Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-1125/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,91 @@
+---
+base_model: mistralai/Mistral-Nemo-Instruct-2407
+library_name: peft
+license: other
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: sft_dpo_fs
+  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. -->
+
+# sft_dpo_fs
+
+This model is a fine-tuned version of [mistralai/Mistral-Nemo-Instruct-2407](https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407) on the heat_transfer_dpo dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.1535
+- Rewards/chosen: 17.2823
+- Rewards/rejected: 11.3004
+- Rewards/accuracies: 0.9610
+- Rewards/margins: 5.9819
+- Logps/chosen: -2.2063
+- Logps/rejected: -60.6033
+- Logits/chosen: 0.0035
+- Logits/rejected: -0.0076
+
+## 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-06
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- total_train_batch_size: 8
+- total_eval_batch_size: 8
+- 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 | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/chosen | Logps/rejected | Logits/chosen | Logits/rejected |
+|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:------------:|:--------------:|:-------------:|:---------------:|
+| 0.3835        | 0.0533 | 60   | 0.3287          | 17.1482        | 15.7095          | 0.9280             | 1.4386          | -3.5472      | -16.5118       | -0.5827       | -0.5914         |
+| 0.2552        | 0.1067 | 120  | 0.1900          | 17.1335        | 13.7535          | 0.9320             | 3.3799          | -3.6944      | -36.0722       | -0.2065       | -0.2218         |
+| 0.2362        | 0.16   | 180  | 0.2024          | 17.0614        | 11.9722          | 0.9510             | 5.0892          | -4.4150      | -53.8850       | -0.1087       | -0.1222         |
+| 0.1781        | 0.2133 | 240  | 0.1546          | 17.0620        | 12.2862          | 0.9500             | 4.7758          | -4.4089      | -50.7448       | -0.1243       | -0.1381         |
+| 0.265         | 0.2667 | 300  | 0.1536          | 17.2493        | 12.6444          | 0.9440             | 4.6050          | -2.5355      | -47.1637       | -0.1744       | -0.1856         |
+| 0.1605        | 0.32   | 360  | 0.3194          | 17.3612        | 12.2655          | 0.9210             | 5.0958          | -1.4165      | -50.9525       | -0.1062       | -0.1173         |
+| 0.2894        | 0.3733 | 420  | 0.1679          | 17.3116        | 12.2496          | 0.9450             | 5.0620          | -1.9131      | -51.1113       | -0.0905       | -0.1026         |
+| 0.1149        | 0.4267 | 480  | 0.2951          | 17.0540        | 11.9844          | 0.9230             | 5.0696          | -4.4890      | -53.7628       | -0.0770       | -0.0883         |
+| 0.0384        | 0.48   | 540  | 0.1739          | 17.2042        | 12.1334          | 0.9490             | 5.0708          | -2.9873      | -52.2731       | -0.0512       | -0.0612         |
+| 0.4008        | 0.5333 | 600  | 0.1706          | 17.2853        | 11.6981          | 0.9470             | 5.5872          | -2.1760      | -56.6266       | -0.0358       | -0.0469         |
+| 0.1678        | 0.5867 | 660  | 0.2050          | 17.2021        | 11.5656          | 0.9450             | 5.6365          | -3.0082      | -57.9516       | -0.0160       | -0.0270         |
+| 0.2272        | 0.64   | 720  | 0.1402          | 17.3928        | 11.7696          | 0.9520             | 5.6233          | -1.1005      | -55.9117       | -0.0229       | -0.0322         |
+| 0.1915        | 0.6933 | 780  | 0.2441          | 17.3947        | 11.7656          | 0.9320             | 5.6290          | -1.0823      | -55.9507       | -0.0166       | -0.0266         |
+| 0.0635        | 0.7467 | 840  | 0.1689          | 17.3812        | 11.5343          | 0.9450             | 5.8469          | -1.2169      | -58.2643       | -0.0111       | -0.0217         |
+| 0.1703        | 0.8    | 900  | 0.1400          | 17.3271        | 11.3817          | 0.9610             | 5.9455          | -1.7577      | -59.7906       | 0.0002        | -0.0105         |
+| 0.1138        | 0.8533 | 960  | 0.1441          | 17.3149        | 11.3432          | 0.9630             | 5.9718          | -1.8795      | -60.1756       | 0.0015        | -0.0094         |
+| 0.0513        | 0.9067 | 1020 | 0.1412          | 17.3211        | 11.3263          | 0.9610             | 5.9948          | -1.8178      | -60.3445       | 0.0045        | -0.0065         |
+| 0.1189        | 0.96   | 1080 | 0.1508          | 17.2887        | 11.3001          | 0.9610             | 5.9886          | -2.1420      | -60.6061       | 0.0074        | -0.0036         |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.46.0
+- Pytorch 2.4.0+cu121
+- Datasets 2.21.0
+- Tokenizers 0.20.1

adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "down_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6404a5e1c60aa26a99a4ccbb402f485483d5eea90c08441f0448e29c51e2b1d0
+size 114106856

all_results.json

@@ -0,0 +1,20 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": 0.0034948738757520914,
+    "eval_logits/rejected": -0.007565807551145554,
+    "eval_logps/chosen": -2.20627498626709,
+    "eval_logps/rejected": -60.60332489013672,
+    "eval_loss": 0.15350937843322754,
+    "eval_rewards/accuracies": 0.9610000848770142,
+    "eval_rewards/chosen": 17.28226661682129,
+    "eval_rewards/margins": 5.981877326965332,
+    "eval_rewards/rejected": 11.30038833618164,
+    "eval_runtime": 361.5652,
+    "eval_samples_per_second": 2.766,
+    "eval_steps_per_second": 0.346,
+    "total_flos": 1.4338459346927616e+18,
+    "train_loss": 0.23597783709896936,
+    "train_runtime": 13850.7044,
+    "train_samples_per_second": 0.65,
+    "train_steps_per_second": 0.081
+}

checkpoint-1000/README.md

@@ -0,0 +1,202 @@
+---
+base_model: mistralai/Mistral-Nemo-Instruct-2407
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.12.0

checkpoint-1000/adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "down_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1000/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b661db279de1245ec057ccc9ea7b2a9d8e593601b17b234d71a8b65eea724cc
+size 114106856

checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5df0d088d43a257985e1c6b5e7afebb17236c24d7243b77ad1477e87251acd3
+size 228536930

checkpoint-1000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41d62848396b464be71e7417b09981050d5f1d377e9e0c4894645772a871d81e
+size 14512

checkpoint-1000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a408fa3a85250385243240a3c4f6e6dc4513da1d97421e8faae4c8d880e7d1a
+size 14512

checkpoint-1000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4549dc1554a9cef5b3b89b84239d0624d2e4f7b2af98173873173c77ff786861
+size 1064

checkpoint-1000/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1000/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

