
+      "rewards/margins": 2.875,
+      "rewards/rejected": -2.15625,
+      "step": 340
+    },
+    {
+      "epoch": 1.1182108626198084,
+      "grad_norm": 24.09426865119445,
+      "learning_rate": 3.485207100591716e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.34375,
+      "logps/chosen": -101.0,
+      "logps/rejected": -124.5,
+      "loss": 0.1584,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.89453125,
+      "rewards/margins": 3.421875,
+      "rewards/rejected": -2.53125,
+      "step": 350
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 25.308356305497963,
+      "learning_rate": 3.42603550295858e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.234375,
+      "logps/chosen": -99.0,
+      "logps/rejected": -116.5,
+      "loss": 0.1927,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 0.80078125,
+      "rewards/margins": 3.078125,
+      "rewards/rejected": -2.28125,
+      "step": 360
+    },
+    {
+      "epoch": 1.182108626198083,
+      "grad_norm": 20.87093717123946,
+      "learning_rate": 3.366863905325444e-07,
+      "logits/chosen": -3.5,
+      "logits/rejected": -3.390625,
+      "logps/chosen": -91.5,
+      "logps/rejected": -129.0,
+      "loss": 0.1332,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.734375,
+      "rewards/margins": 3.484375,
+      "rewards/rejected": -2.75,
+      "step": 370
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 33.11495125656759,
+      "learning_rate": 3.3076923076923075e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.375,
+      "logps/chosen": -99.0,
+      "logps/rejected": -148.0,
+      "loss": 0.1292,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.578125,
+      "rewards/margins": 4.4375,
+      "rewards/rejected": -3.859375,
+      "step": 380
+    },
+    {
+      "epoch": 1.2460063897763578,
+      "grad_norm": 67.10626864365061,
+      "learning_rate": 3.2485207100591716e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.296875,
+      "logps/chosen": -95.0,
+      "logps/rejected": -138.0,
+      "loss": 0.2115,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.765625,
+      "rewards/margins": 3.9375,
+      "rewards/rejected": -3.171875,
+      "step": 390
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 18.06304275803513,
+      "learning_rate": 3.189349112426035e-07,
+      "logits/chosen": -3.421875,
+      "logits/rejected": -3.28125,
+      "logps/chosen": -100.0,
+      "logps/rejected": -136.0,
+      "loss": 0.1318,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.5234375,
+      "rewards/margins": 3.828125,
+      "rewards/rejected": -3.3125,
+      "step": 400
+    },
+    {
+      "epoch": 1.3099041533546325,
+      "grad_norm": 38.28309104781088,
+      "learning_rate": 3.130177514792899e-07,
+      "logits/chosen": -3.40625,
+      "logits/rejected": -3.46875,
+      "logps/chosen": -90.0,
+      "logps/rejected": -130.0,
+      "loss": 0.138,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 0.53125,
+      "rewards/margins": 3.578125,
+      "rewards/rejected": -3.046875,
+      "step": 410
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 15.178754681689357,
+      "learning_rate": 3.071005917159763e-07,
+      "logits/chosen": -3.4375,
+      "logits/rejected": -3.40625,
+      "logps/chosen": -92.5,
+      "logps/rejected": -137.0,
+      "loss": 0.1377,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.388671875,
+      "rewards/margins": 3.890625,
+      "rewards/rejected": -3.5,
+      "step": 420
+    },
+    {
+      "epoch": 1.3738019169329074,
+      "grad_norm": 38.06252515615029,
+      "learning_rate": 3.011834319526627e-07,
+      "logits/chosen": -3.5,
+      "logits/rejected": -3.484375,
+      "logps/chosen": -102.5,
+      "logps/rejected": -136.0,
+      "loss": 0.1334,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.5078125,
+      "rewards/margins": 3.71875,
+      "rewards/rejected": -3.203125,
+      "step": 430
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 19.397383343646702,
+      "learning_rate": 2.952662721893491e-07,
+      "logits/chosen": -3.609375,
+      "logits/rejected": -3.53125,
+      "logps/chosen": -102.0,
+      "logps/rejected": -131.0,
+      "loss": 0.1618,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.95703125,
+      "rewards/margins": 3.984375,
+      "rewards/rejected": -3.03125,
+      "step": 440
+    },
+    {
+      "epoch": 1.4376996805111821,
+      "grad_norm": 21.51659900182813,