checkpoint-1000/trainer_state.json

@@ -0,0 +1,1789 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8888888888888888,
+  "eval_steps": 60,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 8.760091781616211,
+      "learning_rate": 4.4247787610619474e-07,
+      "logits/chosen": -0.8248252868652344,
+      "logits/rejected": -0.8263720273971558,
+      "logps/chosen": -0.36086463928222656,
+      "logps/rejected": -5.696224689483643,
+      "loss": 1.1038,
+      "rewards/accuracies": 0.5125000476837158,
+      "rewards/chosen": 17.43745994567871,
+      "rewards/margins": 0.5984855890274048,
+      "rewards/rejected": 16.838973999023438,
+      "step": 10
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 8.855981826782227,
+      "learning_rate": 8.849557522123895e-07,
+      "logits/chosen": -0.8169006109237671,
+      "logits/rejected": -0.819770872592926,
+      "logps/chosen": -0.12464660406112671,
+      "logps/rejected": -7.139842987060547,
+      "loss": 1.1887,
+      "rewards/accuracies": 0.4000000059604645,
+      "rewards/chosen": 17.17649269104004,
+      "rewards/margins": 0.19107049703598022,
+      "rewards/rejected": 16.98542022705078,
+      "step": 20
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 16.764184951782227,
+      "learning_rate": 1.3274336283185843e-06,
+      "logits/chosen": -0.8003113865852356,
+      "logits/rejected": -0.8030117750167847,
+      "logps/chosen": -0.34651467204093933,
+      "logps/rejected": -6.967917442321777,
+      "loss": 1.0563,
+      "rewards/accuracies": 0.44999998807907104,
+      "rewards/chosen": 17.280975341796875,
+      "rewards/margins": 0.40005987882614136,
+      "rewards/rejected": 16.88091468811035,
+      "step": 30
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 8.33682918548584,
+      "learning_rate": 1.769911504424779e-06,
+      "logits/chosen": -0.7695047855377197,
+      "logits/rejected": -0.7739207148551941,
+      "logps/chosen": -1.5993006229400635,
+      "logps/rejected": -8.504932403564453,
+      "loss": 0.7596,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 17.283912658691406,
+      "rewards/margins": 0.6976072192192078,
+      "rewards/rejected": 16.5863037109375,
+      "step": 40
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 4.494723320007324,
+      "learning_rate": 2.212389380530974e-06,
+      "logits/chosen": -0.7154140472412109,
+      "logits/rejected": -0.7225576043128967,
+      "logps/chosen": -3.112199068069458,
+      "logps/rejected": -12.212080001831055,
+      "loss": 0.6083,
+      "rewards/accuracies": 0.4625000059604645,
+      "rewards/chosen": 17.03064727783203,
+      "rewards/margins": 0.7148451805114746,
+      "rewards/rejected": 16.3158016204834,
+      "step": 50
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 5.110287666320801,
+      "learning_rate": 2.6548672566371687e-06,
+      "logits/chosen": -0.6322453022003174,
+      "logits/rejected": -0.6387485265731812,
+      "logps/chosen": -5.650620460510254,
+      "logps/rejected": -12.759811401367188,
+      "loss": 0.3835,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.101289749145508,
+      "rewards/margins": 1.1824612617492676,
+      "rewards/rejected": 15.918828964233398,
+      "step": 60
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "eval_logits/chosen": -0.5826543569564819,
+      "eval_logits/rejected": -0.5914276838302612,
+      "eval_logps/chosen": -3.5471787452697754,
+      "eval_logps/rejected": -16.51181983947754,
+      "eval_loss": 0.3286525011062622,
+      "eval_rewards/accuracies": 0.9280000925064087,
+      "eval_rewards/chosen": 17.148174285888672,
+      "eval_rewards/margins": 1.4386365413665771,
+      "eval_rewards/rejected": 15.709539413452148,
+      "eval_runtime": 372.0227,
+      "eval_samples_per_second": 2.688,
+      "eval_steps_per_second": 0.336,
+      "step": 60
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 5.098133563995361,
+      "learning_rate": 3.097345132743363e-06,
+      "logits/chosen": -0.5378152132034302,
+      "logits/rejected": -0.5494933724403381,
+      "logps/chosen": -1.5099802017211914,
+      "logps/rejected": -21.206321716308594,
+      "loss": 0.2931,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.083791732788086,
+      "rewards/margins": 1.5844331979751587,
+      "rewards/rejected": 15.499359130859375,
+      "step": 70
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 29.787437438964844,
+      "learning_rate": 3.539823008849558e-06,
+      "logits/chosen": -0.443774938583374,
+      "logits/rejected": -0.45571577548980713,
+      "logps/chosen": -1.5804342031478882,
+      "logps/rejected": -22.606929779052734,
+      "loss": 0.202,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.302125930786133,
+      "rewards/margins": 2.174014091491699,
+      "rewards/rejected": 15.128110885620117,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.14398193359375,
+      "learning_rate": 3.982300884955752e-06,
+      "logits/chosen": -0.3626072406768799,
+      "logits/rejected": -0.3787815570831299,
+      "logps/chosen": -2.203828811645508,
+      "logps/rejected": -29.433551788330078,
+      "loss": 0.2123,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.00284194946289,
+      "rewards/margins": 2.320391893386841,
+      "rewards/rejected": 14.682450294494629,
+      "step": 90
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 29.672739028930664,
+      "learning_rate": 4.424778761061948e-06,
+      "logits/chosen": -0.3035663962364197,
+      "logits/rejected": -0.31762221455574036,
+      "logps/chosen": -3.433589458465576,
+      "logps/rejected": -29.9322509765625,
+      "loss": 0.2592,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.929956436157227,
+      "rewards/margins": 2.31272029876709,
+      "rewards/rejected": 14.617237091064453,
+      "step": 100
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.873722791671753,
+      "learning_rate": 4.867256637168142e-06,
+      "logits/chosen": -0.2679600715637207,
+      "logits/rejected": -0.2826440930366516,
+      "logps/chosen": -0.9653514623641968,
+      "logps/rejected": -30.235322952270508,
+      "loss": 0.1336,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.462385177612305,
+      "rewards/margins": 3.1994175910949707,
+      "rewards/rejected": 14.26296615600586,
+      "step": 110
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.6913721561431885,
+      "learning_rate": 4.999409761242696e-06,
+      "logits/chosen": -0.22222033143043518,
+      "logits/rejected": -0.23720571398735046,
+      "logps/chosen": -4.4953508377075195,
+      "logps/rejected": -34.074745178222656,
+      "loss": 0.2552,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.04866600036621,
+      "rewards/margins": 3.1014418601989746,
+      "rewards/rejected": 13.947224617004395,
+      "step": 120
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "eval_logits/chosen": -0.206527978181839,
+      "eval_logits/rejected": -0.22178640961647034,
+      "eval_logps/chosen": -3.69442081451416,
+      "eval_logps/rejected": -36.072166442871094,
+      "eval_loss": 0.18996010720729828,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.133451461791992,
+      "eval_rewards/margins": 3.379946708679199,
+      "eval_rewards/rejected": 13.753504753112793,
+      "eval_runtime": 361.5279,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 120
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 61.80262756347656,
+      "learning_rate": 4.996519466816778e-06,
+      "logits/chosen": -0.18473535776138306,
+      "logits/rejected": -0.1988501250743866,
+      "logps/chosen": -3.7009687423706055,
+      "logps/rejected": -39.289939880371094,
+      "loss": 0.1394,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.106964111328125,
+      "rewards/margins": 3.633338212966919,
+      "rewards/rejected": 13.473625183105469,
+      "step": 130
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.6732702255249023,
+      "learning_rate": 4.9912234871722805e-06,
+      "logits/chosen": -0.16134041547775269,
+      "logits/rejected": -0.17547868192195892,
+      "logps/chosen": -3.0637736320495605,
+      "logps/rejected": -40.07548522949219,
+      "loss": 0.1408,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.392223358154297,
+      "rewards/margins": 4.242353439331055,
+      "rewards/rejected": 13.149867057800293,
+      "step": 140
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.346453994512558,
+      "learning_rate": 4.98352692559805e-06,
+      "logits/chosen": -0.13797929883003235,
+      "logits/rejected": -0.15283086895942688,
+      "logps/chosen": -5.14492130279541,
+      "logps/rejected": -47.97212219238281,
+      "loss": 0.2153,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.896778106689453,
+      "rewards/margins": 4.227695465087891,
+      "rewards/rejected": 12.669081687927246,
+      "step": 150
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.21871662139892578,
+      "learning_rate": 4.973437198621237e-06,
+      "logits/chosen": -0.12396670132875443,
+      "logits/rejected": -0.13780555129051208,
+      "logps/chosen": -6.108860015869141,
+      "logps/rejected": -54.90739440917969,
+      "loss": 0.0388,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 16.75935935974121,
+      "rewards/margins": 4.755282878875732,
+      "rewards/rejected": 12.004077911376953,
+      "step": 160
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 235.12429809570312,
+      "learning_rate": 4.960964028860621e-06,
+      "logits/chosen": -0.1140839159488678,
+      "logits/rejected": -0.1263057291507721,
+      "logps/chosen": -12.605452537536621,
+      "logps/rejected": -53.81230926513672,
+      "loss": 0.4651,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.101238250732422,
+      "rewards/margins": 3.9864249229431152,
+      "rewards/rejected": 12.114812850952148,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 190.97048950195312,
+      "learning_rate": 4.946119435657738e-06,
+      "logits/chosen": -0.10746976733207703,
+      "logits/rejected": -0.11878640949726105,
+      "logps/chosen": -8.5105562210083,
+      "logps/rejected": -51.314781188964844,
+      "loss": 0.2362,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.719980239868164,
+      "rewards/margins": 4.549674034118652,
+      "rewards/rejected": 12.170306205749512,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "eval_logits/chosen": -0.10870806127786636,
+      "eval_logits/rejected": -0.12223993986845016,
+      "eval_logps/chosen": -4.414996147155762,
+      "eval_logps/rejected": -53.885032653808594,
+      "eval_loss": 0.20236633718013763,
+      "eval_rewards/accuracies": 0.9510000944137573,
+      "eval_rewards/chosen": 17.06139373779297,
+      "eval_rewards/margins": 5.089176177978516,
+      "eval_rewards/rejected": 11.97221851348877,
+      "eval_runtime": 361.4355,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 180
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 56.81266784667969,
+      "learning_rate": 4.928917723494854e-06,
+      "logits/chosen": -0.10682469606399536,
+      "logits/rejected": -0.12124393880367279,
+      "logps/chosen": -3.058413028717041,
+      "logps/rejected": -55.052528381347656,
+      "loss": 0.2442,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.058589935302734,
+      "rewards/margins": 5.056097984313965,
+      "rewards/rejected": 12.002490043640137,
+      "step": 190
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 175.06552124023438,
+      "learning_rate": 4.909375468210947e-06,
+      "logits/chosen": -0.10520349442958832,
+      "logits/rejected": -0.12018950283527374,
+      "logps/chosen": -4.114959716796875,
+      "logps/rejected": -55.9394645690918,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.98603057861328,
+      "rewards/margins": 5.105838775634766,
+      "rewards/rejected": 11.880191802978516,
+      "step": 200