+      "learning_rate": 2.893491124260355e-07,
+      "logits/chosen": -3.4375,
+      "logits/rejected": -3.359375,
+      "logps/chosen": -97.0,
+      "logps/rejected": -147.0,
+      "loss": 0.1678,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 0.291015625,
+      "rewards/margins": 4.28125,
+      "rewards/rejected": -3.96875,
+      "step": 450
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 4.91974930909796,
+      "learning_rate": 2.834319526627219e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.46875,
+      "logps/chosen": -87.0,
+      "logps/rejected": -159.0,
+      "loss": 0.1759,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.140625,
+      "rewards/margins": 4.0625,
+      "rewards/rejected": -3.921875,
+      "step": 460
+    },
+    {
+      "epoch": 1.5015974440894568,
+      "grad_norm": 65.85227279676343,
+      "learning_rate": 2.7751479289940824e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.3125,
+      "logps/chosen": -94.0,
+      "logps/rejected": -149.0,
+      "loss": 0.142,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.2373046875,
+      "rewards/margins": 4.03125,
+      "rewards/rejected": -3.8125,
+      "step": 470
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 18.354377975306473,
+      "learning_rate": 2.715976331360947e-07,
+      "logits/chosen": -3.46875,
+      "logits/rejected": -3.375,
+      "logps/chosen": -98.0,
+      "logps/rejected": -155.0,
+      "loss": 0.1047,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.287109375,
+      "rewards/margins": 4.625,
+      "rewards/rejected": -4.34375,
+      "step": 480
+    },
+    {
+      "epoch": 1.5654952076677318,
+      "grad_norm": 16.420773639490726,
+      "learning_rate": 2.6568047337278106e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.390625,
+      "logps/chosen": -100.0,
+      "logps/rejected": -130.0,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.76171875,
+      "rewards/margins": 4.125,
+      "rewards/rejected": -3.359375,
+      "step": 490
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 43.48032898852333,
+      "learning_rate": 2.5976331360946746e-07,
+      "logits/chosen": -3.375,
+      "logits/rejected": -3.359375,
+      "logps/chosen": -94.0,
+      "logps/rejected": -151.0,
+      "loss": 0.1149,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.1669921875,
+      "rewards/margins": 4.75,
+      "rewards/rejected": -4.59375,
+      "step": 500
+    },
+    {
+      "epoch": 1.6293929712460065,
+      "grad_norm": 39.88152223306465,
+      "learning_rate": 2.538461538461538e-07,
+      "logits/chosen": -3.390625,
+      "logits/rejected": -3.328125,
+      "logps/chosen": -84.5,
+      "logps/rejected": -157.0,
+      "loss": 0.1571,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.201171875,
+      "rewards/margins": 4.1875,
+      "rewards/rejected": -4.375,
+      "step": 510
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 16.23543571951759,
+      "learning_rate": 2.479289940828402e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.390625,
+      "logps/chosen": -97.0,
+      "logps/rejected": -145.0,
+      "loss": 0.1344,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.40234375,
+      "rewards/margins": 4.15625,
+      "rewards/rejected": -3.765625,
+      "step": 520
+    },
+    {
+      "epoch": 1.6932907348242812,
+      "grad_norm": 30.6571237995008,
+      "learning_rate": 2.420118343195266e-07,
+      "logits/chosen": -3.390625,
+      "logits/rejected": -3.1875,
+      "logps/chosen": -87.5,
+      "logps/rejected": -152.0,
+      "loss": 0.1264,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.0791015625,
+      "rewards/margins": 4.5,
+      "rewards/rejected": -4.40625,
+      "step": 530
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 13.821155728848668,
+      "learning_rate": 2.36094674556213e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.25,
+      "logps/chosen": -108.0,
+      "logps/rejected": -135.0,
+      "loss": 0.1009,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.71875,
+      "rewards/margins": 4.46875,
+      "rewards/rejected": -3.75,
+      "step": 540
+    },
+    {
+      "epoch": 1.7571884984025559,
+      "grad_norm": 25.951691798178203,
+      "learning_rate": 2.301775147928994e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.25,
+      "logps/chosen": -99.5,
+      "logps/rejected": -141.0,
+      "loss": 0.1441,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.4921875,
+      "rewards/margins": 5.0,
+      "rewards/rejected": -4.5,
+      "step": 550
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 12.667867070291107,
+      "learning_rate": 2.242603550295858e-07,
+      "logits/chosen": -3.34375,
+      "logits/rejected": -3.25,
+      "logps/chosen": -103.0,
+      "logps/rejected": -158.0,
+      "loss": 0.1124,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 0.07958984375,
+      "rewards/margins": 4.6875,
+      "rewards/rejected": -4.59375,
+      "step": 560
+    },
+    {
+      "epoch": 1.8210862619808306,
+      "grad_norm": 66.98788513046767,
+      "learning_rate": 2.1834319526627217e-07,
+      "logits/chosen": -3.484375,
+      "logits/rejected": -3.359375,
+      "logps/chosen": -100.5,
+      "logps/rejected": -148.0,
+      "loss": 0.1575,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 0.44140625,
+      "rewards/margins": 4.125,
+      "rewards/rejected": -3.6875,
+      "step": 570
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 22.80146314369706,
+      "learning_rate": 2.1242603550295858e-07,
+      "logits/chosen": -3.515625,
+      "logits/rejected": -3.28125,
+      "logps/chosen": -103.5,
+      "logps/rejected": -146.0,
+      "loss": 0.146,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 0.392578125,
+      "rewards/margins": 4.375,
+      "rewards/rejected": -3.984375,
+      "step": 580
+    },
+    {
+      "epoch": 1.8849840255591053,
+      "grad_norm": 22.73841977980904,
+      "learning_rate": 2.0650887573964496e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.34375,
+      "logps/chosen": -103.5,
+      "logps/rejected": -149.0,
+      "loss": 0.1338,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 0.30078125,
+      "rewards/margins": 4.5625,
+      "rewards/rejected": -4.25,
+      "step": 590
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 38.211271754825084,
+      "learning_rate": 2.0059171597633133e-07,
+      "logits/chosen": -3.5625,
+      "logits/rejected": -3.296875,
+      "logps/chosen": -101.5,
+      "logps/rejected": -138.0,
+      "loss": 0.1455,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 0.578125,
+      "rewards/margins": 4.53125,
+      "rewards/rejected": -3.96875,
+      "step": 600
+    },
+    {
+      "epoch": 1.9488817891373802,
+      "grad_norm": 20.886358667126427,
+      "learning_rate": 1.9467455621301774e-07,
+      "logits/chosen": -3.21875,
+      "logits/rejected": -3.296875,
+      "logps/chosen": -86.5,
+      "logps/rejected": -165.0,
+      "loss": 0.1428,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.004638671875,
+      "rewards/margins": 4.40625,
+      "rewards/rejected": -4.40625,
+      "step": 610
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 26.719930163743193,
+      "learning_rate": 1.8875739644970412e-07,
+      "logits/chosen": -3.421875,
+      "logits/rejected": -3.28125,
+      "logps/chosen": -101.5,
+      "logps/rejected": -154.0,
+      "loss": 0.1191,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.29296875,
+      "rewards/margins": 5.15625,
+      "rewards/rejected": -4.875,
+      "step": 620
+    },
+    {
+      "epoch": 2.0,
+      "eval_logits/chosen": -3.4375,
+      "eval_logits/rejected": -3.34375,
+      "eval_logps/chosen": -108.5,
+      "eval_logps/rejected": -137.0,
+      "eval_loss": 0.6070312261581421,
+      "eval_rewards/accuracies": 0.6964285969734192,
+      "eval_rewards/chosen": -1.125,
+      "eval_rewards/margins": 1.3828125,
+      "eval_rewards/rejected": -2.515625,
+      "eval_runtime": 12.3074,
+      "eval_samples_per_second": 16.25,
+      "eval_steps_per_second": 0.569,
+      "step": 626
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 233.3927856017294,
+      "learning_rate": 1.8284023668639053e-07,
+      "logits/chosen": -3.453125,
+      "logits/rejected": -3.390625,
+      "logps/chosen": -90.0,
+      "logps/rejected": -156.0,
+      "loss": 0.1624,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 0.060791015625,
+      "rewards/margins": 4.25,
+      "rewards/rejected": -4.21875,
+      "step": 630
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 6.8655748966408305,
+      "learning_rate": 1.7692307692307693e-07,
+      "logits/chosen": -3.40625,
+      "logits/rejected": -3.28125,
+      "logps/chosen": -102.0,
+      "logps/rejected": -150.0,
+      "loss": 0.0546,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.51953125,