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 78.06558990478516,
+      "learning_rate": 4.8875115010289655e-06,
+      "logits/chosen": -0.10475558042526245,
+      "logits/rejected": -0.11949175596237183,
+      "logps/chosen": -6.760301113128662,
+      "logps/rejected": -53.91607666015625,
+      "loss": 0.2843,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.857545852661133,
+      "rewards/margins": 4.917357921600342,
+      "rewards/rejected": 11.94018840789795,
+      "step": 210
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 15.880486488342285,
+      "learning_rate": 4.863346890409768e-06,
+      "logits/chosen": -0.11213523149490356,
+      "logits/rejected": -0.12581588327884674,
+      "logps/chosen": -6.759585380554199,
+      "logps/rejected": -51.10936737060547,
+      "loss": 0.5104,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.859071731567383,
+      "rewards/margins": 4.638372898101807,
+      "rewards/rejected": 12.220698356628418,
+      "step": 220
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 46.97845458984375,
+      "learning_rate": 4.836904921750224e-06,
+      "logits/chosen": -0.11947059631347656,
+      "logits/rejected": -0.1329912692308426,
+      "logps/chosen": -3.608184814453125,
+      "logps/rejected": -48.794761657714844,
+      "loss": 0.2134,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.235904693603516,
+      "rewards/margins": 4.859888076782227,
+      "rewards/rejected": 12.376014709472656,
+      "step": 230
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 24.032859802246094,
+      "learning_rate": 4.808211074945042e-06,
+      "logits/chosen": -0.1200513243675232,
+      "logits/rejected": -0.1333036869764328,
+      "logps/chosen": -3.7552154064178467,
+      "logps/rejected": -49.87453079223633,
+      "loss": 0.1781,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.094650268554688,
+      "rewards/margins": 4.68077278137207,
+      "rewards/rejected": 12.41387939453125,
+      "step": 240
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "eval_logits/chosen": -0.12433278560638428,
+      "eval_logits/rejected": -0.13808581233024597,
+      "eval_logps/chosen": -4.408891201019287,
+      "eval_logps/rejected": -50.744781494140625,
+      "eval_loss": 0.1546352356672287,
+      "eval_rewards/accuracies": 0.9500000476837158,
+      "eval_rewards/chosen": 17.06200408935547,
+      "eval_rewards/margins": 4.775761604309082,
+      "eval_rewards/rejected": 12.286243438720703,
+      "eval_runtime": 361.4974,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 240
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.25737640261650085,
+      "learning_rate": 4.7772929998339485e-06,
+      "logits/chosen": -0.12348780035972595,
+      "logits/rejected": -0.13704943656921387,
+      "logps/chosen": -4.4299187660217285,
+      "logps/rejected": -53.074607849121094,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.087068557739258,
+      "rewards/margins": 5.06691837310791,
+      "rewards/rejected": 12.020149230957031,
+      "step": 250
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 0.1839389204978943,
+      "learning_rate": 4.744180489557859e-06,
+      "logits/chosen": -0.12177034467458725,
+      "logits/rejected": -0.1342695653438568,
+      "logps/chosen": -3.775188446044922,
+      "logps/rejected": -53.98720932006836,
+      "loss": 0.1896,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.12021255493164,
+      "rewards/margins": 5.148064613342285,
+      "rewards/rejected": 11.972146987915039,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.258485794067383,
+      "learning_rate": 4.708905451849754e-06,
+      "logits/chosen": -0.11067859083414078,
+      "logits/rejected": -0.12377731502056122,
+      "logps/chosen": -6.418317794799805,
+      "logps/rejected": -56.57402801513672,
+      "loss": 0.2315,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.738832473754883,
+      "rewards/margins": 4.884931564331055,
+      "rewards/rejected": 11.853900909423828,
+      "step": 270
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 77.56194305419922,
+      "learning_rate": 4.671501878287879e-06,
+      "logits/chosen": -0.1184445172548294,
+      "logits/rejected": -0.1339874565601349,
+      "logps/chosen": -10.12116527557373,
+      "logps/rejected": -53.403907775878906,
+      "loss": 0.5343,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 16.458633422851562,
+      "rewards/margins": 4.402472496032715,
+      "rewards/rejected": 12.056160926818848,
+      "step": 280
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 67.53883361816406,
+      "learning_rate": 4.6320058115409295e-06,
+      "logits/chosen": -0.1448262631893158,
+      "logits/rejected": -0.15793387591838837,
+      "logps/chosen": -3.4666190147399902,
+      "logps/rejected": -48.79213333129883,
+      "loss": 0.5017,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 16.945899963378906,
+      "rewards/margins": 4.2686333656311035,
+      "rewards/rejected": 12.677268028259277,
+      "step": 290
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.17521341145038605,
+      "learning_rate": 4.590455310636778e-06,
+      "logits/chosen": -0.16128253936767578,
+      "logits/rejected": -0.17375555634498596,
+      "logps/chosen": -2.9032950401306152,
+      "logps/rejected": -47.69734191894531,
+      "loss": 0.265,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.18383026123047,
+      "rewards/margins": 4.541309356689453,
+      "rewards/rejected": 12.642518997192383,
+      "step": 300
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "eval_logits/chosen": -0.17444846034049988,
+      "eval_logits/rejected": -0.18559777736663818,
+      "eval_logps/chosen": -2.535512924194336,
+      "eval_logps/rejected": -47.16367721557617,
+      "eval_loss": 0.15360687673091888,
+      "eval_rewards/accuracies": 0.9440000653266907,
+      "eval_rewards/chosen": 17.249343872070312,
+      "eval_rewards/margins": 4.604989051818848,
+      "eval_rewards/rejected": 12.644353866577148,
+      "eval_runtime": 361.4575,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 300
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 0.5040452480316162,
+      "learning_rate": 4.54689041428819e-06,
+      "logits/chosen": -0.16974106431007385,
+      "logits/rejected": -0.1810058057308197,
+      "logps/chosen": -1.233938217163086,
+      "logps/rejected": -49.907745361328125,
+      "loss": 0.1132,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.34117889404297,
+      "rewards/margins": 4.934173583984375,
+      "rewards/rejected": 12.407005310058594,
+      "step": 310
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 100.02949523925781,
+      "learning_rate": 4.501353102310901e-06,
+      "logits/chosen": -0.15705889463424683,
+      "logits/rejected": -0.1695334017276764,
+      "logps/chosen": -1.0820492506027222,
+      "logps/rejected": -52.577110290527344,
+      "loss": 0.1194,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.33388900756836,
+      "rewards/margins": 5.154760837554932,
+      "rewards/rejected": 12.179126739501953,
+      "step": 320
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 0.2689219117164612,
+      "learning_rate": 4.453887255171206e-06,
+      "logits/chosen": -0.13849371671676636,
+      "logits/rejected": -0.14990833401679993,
+      "logps/chosen": -1.8435032367706299,
+      "logps/rejected": -54.79044723510742,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.2423095703125,
+      "rewards/margins": 5.28987979888916,
+      "rewards/rejected": 11.952428817749023,
+      "step": 330
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 0.09305431693792343,
+      "learning_rate": 4.404538611702055e-06,
+      "logits/chosen": -0.12299702316522598,
+      "logits/rejected": -0.13453055918216705,
+      "logps/chosen": -2.9897143840789795,
+      "logps/rejected": -52.954498291015625,
+      "loss": 0.2873,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.17474365234375,
+      "rewards/margins": 5.071004867553711,
+      "rewards/rejected": 12.103739738464355,
+      "step": 340
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 59.282073974609375,
+      "learning_rate": 4.3533547250284015e-06,
+      "logits/chosen": -0.11913029849529266,
+      "logits/rejected": -0.12785324454307556,
+      "logps/chosen": -3.9456872940063477,
+      "logps/rejected": -48.68487548828125,
+      "loss": 0.4332,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 17.12805938720703,
+      "rewards/margins": 4.669450283050537,
+      "rewards/rejected": 12.458610534667969,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31101909279823303,
+      "learning_rate": 4.300384916744261e-06,
+      "logits/chosen": -0.11280188709497452,
+      "logits/rejected": -0.12300585210323334,
+      "logps/chosen": -2.1714723110198975,
+      "logps/rejected": -54.74174118041992,
+      "loss": 0.1605,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.326162338256836,
+      "rewards/margins": 5.467062473297119,
+      "rewards/rejected": 11.859098434448242,
+      "step": 360
+    },
+    {
+      "epoch": 0.32,
+      "eval_logits/chosen": -0.10620756447315216,
+      "eval_logits/rejected": -0.11727114766836166,
+      "eval_logps/chosen": -1.4165427684783936,
+      "eval_logps/rejected": -50.9525146484375,
+      "eval_loss": 0.3194349706172943,
+      "eval_rewards/accuracies": 0.9210000038146973,
+      "eval_rewards/chosen": 17.36124038696289,
+      "eval_rewards/margins": 5.095769882202148,
+      "eval_rewards/rejected": 12.26546859741211,
+      "eval_runtime": 361.5072,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 360
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 6.1126532554626465,
+      "learning_rate": 4.24568022938566e-06,
+      "logits/chosen": -0.10354311764240265,
+      "logits/rejected": -0.11526636779308319,
+      "logps/chosen": -1.2935255765914917,
+      "logps/rejected": -55.57566833496094,
+      "loss": 0.1711,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.439346313476562,
+      "rewards/margins": 5.700921058654785,
+      "rewards/rejected": 11.738424301147461,
+      "step": 370
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 34.15927505493164,
+      "learning_rate": 4.189293377245241e-06,
+      "logits/chosen": -0.1029932051897049,
+      "logits/rejected": -0.11382515728473663,
+      "logps/chosen": -2.5132687091827393,
+      "logps/rejected": -55.50346374511719,
+      "loss": 0.4359,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.731037139892578,
+      "rewards/margins": 4.368172645568848,
+      "rewards/rejected": 12.362865447998047,
+      "step": 380
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 2.8422904014587402,
+      "learning_rate": 4.131278695575952e-06,
+      "logits/chosen": -0.10793520510196686,
+      "logits/rejected": -0.12109285593032837,
+      "logps/chosen": -3.014652729034424,
+      "logps/rejected": -53.98411560058594,
+      "loss": 0.2161,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.137393951416016,
+      "rewards/margins": 5.105995178222656,
+      "rewards/rejected": 12.03139877319336,
+      "step": 390
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 54.0329475402832,
+      "learning_rate": 4.071692088232743e-06,
+      "logits/chosen": -0.10393750667572021,
+      "logits/rejected": -0.11834606528282166,
+      "logps/chosen": -2.1508543491363525,
+      "logps/rejected": -45.60733413696289,
+      "loss": 0.2077,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.586124420166016,
+      "rewards/margins": 5.077212333679199,
+      "rewards/rejected": 12.5089111328125,
+      "step": 400
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 81.61144256591797,
+      "learning_rate": 4.010590973802737e-06,
+      "logits/chosen": -0.09564584493637085,
+      "logits/rejected": -0.10617707669734955,
+      "logps/chosen": -3.4572842121124268,