+      "rewards/margins": 5.3125,
+      "rewards/rejected": -4.8125,
+      "step": 640
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 5.52148734018718,
+      "learning_rate": 1.710059171597633e-07,
+      "logits/chosen": -3.46875,
+      "logits/rejected": -3.125,
+      "logps/chosen": -103.0,
+      "logps/rejected": -152.0,
+      "loss": 0.0275,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.64453125,
+      "rewards/margins": 5.8125,
+      "rewards/rejected": -5.1875,
+      "step": 650
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 5.945489873996584,
+      "learning_rate": 1.650887573964497e-07,
+      "logits/chosen": -3.53125,
+      "logits/rejected": -3.140625,
+      "logps/chosen": -94.0,
+      "logps/rejected": -163.0,
+      "loss": 0.0346,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.3125,
+      "rewards/margins": 6.28125,
+      "rewards/rejected": -5.96875,
+      "step": 660
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 15.641153990271073,
+      "learning_rate": 1.591715976331361e-07,
+      "logits/chosen": -3.46875,
+      "logits/rejected": -3.203125,
+      "logps/chosen": -95.0,
+      "logps/rejected": -172.0,
+      "loss": 0.0404,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.28125,
+      "rewards/margins": 6.5625,
+      "rewards/rejected": -6.28125,
+      "step": 670
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 12.428538162453988,
+      "learning_rate": 1.5325443786982248e-07,
+      "logits/chosen": -3.265625,
+      "logits/rejected": -3.109375,
+      "logps/chosen": -107.5,
+      "logps/rejected": -170.0,
+      "loss": 0.037,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.265625,
+      "rewards/margins": 6.625,
+      "rewards/rejected": -6.375,
+      "step": 680
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 4.017973551269618,
+      "learning_rate": 1.4733727810650885e-07,
+      "logits/chosen": -3.25,
+      "logits/rejected": -3.109375,
+      "logps/chosen": -108.0,
+      "logps/rejected": -173.0,
+      "loss": 0.037,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 0.302734375,
+      "rewards/margins": 7.0,
+      "rewards/rejected": -6.6875,
+      "step": 690
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 86.48671040112693,
+      "learning_rate": 1.4142011834319526e-07,
+      "logits/chosen": -3.28125,
+      "logits/rejected": -3.0625,
+      "logps/chosen": -97.0,
+      "logps/rejected": -173.0,
+      "loss": 0.0335,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.2236328125,
+      "rewards/margins": 7.4375,
+      "rewards/rejected": -7.65625,
+      "step": 700
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 3.430888312255222,
+      "learning_rate": 1.3550295857988164e-07,
+      "logits/chosen": -3.15625,
+      "logits/rejected": -3.125,
+      "logps/chosen": -108.0,
+      "logps/rejected": -192.0,
+      "loss": 0.0355,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.9765625,
+      "rewards/margins": 7.53125,
+      "rewards/rejected": -8.5,
+      "step": 710
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 7.379623913134848,
+      "learning_rate": 1.2958579881656802e-07,
+      "logits/chosen": -3.34375,
+      "logits/rejected": -3.25,
+      "logps/chosen": -101.5,
+      "logps/rejected": -163.0,
+      "loss": 0.0452,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 0.0230712890625,
+      "rewards/margins": 6.53125,
+      "rewards/rejected": -6.5,
+      "step": 720
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 16.067536241252174,
+      "learning_rate": 1.2366863905325443e-07,
+      "logits/chosen": -3.40625,
+      "logits/rejected": -3.25,
+      "logps/chosen": -104.0,
+      "logps/rejected": -183.0,
+      "loss": 0.0485,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -0.265625,
+      "rewards/margins": 7.0625,
+      "rewards/rejected": -7.3125,
+      "step": 730
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 8.775321084313918,
+      "learning_rate": 1.1775147928994082e-07,
+      "logits/chosen": -3.25,
+      "logits/rejected": -3.1875,
+      "logps/chosen": -101.5,
+      "logps/rejected": -198.0,
+      "loss": 0.0573,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -1.109375,
+      "rewards/margins": 7.1875,
+      "rewards/rejected": -8.3125,
+      "step": 740
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 2.7323617705921777,
+      "learning_rate": 1.1183431952662721e-07,