+      "logps/rejected": -50.92162322998047,
+      "loss": 0.2478,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 17.010910034179688,
+      "rewards/margins": 4.556198596954346,
+      "rewards/rejected": 12.454713821411133,
+      "step": 410
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 0.30974289774894714,
+      "learning_rate": 3.948034230275781e-06,
+      "logits/chosen": -0.09134417027235031,
+      "logits/rejected": -0.1020016297698021,
+      "logps/chosen": -5.046698570251465,
+      "logps/rejected": -48.908958435058594,
+      "loss": 0.2894,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.007888793945312,
+      "rewards/margins": 4.53641414642334,
+      "rewards/rejected": 12.471475601196289,
+      "step": 420
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "eval_logits/chosen": -0.09054450690746307,
+      "eval_logits/rejected": -0.10264354199171066,
+      "eval_logps/chosen": -1.913105845451355,
+      "eval_logps/rejected": -51.11127471923828,
+      "eval_loss": 0.16789735853672028,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.311582565307617,
+      "eval_rewards/margins": 5.061989784240723,
+      "eval_rewards/rejected": 12.249593734741211,
+      "eval_runtime": 361.5337,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 420
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 12.824393272399902,
+      "learning_rate": 3.884082138308699e-06,
+      "logits/chosen": -0.08666776865720749,
+      "logits/rejected": -0.0997733399271965,
+      "logps/chosen": -1.7306327819824219,
+      "logps/rejected": -54.273292541503906,
+      "loss": 0.2298,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.167621612548828,
+      "rewards/margins": 5.065673351287842,
+      "rewards/rejected": 12.101947784423828,
+      "step": 430
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.30713599920272827,
+      "learning_rate": 3.818796323137896e-06,
+      "logits/chosen": -0.09174907952547073,
+      "logits/rejected": -0.10376611351966858,
+      "logps/chosen": -1.489154577255249,
+      "logps/rejected": -54.580726623535156,
+      "loss": 0.2513,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.22280502319336,
+      "rewards/margins": 5.175349235534668,
+      "rewards/rejected": 12.047454833984375,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 87.4791488647461,
+      "learning_rate": 3.7522396951963303e-06,
+      "logits/chosen": -0.09688778221607208,
+      "logits/rejected": -0.10897806286811829,
+      "logps/chosen": -3.157695770263672,
+      "logps/rejected": -50.96417236328125,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.345651626586914,
+      "rewards/margins": 5.245656967163086,
+      "rewards/rejected": 12.099993705749512,
+      "step": 450
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 146.2008056640625,
+      "learning_rate": 3.684476389492026e-06,
+      "logits/chosen": -0.09378582239151001,
+      "logits/rejected": -0.10475654900074005,
+      "logps/chosen": -0.5611928701400757,
+      "logps/rejected": -56.518890380859375,
+      "loss": 0.1981,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.113712310791016,
+      "rewards/margins": 5.068872928619385,
+      "rewards/rejected": 12.044839859008789,
+      "step": 460
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.9137721061706543,
+      "learning_rate": 3.6155717038065783e-06,
+      "logits/chosen": -0.08695463836193085,
+      "logits/rejected": -0.09596743434667587,
+      "logps/chosen": -1.5298550128936768,
+      "logps/rejected": -50.27445983886719,
+      "loss": 0.2066,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.35186004638672,
+      "rewards/margins": 5.014693260192871,
+      "rewards/rejected": 12.337167739868164,
+      "step": 470
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 84.80391693115234,
+      "learning_rate": 3.545592035773192e-06,
+      "logits/chosen": -0.0746893435716629,
+      "logits/rejected": -0.08653923869132996,
+      "logps/chosen": -2.0052125453948975,
+      "logps/rejected": -57.502811431884766,
+      "loss": 0.1149,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.14373016357422,
+      "rewards/margins": 5.360415935516357,
+      "rewards/rejected": 11.783313751220703,
+      "step": 480
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "eval_logits/chosen": -0.07700399309396744,
+      "eval_logits/rejected": -0.08828537166118622,
+      "eval_logps/chosen": -4.48896598815918,
+      "eval_logps/rejected": -53.76282501220703,
+      "eval_loss": 0.29511645436286926,
+      "eval_rewards/accuracies": 0.9230000376701355,
+      "eval_rewards/chosen": 17.053997039794922,
+      "eval_rewards/margins": 5.069558143615723,
+      "eval_rewards/rejected": 11.984437942504883,
+      "eval_runtime": 361.5035,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 480
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 82.9616470336914,
+      "learning_rate": 3.4746048188948806e-06,
+      "logits/chosen": -0.06675051152706146,
+      "logits/rejected": -0.07860895991325378,
+      "logps/chosen": -4.162237167358398,
+      "logps/rejected": -54.77789306640625,
+      "loss": 0.2979,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.047603607177734,
+      "rewards/margins": 5.138361930847168,
+      "rewards/rejected": 11.909242630004883,
+      "step": 490
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.04293210059404373,
+      "learning_rate": 3.4026784575644887e-06,
+      "logits/chosen": -0.06424491107463837,
+      "logits/rejected": -0.07567107677459717,
+      "logps/chosen": -2.05729603767395,
+      "logps/rejected": -56.646087646484375,
+      "loss": 0.4378,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.947803497314453,
+      "rewards/margins": 4.919981956481934,
+      "rewards/rejected": 12.02782154083252,
+      "step": 500
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 0.07136644423007965,
+      "learning_rate": 3.329882261149148e-06,
+      "logits/chosen": -0.06423303484916687,
+      "logits/rejected": -0.07512776553630829,
+      "logps/chosen": -3.1519265174865723,
+      "logps/rejected": -53.53008270263672,
+      "loss": 0.2613,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.112144470214844,
+      "rewards/margins": 5.046430587768555,
+      "rewards/rejected": 12.065712928771973,
+      "step": 510
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 0.29279613494873047,
+      "learning_rate": 3.25628637720269e-06,
+      "logits/chosen": -0.060233693569898605,
+      "logits/rejected": -0.07076811790466309,
+      "logps/chosen": -1.2358124256134033,
+      "logps/rejected": -50.61806106567383,
+      "loss": 0.1517,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.457351684570312,
+      "rewards/margins": 5.2062835693359375,
+      "rewards/rejected": 12.251070022583008,
+      "step": 520
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 0.7946074604988098,
+      "learning_rate": 3.181961723870359e-06,
+      "logits/chosen": -0.054482050240039825,
+      "logits/rejected": -0.0657092034816742,
+      "logps/chosen": -0.7682158946990967,
+      "logps/rejected": -58.81409454345703,
+      "loss": 0.2304,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.075973510742188,
+      "rewards/margins": 5.222441673278809,
+      "rewards/rejected": 11.853530883789062,
+      "step": 530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1405210494995117,
+      "learning_rate": 3.1069799215509847e-06,
+      "logits/chosen": -0.05050881579518318,
+      "logits/rejected": -0.061149902641773224,
+      "logps/chosen": -0.954046368598938,
+      "logps/rejected": -52.70227813720703,
+      "loss": 0.0384,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.378849029541016,
+      "rewards/margins": 5.253483295440674,
+      "rewards/rejected": 12.125364303588867,
+      "step": 540
+    },
+    {
+      "epoch": 0.48,
+      "eval_logits/chosen": -0.05120665580034256,
+      "eval_logits/rejected": -0.06123337894678116,
+      "eval_logps/chosen": -2.9872913360595703,
+      "eval_logps/rejected": -52.27314758300781,
+      "eval_loss": 0.17387841641902924,
+      "eval_rewards/accuracies": 0.9490000605583191,
+      "eval_rewards/chosen": 17.204164505004883,
+      "eval_rewards/margins": 5.070757865905762,
+      "eval_rewards/rejected": 12.133406639099121,
+      "eval_runtime": 361.5449,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 540
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.07270358502864838,
+      "learning_rate": 3.0314132238824416e-06,
+      "logits/chosen": -0.05125313252210617,
+      "logits/rejected": -0.06174170970916748,
+      "logps/chosen": -3.2163877487182617,
+      "logps/rejected": -47.79279327392578,
+      "loss": 0.2087,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.408517837524414,
+      "rewards/margins": 5.061524391174316,
+      "rewards/rejected": 12.346992492675781,
+      "step": 550
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.10005924850702286,
+      "learning_rate": 2.955334448116915e-06,
+      "logits/chosen": -0.041773442178964615,
+      "logits/rejected": -0.05364570394158363,
+      "logps/chosen": -0.363404780626297,
+      "logps/rejected": -56.32415008544922,
+      "loss": 0.0969,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.3824462890625,
+      "rewards/margins": 5.5404510498046875,
+      "rewards/rejected": 11.841995239257812,
+      "step": 560
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 7.818356990814209,
+      "learning_rate": 2.8788169049530533e-06,
+      "logits/chosen": -0.04309462010860443,
+      "logits/rejected": -0.05494442582130432,
+      "logps/chosen": -2.2242724895477295,
+      "logps/rejected": -56.444740295410156,
+      "loss": 0.1447,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.275390625,
+      "rewards/margins": 5.55007791519165,
+      "rewards/rejected": 11.725313186645508,
+      "step": 570
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 0.03519747406244278,
+      "learning_rate": 2.8019343278926397e-06,
+      "logits/chosen": -0.03506368771195412,
+      "logits/rejected": -0.046854715794324875,
+      "logps/chosen": -0.5200096964836121,
+      "logps/rejected": -59.05330276489258,
+      "loss": 0.0998,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.320554733276367,
+      "rewards/margins": 5.727260589599609,
+      "rewards/rejected": 11.593294143676758,
+      "step": 580
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 0.04519123584032059,
+      "learning_rate": 2.7247608021898265e-06,
+      "logits/chosen": -0.03204537555575371,
+      "logits/rejected": -0.04383891448378563,
+      "logps/chosen": -1.1271060705184937,
+      "logps/rejected": -59.308895111083984,
+      "loss": 0.1162,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.256351470947266,
+      "rewards/margins": 5.6881890296936035,
+      "rewards/rejected": 11.568161010742188,
+      "step": 590
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.028689857572317123,
+      "learning_rate": 2.647370693461432e-06,
+      "logits/chosen": -0.02834726870059967,
+      "logits/rejected": -0.03827046602964401,
+      "logps/chosen": -5.673943519592285,
+      "logps/rejected": -55.72624588012695,
+      "loss": 0.4008,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.761056900024414,
+      "rewards/margins": 4.804043769836426,
+      "rewards/rejected": 11.957012176513672,
+      "step": 600
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "eval_logits/chosen": -0.035840023308992386,
+      "eval_logits/rejected": -0.04687971621751785,
+      "eval_logps/chosen": -2.1760435104370117,
+      "eval_logps/rejected": -56.62664031982422,
+      "eval_loss": 0.17061151564121246,
+      "eval_rewards/accuracies": 0.9470000267028809,
+      "eval_rewards/chosen": 17.285289764404297,
+      "eval_rewards/margins": 5.587231636047363,
+      "eval_rewards/rejected": 11.698057174682617,
+      "eval_runtime": 361.5056,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 600
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 1.1994622945785522,
+      "learning_rate": 2.569838576027068e-06,
+      "logits/chosen": -0.03231767192482948,
+      "logits/rejected": -0.04397805407643318,
+      "logps/chosen": -1.5904741287231445,
+      "logps/rejected": -62.10520553588867,
+      "loss": 0.1265,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.11206817626953,
+      "rewards/margins": 5.734784126281738,
+      "rewards/rejected": 11.37728500366211,
+      "step": 610
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 0.14163845777511597,
+      "learning_rate": 2.4922391610481544e-06,
+      "logits/chosen": -0.03293662518262863,
+      "logits/rejected": -0.04335154965519905,
+      "logps/chosen": -1.3568997383117676,
+      "logps/rejected": -58.347434997558594,
+      "loss": 0.097,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.345399856567383,
+      "rewards/margins": 5.823373794555664,
+      "rewards/rejected": 11.522026062011719,
+      "step": 620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.125027060508728,
+      "learning_rate": 2.4146472245350804e-06,
+      "logits/chosen": -0.02864963933825493,
+      "logits/rejected": -0.03835710883140564,
+      "logps/chosen": -2.494175434112549,
+      "logps/rejected": -55.33067321777344,
+      "loss": 0.3115,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.262205123901367,
+      "rewards/margins": 5.4287261962890625,
+      "rewards/rejected": 11.833479881286621,
+      "step": 630
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.03514016419649124,
+      "learning_rate": 2.337137535291868e-06,
+      "logits/chosen": -0.02757749892771244,
+      "logits/rejected": -0.0376611053943634,
+      "logps/chosen": -2.3163387775421143,
+      "logps/rejected": -53.54579162597656,
+      "loss": 0.1268,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.542863845825195,
+      "rewards/margins": 5.799897193908691,
+      "rewards/rejected": 11.742965698242188,
+      "step": 640
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 34.20791244506836,
+      "learning_rate": 2.259784782867782e-06,
+      "logits/chosen": -0.02032250165939331,
+      "logits/rejected": -0.031542714685201645,
+      "logps/chosen": -1.7910137176513672,
+      "logps/rejected": -62.980018615722656,
+      "loss": 0.1831,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.982879638671875,
+      "rewards/margins": 5.589078903198242,
+      "rewards/rejected": 11.393800735473633,
+      "step": 650
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 32.356956481933594,
+      "learning_rate": 2.182663505585314e-06,
+      "logits/chosen": -0.01711965538561344,
+      "logits/rejected": -0.028038471937179565,
+      "logps/chosen": -2.662904739379883,
+      "logps/rejected": -63.326297760009766,
+      "loss": 0.1678,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 16.997777938842773,
+      "rewards/margins": 5.741157531738281,
+      "rewards/rejected": 11.256620407104492,
+      "step": 660
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "eval_logits/chosen": -0.016018809750676155,
+      "eval_logits/rejected": -0.027020033448934555,
+      "eval_logps/chosen": -3.00822114944458,
+      "eval_logps/rejected": -57.951629638671875,
+      "eval_loss": 0.2050127536058426,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.202072143554688,
+      "eval_rewards/margins": 5.636512756347656,
+      "eval_rewards/rejected": 11.565557479858398,
+      "eval_runtime": 361.5073,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 660
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 86.53874969482422,
+      "learning_rate": 2.1058480187138863e-06,
+      "logits/chosen": -0.012352555990219116,
+      "logits/rejected": -0.023966707289218903,
+      "logps/chosen": -2.129209041595459,
+      "logps/rejected": -62.1393928527832,
+      "loss": 0.1025,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.069538116455078,
+      "rewards/margins": 5.678930759429932,
+      "rewards/rejected": 11.390605926513672,
+      "step": 670
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 2.892672300338745,
+      "learning_rate": 2.0294123428584985e-06,
+      "logits/chosen": -0.011562807485461235,
+      "logits/rejected": -0.020859256386756897,
+      "logps/chosen": -3.2486608028411865,
+      "logps/rejected": -59.56721496582031,
+      "loss": 0.1961,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.051376342773438,
+      "rewards/margins": 5.515361785888672,
+      "rewards/rejected": 11.536016464233398,
+      "step": 680
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 30.26588249206543,
+      "learning_rate": 1.953430132632311e-06,
+      "logits/chosen": -0.011488726362586021,
+      "logits/rejected": -0.021527227014303207,
+      "logps/chosen": -1.794357180595398,
+      "logps/rejected": -60.649505615234375,
+      "loss": 0.2821,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.114843368530273,
+      "rewards/margins": 5.612217903137207,
+      "rewards/rejected": 11.50262451171875,
+      "step": 690
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 0.030314341187477112,
+      "learning_rate": 1.8779746056819104e-06,
+      "logits/chosen": -0.014436552301049232,
+      "logits/rejected": -0.026044374331831932,
+      "logps/chosen": -3.1617178916931152,
+      "logps/rejected": -60.94548797607422,
+      "loss": 0.3141,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.08698272705078,
+      "rewards/margins": 5.719264984130859,
+      "rewards/rejected": 11.367716789245605,
+      "step": 700
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 0.06711317598819733,
+      "learning_rate": 1.8031184721336364e-06,
+      "logits/chosen": -0.017312290146946907,
+      "logits/rejected": -0.026554957032203674,
+      "logps/chosen": -5.120705604553223,
+      "logps/rejected": -54.33483123779297,
+      "loss": 0.3827,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 17.1247501373291,
+      "rewards/margins": 5.327882766723633,
+      "rewards/rejected": 11.796867370605469,
+      "step": 710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 60.36637496948242,
+      "learning_rate": 1.7289338645289711e-06,
+      "logits/chosen": -0.01987219974398613,
+      "logits/rejected": -0.02766304835677147,
+      "logps/chosen": -2.1757986545562744,
+      "logps/rejected": -54.02531433105469,
+      "loss": 0.2272,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.239776611328125,
+      "rewards/margins": 5.250313758850098,
+      "rewards/rejected": 11.989462852478027,
+      "step": 720
+    },
+    {
+      "epoch": 0.64,
+      "eval_logits/chosen": -0.022905193269252777,
+      "eval_logits/rejected": -0.0321992002427578,
+      "eval_logps/chosen": -1.100506067276001,
+      "eval_logps/rejected": -55.91169738769531,
+      "eval_loss": 0.14018221199512482,
+      "eval_rewards/accuracies": 0.9520000219345093,
+      "eval_rewards/chosen": 17.392841339111328,
+      "eval_rewards/margins": 5.623291492462158,
+      "eval_rewards/rejected": 11.769551277160645,
+      "eval_runtime": 361.416,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 720
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 0.029294608160853386,
+      "learning_rate": 1.6554922683164875e-06,
+      "logits/chosen": -0.016416028141975403,
+      "logits/rejected": -0.025527067482471466,
+      "logps/chosen": -0.906692385673523,
+      "logps/rejected": -59.142173767089844,
+      "loss": 0.1171,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.282169342041016,
+      "rewards/margins": 5.6978583335876465,
+      "rewards/rejected": 11.584310531616211,
+      "step": 730
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 33.89070510864258,
+      "learning_rate": 1.5828644529673592e-06,
+      "logits/chosen": -0.016044551506638527,
+      "logits/rejected": -0.025811903178691864,
+      "logps/chosen": -1.1167538166046143,
+      "logps/rejected": -59.84492492675781,
+      "loss": 0.2001,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.266983032226562,
+      "rewards/margins": 5.749438285827637,
+      "rewards/rejected": 11.51754379272461,
+      "step": 740
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.53726726770401,
+      "learning_rate": 1.5111204037807844e-06,
+      "logits/chosen": -0.012529855594038963,
+      "logits/rejected": -0.023765765130519867,
+      "logps/chosen": -0.3461765646934509,
+      "logps/rejected": -55.841102600097656,
+      "loss": 0.1692,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.61865997314453,
+      "rewards/margins": 6.019055366516113,
+      "rewards/rejected": 11.599604606628418,
+      "step": 750
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.7173987627029419,
+      "learning_rate": 1.4403292544450625e-06,
+      "logits/chosen": -0.013380522839725018,
+      "logits/rejected": -0.022107835859060287,
+      "logps/chosen": -3.2759666442871094,
+      "logps/rejected": -53.836822509765625,
+      "loss": 0.2418,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.383333206176758,
+      "rewards/margins": 5.590358734130859,
+      "rewards/rejected": 11.792974472045898,
+      "step": 760
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 2.744900941848755,
+      "learning_rate": 1.3705592204192853e-06,
+      "logits/chosen": -0.011318420059978962,
+      "logits/rejected": -0.021840626373887062,
+      "logps/chosen": -2.592001438140869,
+      "logps/rejected": -57.06926345825195,
+      "loss": 0.1749,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.286128997802734,
+      "rewards/margins": 5.662715911865234,
+      "rewards/rejected": 11.623414993286133,
+      "step": 770
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 70.76551818847656,
+      "learning_rate": 1.301877533199859e-06,
+      "logits/chosen": -0.013944407925009727,
+      "logits/rejected": -0.025667501613497734,
+      "logps/chosen": -2.1070234775543213,
+      "logps/rejected": -57.72039794921875,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.28545379638672,
+      "rewards/margins": 5.708344459533691,
+      "rewards/rejected": 11.577108383178711,
+      "step": 780
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "eval_logits/chosen": -0.01664295792579651,
+      "eval_logits/rejected": -0.026626665145158768,
+      "eval_logps/chosen": -1.082255244255066,
+      "eval_logps/rejected": -55.95073699951172,
+      "eval_loss": 0.24412688612937927,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.394668579101562,
+      "eval_rewards/margins": 5.629020690917969,
+      "eval_rewards/rejected": 11.765647888183594,
+      "eval_runtime": 361.4602,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 780
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 0.41359376907348633,
+      "learning_rate": 1.2343503755351729e-06,
+      "logits/chosen": -0.012772129848599434,
+      "logits/rejected": -0.023186586797237396,
+      "logps/chosen": -0.8030359148979187,
+      "logps/rejected": -57.84947967529297,
+      "loss": 0.2143,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.32242202758789,
+      "rewards/margins": 5.645486831665039,
+      "rewards/rejected": 11.676933288574219,
+      "step": 790
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.41748157143592834,
+      "learning_rate": 1.168042817650881e-06,
+      "logits/chosen": -0.012502101249992847,
+      "logits/rejected": -0.023272844031453133,
+      "logps/chosen": -1.0377256870269775,
+      "logps/rejected": -57.05615997314453,
+      "loss": 0.0792,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.60501480102539,
+      "rewards/margins": 6.159370422363281,
+      "rewards/rejected": 11.445646286010742,
+      "step": 800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.039210401475429535,