+      "logits/chosen": -3.203125,
+      "logits/rejected": -3.125,
+      "logps/chosen": -101.0,
+      "logps/rejected": -190.0,
+      "loss": 0.0264,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -0.60546875,
+      "rewards/margins": 8.0,
+      "rewards/rejected": -8.625,
+      "step": 750
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 26.885338666699912,
+      "learning_rate": 1.059171597633136e-07,
+      "logits/chosen": -3.296875,
+      "logits/rejected": -3.15625,
+      "logps/chosen": -100.0,
+      "logps/rejected": -168.0,
+      "loss": 0.0437,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": -0.12255859375,
+      "rewards/margins": 6.75,
+      "rewards/rejected": -6.875,
+      "step": 760
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 4.85033080386058,
+      "learning_rate": 1e-07,
+      "logits/chosen": -3.171875,
+      "logits/rejected": -3.265625,
+      "logps/chosen": -97.0,
+      "logps/rejected": -188.0,
+      "loss": 0.0788,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -0.474609375,
+      "rewards/margins": 7.6875,
+      "rewards/rejected": -8.125,
+      "step": 770
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 205.54438620209856,
+      "learning_rate": 9.408284023668639e-08,
+      "logits/chosen": -3.25,
+      "logits/rejected": -3.234375,
+      "logps/chosen": -96.0,
+      "logps/rejected": -186.0,
+      "loss": 0.0678,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.177734375,
+      "rewards/margins": 7.3125,
+      "rewards/rejected": -7.5,
+      "step": 780
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 28.433975567152338,
+      "learning_rate": 8.816568047337278e-08,
+      "logits/chosen": -3.15625,
+      "logits/rejected": -3.03125,
+      "logps/chosen": -98.5,
+      "logps/rejected": -188.0,
+      "loss": 0.031,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -0.91796875,
+      "rewards/margins": 7.5,
+      "rewards/rejected": -8.4375,
+      "step": 790
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 4.2386789501934965,
+      "learning_rate": 8.224852071005916e-08,
+      "logits/chosen": -3.265625,
+      "logits/rejected": -3.03125,
+      "logps/chosen": -110.5,
+      "logps/rejected": -175.0,
+      "loss": 0.0499,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -0.353515625,
+      "rewards/margins": 7.15625,
+      "rewards/rejected": -7.5,
+      "step": 800
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 6.261571908946508,
+      "learning_rate": 7.633136094674555e-08,
+      "logits/chosen": -3.234375,
+      "logits/rejected": -3.09375,
+      "logps/chosen": -104.5,
+      "logps/rejected": -185.0,
+      "loss": 0.0425,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -0.7890625,
+      "rewards/margins": 7.3125,
+      "rewards/rejected": -8.125,
+      "step": 810
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 43.7884576040152,
+      "learning_rate": 7.041420118343195e-08,
+      "logits/chosen": -3.15625,
+      "logits/rejected": -3.140625,
+      "logps/chosen": -107.0,
+      "logps/rejected": -193.0,
+      "loss": 0.0592,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -0.404296875,
+      "rewards/margins": 7.59375,
+      "rewards/rejected": -8.0,
+      "step": 820
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 7.063030707335778,
+      "learning_rate": 6.449704142011835e-08,
+      "logits/chosen": -3.203125,
+      "logits/rejected": -3.25,
+      "logps/chosen": -107.5,
+      "logps/rejected": -219.0,
+      "loss": 0.0132,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.0,
+      "rewards/margins": 8.6875,
+      "rewards/rejected": -9.6875,
+      "step": 830
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 31.69985105836376,
+      "learning_rate": 5.857988165680473e-08,
+      "logits/chosen": -3.234375,
+      "logits/rejected": -3.125,
+      "logps/chosen": -110.0,
+      "logps/rejected": -194.0,
+      "loss": 0.0767,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -1.40625,
+      "rewards/margins": 7.375,
+      "rewards/rejected": -8.8125,
+      "step": 840
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 36.344404095239184,
+      "learning_rate": 5.266272189349112e-08,
+      "logits/chosen": -3.171875,
+      "logits/rejected": -3.15625,
+      "logps/chosen": -110.0,
+      "logps/rejected": -196.0,
+      "loss": 0.0383,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.234375,
+      "rewards/margins": 7.71875,