+      "learning_rate": 1.1030187545472012e-06,
+      "logits/chosen": -0.008177272044122219,
+      "logits/rejected": -0.01832464337348938,
+      "logps/chosen": -3.8701748847961426,
+      "logps/rejected": -57.661476135253906,
+      "loss": 0.3177,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.14336395263672,
+      "rewards/margins": 5.583393573760986,
+      "rewards/rejected": 11.55997085571289,
+      "step": 810
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.6096951961517334,
+      "learning_rate": 1.0393408444287048e-06,
+      "logits/chosen": -0.006832236424088478,
+      "logits/rejected": -0.01682097464799881,
+      "logps/chosen": -2.1741790771484375,
+      "logps/rejected": -57.273414611816406,
+      "loss": 0.2128,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.426712036132812,
+      "rewards/margins": 5.911205291748047,
+      "rewards/rejected": 11.515506744384766,
+      "step": 820
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 3.7820959091186523,
+      "learning_rate": 9.770704483258782e-07,
+      "logits/chosen": -0.009998206980526447,
+      "logits/rejected": -0.0204261876642704,
+      "logps/chosen": -1.9802953004837036,
+      "logps/rejected": -56.901512145996094,
+      "loss": 0.1296,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.56071662902832,
+      "rewards/margins": 6.169583320617676,
+      "rewards/rejected": 11.391134262084961,
+      "step": 830
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.126626968383789,
+      "learning_rate": 9.162675709666865e-07,
+      "logits/chosen": -0.00826224498450756,
+      "logits/rejected": -0.018977787345647812,
+      "logps/chosen": -1.4256607294082642,
+      "logps/rejected": -61.65986633300781,
+      "loss": 0.0635,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.324975967407227,
+      "rewards/margins": 6.09440803527832,
+      "rewards/rejected": 11.230567932128906,
+      "step": 840
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "eval_logits/chosen": -0.01111944392323494,
+      "eval_logits/rejected": -0.021697774529457092,
+      "eval_logps/chosen": -1.2168633937835693,
+      "eval_logps/rejected": -58.2642822265625,
+      "eval_loss": 0.1689341962337494,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.381206512451172,
+      "eval_rewards/margins": 5.846914291381836,
+      "eval_rewards/rejected": 11.534292221069336,
+      "eval_runtime": 361.6192,
+      "eval_samples_per_second": 2.765,
+      "eval_steps_per_second": 0.346,
+      "step": 840
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 5.250723838806152,
+      "learning_rate": 8.569908029550686e-07,
+      "logits/chosen": -0.006854387000203133,
+      "logits/rejected": -0.018336206674575806,
+      "logps/chosen": -0.6238930821418762,
+      "logps/rejected": -60.925689697265625,
+      "loss": 0.1157,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.443281173706055,
+      "rewards/margins": 6.164813995361328,
+      "rewards/rejected": 11.278467178344727,
+      "step": 850
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 3.1401162147521973,
+      "learning_rate": 7.992972643121227e-07,
+      "logits/chosen": -0.0037835021503269672,
+      "logits/rejected": -0.013135241344571114,
+      "logps/chosen": -0.8492221832275391,
+      "logps/rejected": -55.516075134277344,
+      "loss": 0.2252,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.488588333129883,
+      "rewards/margins": 5.736725807189941,
+      "rewards/rejected": 11.751862525939941,
+      "step": 860
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 55.528812408447266,
+      "learning_rate": 7.432425494343509e-07,
+      "logits/chosen": -0.0033687639515846968,
+      "logits/rejected": -0.013152632862329483,
+      "logps/chosen": -1.3188884258270264,
+      "logps/rejected": -57.9510498046875,
+      "loss": 0.1398,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.412578582763672,
+      "rewards/margins": 5.868515968322754,
+      "rewards/rejected": 11.544061660766602,
+      "step": 870
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.039824869483709335,
+      "learning_rate": 6.888806735220396e-07,
+      "logits/chosen": -0.0010406378423795104,
+      "logits/rejected": -0.012095071375370026,
+      "logps/chosen": -2.0619027614593506,
+      "logps/rejected": -59.65806579589844,
+      "loss": 0.2966,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.192535400390625,
+      "rewards/margins": 5.7006731033325195,
+      "rewards/rejected": 11.491861343383789,
+      "step": 880
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 0.7536466717720032,
+      "learning_rate": 6.362640205293583e-07,
+      "logits/chosen": -0.0016857212176546454,
+      "logits/rejected": -0.010936147533357143,
+      "logps/chosen": -2.1478958129882812,
+      "logps/rejected": -58.20386505126953,
+      "loss": 0.1641,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.243247985839844,
+      "rewards/margins": 5.678771018981934,
+      "rewards/rejected": 11.564477920532227,
+      "step": 890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3857377767562866,
+      "learning_rate": 5.854432926863684e-07,
+      "logits/chosen": 0.00038508616853505373,
+      "logits/rejected": -0.011034643277525902,
+      "logps/chosen": -1.606274127960205,
+      "logps/rejected": -62.58662414550781,
+      "loss": 0.1703,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.211261749267578,
+      "rewards/margins": 5.966868877410889,
+      "rewards/rejected": 11.244392395019531,
+      "step": 900
+    },
+    {
+      "epoch": 0.8,
+      "eval_logits/chosen": 0.00021816430671606213,
+      "eval_logits/rejected": -0.010477552190423012,
+      "eval_logps/chosen": -1.7576563358306885,
+      "eval_logps/rejected": -59.79063415527344,
+      "eval_loss": 0.13995186984539032,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.327129364013672,
+      "eval_rewards/margins": 5.945469856262207,
+      "eval_rewards/rejected": 11.381658554077148,
+      "eval_runtime": 361.4592,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 900
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 0.13143697381019592,
+      "learning_rate": 5.364674616415547e-07,
+      "logits/chosen": 0.0005570838693529367,
+      "logits/rejected": -0.011198626831173897,
+      "logps/chosen": -0.12791283428668976,
+      "logps/rejected": -65.7052993774414,
+      "loss": 0.0507,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.224443435668945,
+      "rewards/margins": 6.163690567016602,
+      "rewards/rejected": 11.060752868652344,
+      "step": 910
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 102.9654541015625,
+      "learning_rate": 4.893837212719859e-07,
+      "logits/chosen": -0.0008557128603570163,
+      "logits/rejected": -0.01163212489336729,
+      "logps/chosen": -1.3292646408081055,
+      "logps/rejected": -61.644893646240234,
+      "loss": 0.0881,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.276538848876953,
+      "rewards/margins": 5.96080207824707,
+      "rewards/rejected": 11.3157377243042,
+      "step": 920
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 146.04498291015625,
+      "learning_rate": 4.442374422065493e-07,
+      "logits/chosen": 0.002922601066529751,
+      "logits/rejected": -0.007130052894353867,
+      "logps/chosen": -1.6069023609161377,
+      "logps/rejected": -61.52588653564453,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.184974670410156,
+      "rewards/margins": 5.7995734214782715,
+      "rewards/rejected": 11.38540267944336,
+      "step": 930
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 23.539485931396484,
+      "learning_rate": 4.0107212810610974e-07,
+      "logits/chosen": 0.0018056132830679417,
+      "logits/rejected": -0.007847340777516365,
+      "logps/chosen": -3.637047290802002,
+      "logps/rejected": -61.21245574951172,
+      "loss": 0.2763,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.960758209228516,
+      "rewards/margins": 5.551811695098877,
+      "rewards/rejected": 11.40894603729248,
+      "step": 940
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 31.52926254272461,
+      "learning_rate": 3.599293737426932e-07,
+      "logits/chosen": 0.0032081177923828363,
+      "logits/rejected": -0.007756482809782028,
+      "logps/chosen": -1.172515869140625,
+      "logps/rejected": -66.5853271484375,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 16.992876052856445,
+      "rewards/margins": 5.886469841003418,
+      "rewards/rejected": 11.106407165527344,
+      "step": 950
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.07474468648433685,
+      "learning_rate": 3.208488249181216e-07,
+      "logits/chosen": 0.0022508346009999514,
+      "logits/rejected": -0.009156409651041031,
+      "logps/chosen": -0.7930470108985901,
+      "logps/rejected": -60.068790435791016,
+      "loss": 0.1138,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.401771545410156,
+      "rewards/margins": 6.0026960372924805,
+      "rewards/rejected": 11.399076461791992,
+      "step": 960
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "eval_logits/chosen": 0.001469604205340147,
+      "eval_logits/rejected": -0.009397665038704872,
+      "eval_logps/chosen": -1.8795456886291504,
+      "eval_logps/rejected": -60.17564010620117,
+      "eval_loss": 0.1441129744052887,
+      "eval_rewards/accuracies": 0.9630000591278076,
+      "eval_rewards/chosen": 17.314937591552734,
+      "eval_rewards/margins": 5.9717817306518555,
+      "eval_rewards/rejected": 11.343156814575195,
+      "eval_runtime": 361.5344,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 960
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 68.90747833251953,
+      "learning_rate": 2.838681402606952e-07,
+      "logits/chosen": 0.004552370868623257,
+      "logits/rejected": -0.005488495342433453,
+      "logps/chosen": -3.7298974990844727,
+      "logps/rejected": -64.72488403320312,
+      "loss": 0.2425,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.832754135131836,
+      "rewards/margins": 5.644216537475586,
+      "rewards/rejected": 11.18853759765625,
+      "step": 970
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 0.028206102550029755,
+      "learning_rate": 2.490229549367443e-07,
+      "logits/chosen": 0.0025807656347751617,
+      "logits/rejected": -0.008657123893499374,
+      "logps/chosen": -1.1825838088989258,
+      "logps/rejected": -63.57493591308594,
+      "loss": 0.0609,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.227130889892578,
+      "rewards/margins": 6.053717136383057,
+      "rewards/rejected": 11.173412322998047,
+      "step": 980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39742231369018555,
+      "learning_rate": 2.1634684631203412e-07,
+      "logits/chosen": 0.0048486413434147835,
+      "logits/rejected": -0.006056814920157194,
+      "logps/chosen": -2.919680595397949,
+      "logps/rejected": -57.8403205871582,
+      "loss": 0.3464,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.175275802612305,
+      "rewards/margins": 5.559727668762207,
+      "rewards/rejected": 11.615548133850098,
+      "step": 990
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 30.22509002685547,
+      "learning_rate": 1.8587130159608196e-07,
+      "logits/chosen": 0.0050649940967559814,
+      "logits/rejected": -0.0071399761363863945,
+      "logps/chosen": -0.16131475567817688,
+      "logps/rejected": -66.25190734863281,
+      "loss": 0.0063,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 17.313472747802734,
+      "rewards/margins": 6.412895679473877,
+      "rewards/rejected": 10.9005765914917,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1125,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2745297197268992e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5286bf380083ef64133f25e161ddfb8549ff44790e432f478753f221ef89e695
+size 5368

checkpoint-1125/README.md

@@ -0,0 +1,202 @@
+---
+base_model: mistralai/Mistral-Nemo-Instruct-2407
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.12.0

checkpoint-1125/adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "down_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1125/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6404a5e1c60aa26a99a4ccbb402f485483d5eea90c08441f0448e29c51e2b1d0
+size 114106856

checkpoint-1125/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:679b631e5cc31300cab94ec2968b5eafa50b74a27dc17dffbac7d227a1ae878e
+size 228536930

checkpoint-1125/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55f6ef5fd81f87dc8a69f1b9a751ba37cc49c37318322e45ba4733ff23a92208
+size 14512

checkpoint-1125/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f510db96e40d7f66609c96cf485c13417fc4eaf253603d2b6591466c3fb5f63a
+size 14512

checkpoint-1125/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb6eb7801fbbb5055fabd75421a32f0209bef58fe235404c9e167738db177888
+size 1064

checkpoint-1125/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1125/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

checkpoint-1125/trainer_state.json

@@ -0,0 +1,2001 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 60,
+  "global_step": 1125,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 8.760091781616211,
+      "learning_rate": 4.4247787610619474e-07,
+      "logits/chosen": -0.8248252868652344,
+      "logits/rejected": -0.8263720273971558,
+      "logps/chosen": -0.36086463928222656,
+      "logps/rejected": -5.696224689483643,
+      "loss": 1.1038,
+      "rewards/accuracies": 0.5125000476837158,
+      "rewards/chosen": 17.43745994567871,
+      "rewards/margins": 0.5984855890274048,
+      "rewards/rejected": 16.838973999023438,
+      "step": 10
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 8.855981826782227,
+      "learning_rate": 8.849557522123895e-07,
+      "logits/chosen": -0.8169006109237671,
+      "logits/rejected": -0.819770872592926,
+      "logps/chosen": -0.12464660406112671,
+      "logps/rejected": -7.139842987060547,
+      "loss": 1.1887,
+      "rewards/accuracies": 0.4000000059604645,
+      "rewards/chosen": 17.17649269104004,
+      "rewards/margins": 0.19107049703598022,
+      "rewards/rejected": 16.98542022705078,
+      "step": 20
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 16.764184951782227,
+      "learning_rate": 1.3274336283185843e-06,
+      "logits/chosen": -0.8003113865852356,
+      "logits/rejected": -0.8030117750167847,
+      "logps/chosen": -0.34651467204093933,
+      "logps/rejected": -6.967917442321777,
+      "loss": 1.0563,
+      "rewards/accuracies": 0.44999998807907104,
+      "rewards/chosen": 17.280975341796875,
+      "rewards/margins": 0.40005987882614136,
+      "rewards/rejected": 16.88091468811035,
+      "step": 30
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 8.33682918548584,
+      "learning_rate": 1.769911504424779e-06,
+      "logits/chosen": -0.7695047855377197,
+      "logits/rejected": -0.7739207148551941,
+      "logps/chosen": -1.5993006229400635,
+      "logps/rejected": -8.504932403564453,
+      "loss": 0.7596,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 17.283912658691406,
+      "rewards/margins": 0.6976072192192078,
+      "rewards/rejected": 16.5863037109375,
+      "step": 40
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 4.494723320007324,
+      "learning_rate": 2.212389380530974e-06,
+      "logits/chosen": -0.7154140472412109,
+      "logits/rejected": -0.7225576043128967,
+      "logps/chosen": -3.112199068069458,
+      "logps/rejected": -12.212080001831055,
+      "loss": 0.6083,
+      "rewards/accuracies": 0.4625000059604645,
+      "rewards/chosen": 17.03064727783203,
+      "rewards/margins": 0.7148451805114746,
+      "rewards/rejected": 16.3158016204834,
+      "step": 50
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 5.110287666320801,
+      "learning_rate": 2.6548672566371687e-06,
+      "logits/chosen": -0.6322453022003174,
+      "logits/rejected": -0.6387485265731812,
+      "logps/chosen": -5.650620460510254,
+      "logps/rejected": -12.759811401367188,
+      "loss": 0.3835,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.101289749145508,
+      "rewards/margins": 1.1824612617492676,
+      "rewards/rejected": 15.918828964233398,
+      "step": 60
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "eval_logits/chosen": -0.5826543569564819,
+      "eval_logits/rejected": -0.5914276838302612,
+      "eval_logps/chosen": -3.5471787452697754,
+      "eval_logps/rejected": -16.51181983947754,
+      "eval_loss": 0.3286525011062622,
+      "eval_rewards/accuracies": 0.9280000925064087,
+      "eval_rewards/chosen": 17.148174285888672,
+      "eval_rewards/margins": 1.4386365413665771,
+      "eval_rewards/rejected": 15.709539413452148,
+      "eval_runtime": 372.0227,
+      "eval_samples_per_second": 2.688,
+      "eval_steps_per_second": 0.336,
+      "step": 60
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 5.098133563995361,
+      "learning_rate": 3.097345132743363e-06,
+      "logits/chosen": -0.5378152132034302,
+      "logits/rejected": -0.5494933724403381,
+      "logps/chosen": -1.5099802017211914,
+      "logps/rejected": -21.206321716308594,
+      "loss": 0.2931,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.083791732788086,
+      "rewards/margins": 1.5844331979751587,
+      "rewards/rejected": 15.499359130859375,
+      "step": 70
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 29.787437438964844,
+      "learning_rate": 3.539823008849558e-06,
+      "logits/chosen": -0.443774938583374,
+      "logits/rejected": -0.45571577548980713,
+      "logps/chosen": -1.5804342031478882,
+      "logps/rejected": -22.606929779052734,
+      "loss": 0.202,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.302125930786133,
+      "rewards/margins": 2.174014091491699,
+      "rewards/rejected": 15.128110885620117,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.14398193359375,
+      "learning_rate": 3.982300884955752e-06,
+      "logits/chosen": -0.3626072406768799,
+      "logits/rejected": -0.3787815570831299,
+      "logps/chosen": -2.203828811645508,
+      "logps/rejected": -29.433551788330078,
+      "loss": 0.2123,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.00284194946289,
+      "rewards/margins": 2.320391893386841,
+      "rewards/rejected": 14.682450294494629,
+      "step": 90
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 29.672739028930664,
+      "learning_rate": 4.424778761061948e-06,
+      "logits/chosen": -0.3035663962364197,
+      "logits/rejected": -0.31762221455574036,
+      "logps/chosen": -3.433589458465576,
+      "logps/rejected": -29.9322509765625,
+      "loss": 0.2592,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.929956436157227,
+      "rewards/margins": 2.31272029876709,
+      "rewards/rejected": 14.617237091064453,
+      "step": 100
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.873722791671753,
+      "learning_rate": 4.867256637168142e-06,
+      "logits/chosen": -0.2679600715637207,
+      "logits/rejected": -0.2826440930366516,
+      "logps/chosen": -0.9653514623641968,
+      "logps/rejected": -30.235322952270508,
+      "loss": 0.1336,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.462385177612305,
+      "rewards/margins": 3.1994175910949707,
+      "rewards/rejected": 14.26296615600586,
+      "step": 110
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.6913721561431885,
+      "learning_rate": 4.999409761242696e-06,
+      "logits/chosen": -0.22222033143043518,
+      "logits/rejected": -0.23720571398735046,
+      "logps/chosen": -4.4953508377075195,
+      "logps/rejected": -34.074745178222656,
+      "loss": 0.2552,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.04866600036621,
+      "rewards/margins": 3.1014418601989746,
+      "rewards/rejected": 13.947224617004395,
+      "step": 120
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "eval_logits/chosen": -0.206527978181839,
+      "eval_logits/rejected": -0.22178640961647034,
+      "eval_logps/chosen": -3.69442081451416,
+      "eval_logps/rejected": -36.072166442871094,
+      "eval_loss": 0.18996010720729828,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.133451461791992,
+      "eval_rewards/margins": 3.379946708679199,
+      "eval_rewards/rejected": 13.753504753112793,
+      "eval_runtime": 361.5279,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 120
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 61.80262756347656,
+      "learning_rate": 4.996519466816778e-06,
+      "logits/chosen": -0.18473535776138306,
+      "logits/rejected": -0.1988501250743866,
+      "logps/chosen": -3.7009687423706055,
+      "logps/rejected": -39.289939880371094,
+      "loss": 0.1394,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.106964111328125,
+      "rewards/margins": 3.633338212966919,
+      "rewards/rejected": 13.473625183105469,
+      "step": 130
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.6732702255249023,
+      "learning_rate": 4.9912234871722805e-06,
+      "logits/chosen": -0.16134041547775269,
+      "logits/rejected": -0.17547868192195892,
+      "logps/chosen": -3.0637736320495605,
+      "logps/rejected": -40.07548522949219,
+      "loss": 0.1408,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.392223358154297,
+      "rewards/margins": 4.242353439331055,
+      "rewards/rejected": 13.149867057800293,
+      "step": 140
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.346453994512558,
+      "learning_rate": 4.98352692559805e-06,
+      "logits/chosen": -0.13797929883003235,
+      "logits/rejected": -0.15283086895942688,
+      "logps/chosen": -5.14492130279541,
+      "logps/rejected": -47.97212219238281,
+      "loss": 0.2153,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.896778106689453,
+      "rewards/margins": 4.227695465087891,
+      "rewards/rejected": 12.669081687927246,
+      "step": 150
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.21871662139892578,
+      "learning_rate": 4.973437198621237e-06,
+      "logits/chosen": -0.12396670132875443,
+      "logits/rejected": -0.13780555129051208,
+      "logps/chosen": -6.108860015869141,
+      "logps/rejected": -54.90739440917969,
+      "loss": 0.0388,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 16.75935935974121,
+      "rewards/margins": 4.755282878875732,
+      "rewards/rejected": 12.004077911376953,
+      "step": 160
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 235.12429809570312,
+      "learning_rate": 4.960964028860621e-06,
+      "logits/chosen": -0.1140839159488678,
+      "logits/rejected": -0.1263057291507721,
+      "logps/chosen": -12.605452537536621,
+      "logps/rejected": -53.81230926513672,
+      "loss": 0.4651,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.101238250732422,
+      "rewards/margins": 3.9864249229431152,
+      "rewards/rejected": 12.114812850952148,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 190.97048950195312,
+      "learning_rate": 4.946119435657738e-06,
+      "logits/chosen": -0.10746976733207703,
+      "logits/rejected": -0.11878640949726105,
+      "logps/chosen": -8.5105562210083,
+      "logps/rejected": -51.314781188964844,
+      "loss": 0.2362,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.719980239868164,
+      "rewards/margins": 4.549674034118652,
+      "rewards/rejected": 12.170306205749512,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "eval_logits/chosen": -0.10870806127786636,
+      "eval_logits/rejected": -0.12223993986845016,
+      "eval_logps/chosen": -4.414996147155762,
+      "eval_logps/rejected": -53.885032653808594,
+      "eval_loss": 0.20236633718013763,
+      "eval_rewards/accuracies": 0.9510000944137573,
+      "eval_rewards/chosen": 17.06139373779297,
+      "eval_rewards/margins": 5.089176177978516,
+      "eval_rewards/rejected": 11.97221851348877,
+      "eval_runtime": 361.4355,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 180
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 56.81266784667969,
+      "learning_rate": 4.928917723494854e-06,
+      "logits/chosen": -0.10682469606399536,
+      "logits/rejected": -0.12124393880367279,
+      "logps/chosen": -3.058413028717041,
+      "logps/rejected": -55.052528381347656,
+      "loss": 0.2442,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.058589935302734,
+      "rewards/margins": 5.056097984313965,
+      "rewards/rejected": 12.002490043640137,
+      "step": 190
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 175.06552124023438,
+      "learning_rate": 4.909375468210947e-06,
+      "logits/chosen": -0.10520349442958832,
+      "logits/rejected": -0.12018950283527374,
+      "logps/chosen": -4.114959716796875,
+      "logps/rejected": -55.9394645690918,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.98603057861328,
+      "rewards/margins": 5.105838775634766,
+      "rewards/rejected": 11.880191802978516,
+      "step": 200
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 78.06558990478516,
+      "learning_rate": 4.8875115010289655e-06,
+      "logits/chosen": -0.10475558042526245,
+      "logits/rejected": -0.11949175596237183,
+      "logps/chosen": -6.760301113128662,
+      "logps/rejected": -53.91607666015625,
+      "loss": 0.2843,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.857545852661133,
+      "rewards/margins": 4.917357921600342,
+      "rewards/rejected": 11.94018840789795,
+      "step": 210
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 15.880486488342285,
+      "learning_rate": 4.863346890409768e-06,
+      "logits/chosen": -0.11213523149490356,
+      "logits/rejected": -0.12581588327884674,
+      "logps/chosen": -6.759585380554199,
+      "logps/rejected": -51.10936737060547,
+      "loss": 0.5104,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.859071731567383,
+      "rewards/margins": 4.638372898101807,
+      "rewards/rejected": 12.220698356628418,
+      "step": 220
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 46.97845458984375,
+      "learning_rate": 4.836904921750224e-06,
+      "logits/chosen": -0.11947059631347656,
+      "logits/rejected": -0.1329912692308426,
+      "logps/chosen": -3.608184814453125,
+      "logps/rejected": -48.794761657714844,
+      "loss": 0.2134,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.235904693603516,
+      "rewards/margins": 4.859888076782227,
+      "rewards/rejected": 12.376014709472656,
+      "step": 230
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 24.032859802246094,
+      "learning_rate": 4.808211074945042e-06,
+      "logits/chosen": -0.1200513243675232,
+      "logits/rejected": -0.1333036869764328,
+      "logps/chosen": -3.7552154064178467,
+      "logps/rejected": -49.87453079223633,
+      "loss": 0.1781,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.094650268554688,
+      "rewards/margins": 4.68077278137207,
+      "rewards/rejected": 12.41387939453125,
+      "step": 240
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "eval_logits/chosen": -0.12433278560638428,
+      "eval_logits/rejected": -0.13808581233024597,
+      "eval_logps/chosen": -4.408891201019287,
+      "eval_logps/rejected": -50.744781494140625,
+      "eval_loss": 0.1546352356672287,
+      "eval_rewards/accuracies": 0.9500000476837158,
+      "eval_rewards/chosen": 17.06200408935547,
+      "eval_rewards/margins": 4.775761604309082,
+      "eval_rewards/rejected": 12.286243438720703,
+      "eval_runtime": 361.4974,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 240
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.25737640261650085,
+      "learning_rate": 4.7772929998339485e-06,
+      "logits/chosen": -0.12348780035972595,
+      "logits/rejected": -0.13704943656921387,
+      "logps/chosen": -4.4299187660217285,
+      "logps/rejected": -53.074607849121094,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.087068557739258,
+      "rewards/margins": 5.06691837310791,
+      "rewards/rejected": 12.020149230957031,