+      "rewards/rejected": -8.9375,
+      "step": 850
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 26.561166415223443,
+      "learning_rate": 4.674556213017751e-08,
+      "logits/chosen": -3.296875,
+      "logits/rejected": -3.328125,
+      "logps/chosen": -109.5,
+      "logps/rejected": -183.0,
+      "loss": 0.0619,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -1.28125,
+      "rewards/margins": 6.5625,
+      "rewards/rejected": -7.84375,
+      "step": 860
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 51.634391271169434,
+      "learning_rate": 4.082840236686391e-08,
+      "logits/chosen": -3.34375,
+      "logits/rejected": -3.078125,
+      "logps/chosen": -115.5,
+      "logps/rejected": -175.0,
+      "loss": 0.0302,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -0.75,
+      "rewards/margins": 7.78125,
+      "rewards/rejected": -8.5625,
+      "step": 870
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 4.862668702951866,
+      "learning_rate": 3.4911242603550294e-08,
+      "logits/chosen": -3.109375,
+      "logits/rejected": -3.1875,
+      "logps/chosen": -110.0,
+      "logps/rejected": -192.0,
+      "loss": 0.0318,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.3125,
+      "rewards/margins": 8.125,
+      "rewards/rejected": -9.375,
+      "step": 880
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 5.808427284559429,
+      "learning_rate": 2.8994082840236687e-08,
+      "logits/chosen": -3.09375,
+      "logits/rejected": -3.21875,
+      "logps/chosen": -118.5,
+      "logps/rejected": -202.0,
+      "loss": 0.0413,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.6171875,
+      "rewards/margins": 8.25,
+      "rewards/rejected": -9.8125,
+      "step": 890
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 1.47350840913393,
+      "learning_rate": 2.3076923076923076e-08,
+      "logits/chosen": -3.21875,
+      "logits/rejected": -3.171875,
+      "logps/chosen": -102.5,
+      "logps/rejected": -202.0,
+      "loss": 0.024,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -1.3046875,
+      "rewards/margins": 8.4375,
+      "rewards/rejected": -9.75,
+      "step": 900
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 35.85817350730659,
+      "learning_rate": 1.7159763313609465e-08,
+      "logits/chosen": -3.265625,
+      "logits/rejected": -3.234375,
+      "logps/chosen": -116.5,
+      "logps/rejected": -181.0,
+      "loss": 0.0377,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": -0.474609375,
+      "rewards/margins": 7.96875,
+      "rewards/rejected": -8.4375,
+      "step": 910
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 2.972228065765212,
+      "learning_rate": 1.1242603550295858e-08,
+      "logits/chosen": -3.1875,
+      "logits/rejected": -3.28125,
+      "logps/chosen": -107.5,
+      "logps/rejected": -200.0,
+      "loss": 0.0497,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": -1.109375,
+      "rewards/margins": 8.0625,
+      "rewards/rejected": -9.1875,
+      "step": 920
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 41.76203114287835,
+      "learning_rate": 5.325443786982248e-09,
+      "logits/chosen": -3.125,
+      "logits/rejected": -3.03125,
+      "logps/chosen": -96.5,
+      "logps/rejected": -183.0,
+      "loss": 0.0546,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": -1.0234375,
+      "rewards/margins": 7.71875,
+      "rewards/rejected": -8.75,
+      "step": 930
+    },
+    {
+      "epoch": 3.0,
+      "eval_logits/chosen": -3.328125,
+      "eval_logits/rejected": -3.1875,
+      "eval_logps/chosen": -133.0,
+      "eval_logps/rejected": -166.0,
+      "eval_loss": 0.8457812666893005,
+      "eval_rewards/accuracies": 0.6785714030265808,
+      "eval_rewards/chosen": -3.546875,
+      "eval_rewards/margins": 1.8359375,
+      "eval_rewards/rejected": -5.40625,
+      "eval_runtime": 15.4802,
+      "eval_samples_per_second": 12.92,
+      "eval_steps_per_second": 0.452,
+      "step": 939
+    },
+    {
+      "epoch": 3.0,
+      "step": 939,
+      "total_flos": 0.0,
+      "train_loss": 0.23619326285787792,
+      "train_runtime": 4850.984,
+      "train_samples_per_second": 6.184,
+      "train_steps_per_second": 0.194
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 939,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "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": 8,
+  "trial_name": null,
+  "trial_params": null
+}