+      "step": 250
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 0.1839389204978943,
+      "learning_rate": 4.744180489557859e-06,
+      "logits/chosen": -0.12177034467458725,
+      "logits/rejected": -0.1342695653438568,
+      "logps/chosen": -3.775188446044922,
+      "logps/rejected": -53.98720932006836,
+      "loss": 0.1896,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.12021255493164,
+      "rewards/margins": 5.148064613342285,
+      "rewards/rejected": 11.972146987915039,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.258485794067383,
+      "learning_rate": 4.708905451849754e-06,
+      "logits/chosen": -0.11067859083414078,
+      "logits/rejected": -0.12377731502056122,
+      "logps/chosen": -6.418317794799805,
+      "logps/rejected": -56.57402801513672,
+      "loss": 0.2315,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.738832473754883,
+      "rewards/margins": 4.884931564331055,
+      "rewards/rejected": 11.853900909423828,
+      "step": 270
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 77.56194305419922,
+      "learning_rate": 4.671501878287879e-06,
+      "logits/chosen": -0.1184445172548294,
+      "logits/rejected": -0.1339874565601349,
+      "logps/chosen": -10.12116527557373,
+      "logps/rejected": -53.403907775878906,
+      "loss": 0.5343,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 16.458633422851562,
+      "rewards/margins": 4.402472496032715,
+      "rewards/rejected": 12.056160926818848,
+      "step": 280
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 67.53883361816406,
+      "learning_rate": 4.6320058115409295e-06,
+      "logits/chosen": -0.1448262631893158,
+      "logits/rejected": -0.15793387591838837,
+      "logps/chosen": -3.4666190147399902,
+      "logps/rejected": -48.79213333129883,
+      "loss": 0.5017,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 16.945899963378906,
+      "rewards/margins": 4.2686333656311035,
+      "rewards/rejected": 12.677268028259277,
+      "step": 290
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.17521341145038605,
+      "learning_rate": 4.590455310636778e-06,
+      "logits/chosen": -0.16128253936767578,
+      "logits/rejected": -0.17375555634498596,
+      "logps/chosen": -2.9032950401306152,
+      "logps/rejected": -47.69734191894531,
+      "loss": 0.265,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.18383026123047,
+      "rewards/margins": 4.541309356689453,
+      "rewards/rejected": 12.642518997192383,
+      "step": 300
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "eval_logits/chosen": -0.17444846034049988,
+      "eval_logits/rejected": -0.18559777736663818,
+      "eval_logps/chosen": -2.535512924194336,
+      "eval_logps/rejected": -47.16367721557617,
+      "eval_loss": 0.15360687673091888,
+      "eval_rewards/accuracies": 0.9440000653266907,
+      "eval_rewards/chosen": 17.249343872070312,
+      "eval_rewards/margins": 4.604989051818848,
+      "eval_rewards/rejected": 12.644353866577148,
+      "eval_runtime": 361.4575,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 300
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 0.5040452480316162,
+      "learning_rate": 4.54689041428819e-06,
+      "logits/chosen": -0.16974106431007385,
+      "logits/rejected": -0.1810058057308197,
+      "logps/chosen": -1.233938217163086,
+      "logps/rejected": -49.907745361328125,
+      "loss": 0.1132,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.34117889404297,
+      "rewards/margins": 4.934173583984375,
+      "rewards/rejected": 12.407005310058594,
+      "step": 310
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 100.02949523925781,
+      "learning_rate": 4.501353102310901e-06,
+      "logits/chosen": -0.15705889463424683,
+      "logits/rejected": -0.1695334017276764,
+      "logps/chosen": -1.0820492506027222,
+      "logps/rejected": -52.577110290527344,
+      "loss": 0.1194,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.33388900756836,
+      "rewards/margins": 5.154760837554932,
+      "rewards/rejected": 12.179126739501953,
+      "step": 320
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 0.2689219117164612,
+      "learning_rate": 4.453887255171206e-06,
+      "logits/chosen": -0.13849371671676636,
+      "logits/rejected": -0.14990833401679993,
+      "logps/chosen": -1.8435032367706299,
+      "logps/rejected": -54.79044723510742,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.2423095703125,
+      "rewards/margins": 5.28987979888916,
+      "rewards/rejected": 11.952428817749023,
+      "step": 330
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 0.09305431693792343,
+      "learning_rate": 4.404538611702055e-06,
+      "logits/chosen": -0.12299702316522598,
+      "logits/rejected": -0.13453055918216705,
+      "logps/chosen": -2.9897143840789795,
+      "logps/rejected": -52.954498291015625,
+      "loss": 0.2873,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.17474365234375,
+      "rewards/margins": 5.071004867553711,
+      "rewards/rejected": 12.103739738464355,
+      "step": 340
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 59.282073974609375,
+      "learning_rate": 4.3533547250284015e-06,
+      "logits/chosen": -0.11913029849529266,
+      "logits/rejected": -0.12785324454307556,
+      "logps/chosen": -3.9456872940063477,
+      "logps/rejected": -48.68487548828125,
+      "loss": 0.4332,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 17.12805938720703,
+      "rewards/margins": 4.669450283050537,
+      "rewards/rejected": 12.458610534667969,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31101909279823303,
+      "learning_rate": 4.300384916744261e-06,
+      "logits/chosen": -0.11280188709497452,
+      "logits/rejected": -0.12300585210323334,
+      "logps/chosen": -2.1714723110198975,
+      "logps/rejected": -54.74174118041992,
+      "loss": 0.1605,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.326162338256836,
+      "rewards/margins": 5.467062473297119,
+      "rewards/rejected": 11.859098434448242,
+      "step": 360
+    },
+    {
+      "epoch": 0.32,
+      "eval_logits/chosen": -0.10620756447315216,
+      "eval_logits/rejected": -0.11727114766836166,
+      "eval_logps/chosen": -1.4165427684783936,
+      "eval_logps/rejected": -50.9525146484375,
+      "eval_loss": 0.3194349706172943,
+      "eval_rewards/accuracies": 0.9210000038146973,
+      "eval_rewards/chosen": 17.36124038696289,
+      "eval_rewards/margins": 5.095769882202148,
+      "eval_rewards/rejected": 12.26546859741211,
+      "eval_runtime": 361.5072,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 360
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 6.1126532554626465,
+      "learning_rate": 4.24568022938566e-06,
+      "logits/chosen": -0.10354311764240265,
+      "logits/rejected": -0.11526636779308319,
+      "logps/chosen": -1.2935255765914917,
+      "logps/rejected": -55.57566833496094,
+      "loss": 0.1711,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.439346313476562,
+      "rewards/margins": 5.700921058654785,
+      "rewards/rejected": 11.738424301147461,
+      "step": 370
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 34.15927505493164,
+      "learning_rate": 4.189293377245241e-06,
+      "logits/chosen": -0.1029932051897049,
+      "logits/rejected": -0.11382515728473663,
+      "logps/chosen": -2.5132687091827393,
+      "logps/rejected": -55.50346374511719,
+      "loss": 0.4359,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.731037139892578,
+      "rewards/margins": 4.368172645568848,
+      "rewards/rejected": 12.362865447998047,
+      "step": 380
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 2.8422904014587402,
+      "learning_rate": 4.131278695575952e-06,
+      "logits/chosen": -0.10793520510196686,
+      "logits/rejected": -0.12109285593032837,
+      "logps/chosen": -3.014652729034424,
+      "logps/rejected": -53.98411560058594,
+      "loss": 0.2161,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.137393951416016,
+      "rewards/margins": 5.105995178222656,
+      "rewards/rejected": 12.03139877319336,
+      "step": 390
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 54.0329475402832,
+      "learning_rate": 4.071692088232743e-06,
+      "logits/chosen": -0.10393750667572021,
+      "logits/rejected": -0.11834606528282166,
+      "logps/chosen": -2.1508543491363525,
+      "logps/rejected": -45.60733413696289,
+      "loss": 0.2077,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.586124420166016,
+      "rewards/margins": 5.077212333679199,
+      "rewards/rejected": 12.5089111328125,
+      "step": 400
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 81.61144256591797,
+      "learning_rate": 4.010590973802737e-06,
+      "logits/chosen": -0.09564584493637085,
+      "logits/rejected": -0.10617707669734955,
+      "logps/chosen": -3.4572842121124268,
+      "logps/rejected": -50.92162322998047,
+      "loss": 0.2478,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 17.010910034179688,
+      "rewards/margins": 4.556198596954346,
+      "rewards/rejected": 12.454713821411133,
+      "step": 410
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 0.30974289774894714,
+      "learning_rate": 3.948034230275781e-06,
+      "logits/chosen": -0.09134417027235031,
+      "logits/rejected": -0.1020016297698021,
+      "logps/chosen": -5.046698570251465,
+      "logps/rejected": -48.908958435058594,
+      "loss": 0.2894,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.007888793945312,
+      "rewards/margins": 4.53641414642334,
+      "rewards/rejected": 12.471475601196289,
+      "step": 420
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "eval_logits/chosen": -0.09054450690746307,
+      "eval_logits/rejected": -0.10264354199171066,
+      "eval_logps/chosen": -1.913105845451355,
+      "eval_logps/rejected": -51.11127471923828,
+      "eval_loss": 0.16789735853672028,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.311582565307617,
+      "eval_rewards/margins": 5.061989784240723,
+      "eval_rewards/rejected": 12.249593734741211,
+      "eval_runtime": 361.5337,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 420
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 12.824393272399902,
+      "learning_rate": 3.884082138308699e-06,
+      "logits/chosen": -0.08666776865720749,
+      "logits/rejected": -0.0997733399271965,
+      "logps/chosen": -1.7306327819824219,
+      "logps/rejected": -54.273292541503906,
+      "loss": 0.2298,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.167621612548828,
+      "rewards/margins": 5.065673351287842,
+      "rewards/rejected": 12.101947784423828,
+      "step": 430
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.30713599920272827,
+      "learning_rate": 3.818796323137896e-06,
+      "logits/chosen": -0.09174907952547073,
+      "logits/rejected": -0.10376611351966858,
+      "logps/chosen": -1.489154577255249,
+      "logps/rejected": -54.580726623535156,
+      "loss": 0.2513,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.22280502319336,
+      "rewards/margins": 5.175349235534668,
+      "rewards/rejected": 12.047454833984375,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 87.4791488647461,
+      "learning_rate": 3.7522396951963303e-06,
+      "logits/chosen": -0.09688778221607208,
+      "logits/rejected": -0.10897806286811829,
+      "logps/chosen": -3.157695770263672,
+      "logps/rejected": -50.96417236328125,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.345651626586914,
+      "rewards/margins": 5.245656967163086,
+      "rewards/rejected": 12.099993705749512,
+      "step": 450
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 146.2008056640625,
+      "learning_rate": 3.684476389492026e-06,
+      "logits/chosen": -0.09378582239151001,
+      "logits/rejected": -0.10475654900074005,
+      "logps/chosen": -0.5611928701400757,
+      "logps/rejected": -56.518890380859375,
+      "loss": 0.1981,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.113712310791016,
+      "rewards/margins": 5.068872928619385,
+      "rewards/rejected": 12.044839859008789,
+      "step": 460
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.9137721061706543,
+      "learning_rate": 3.6155717038065783e-06,
+      "logits/chosen": -0.08695463836193085,
+      "logits/rejected": -0.09596743434667587,
+      "logps/chosen": -1.5298550128936768,
+      "logps/rejected": -50.27445983886719,
+      "loss": 0.2066,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.35186004638672,
+      "rewards/margins": 5.014693260192871,
+      "rewards/rejected": 12.337167739868164,
+      "step": 470
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 84.80391693115234,
+      "learning_rate": 3.545592035773192e-06,
+      "logits/chosen": -0.0746893435716629,
+      "logits/rejected": -0.08653923869132996,
+      "logps/chosen": -2.0052125453948975,
+      "logps/rejected": -57.502811431884766,
+      "loss": 0.1149,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.14373016357422,
+      "rewards/margins": 5.360415935516357,
+      "rewards/rejected": 11.783313751220703,
+      "step": 480
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "eval_logits/chosen": -0.07700399309396744,
+      "eval_logits/rejected": -0.08828537166118622,
+      "eval_logps/chosen": -4.48896598815918,
+      "eval_logps/rejected": -53.76282501220703,
+      "eval_loss": 0.29511645436286926,
+      "eval_rewards/accuracies": 0.9230000376701355,
+      "eval_rewards/chosen": 17.053997039794922,
+      "eval_rewards/margins": 5.069558143615723,
+      "eval_rewards/rejected": 11.984437942504883,
+      "eval_runtime": 361.5035,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 480
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 82.9616470336914,
+      "learning_rate": 3.4746048188948806e-06,
+      "logits/chosen": -0.06675051152706146,
+      "logits/rejected": -0.07860895991325378,
+      "logps/chosen": -4.162237167358398,
+      "logps/rejected": -54.77789306640625,
+      "loss": 0.2979,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.047603607177734,
+      "rewards/margins": 5.138361930847168,
+      "rewards/rejected": 11.909242630004883,
+      "step": 490
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.04293210059404373,
+      "learning_rate": 3.4026784575644887e-06,
+      "logits/chosen": -0.06424491107463837,
+      "logits/rejected": -0.07567107677459717,
+      "logps/chosen": -2.05729603767395,
+      "logps/rejected": -56.646087646484375,
+      "loss": 0.4378,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.947803497314453,
+      "rewards/margins": 4.919981956481934,
+      "rewards/rejected": 12.02782154083252,
+      "step": 500
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 0.07136644423007965,
+      "learning_rate": 3.329882261149148e-06,
+      "logits/chosen": -0.06423303484916687,
+      "logits/rejected": -0.07512776553630829,
+      "logps/chosen": -3.1519265174865723,
+      "logps/rejected": -53.53008270263672,
+      "loss": 0.2613,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.112144470214844,
+      "rewards/margins": 5.046430587768555,
+      "rewards/rejected": 12.065712928771973,
+      "step": 510
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 0.29279613494873047,
+      "learning_rate": 3.25628637720269e-06,
+      "logits/chosen": -0.060233693569898605,
+      "logits/rejected": -0.07076811790466309,
+      "logps/chosen": -1.2358124256134033,
+      "logps/rejected": -50.61806106567383,
+      "loss": 0.1517,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.457351684570312,
+      "rewards/margins": 5.2062835693359375,
+      "rewards/rejected": 12.251070022583008,
+      "step": 520
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 0.7946074604988098,
+      "learning_rate": 3.181961723870359e-06,
+      "logits/chosen": -0.054482050240039825,
+      "logits/rejected": -0.0657092034816742,
+      "logps/chosen": -0.7682158946990967,
+      "logps/rejected": -58.81409454345703,
+      "loss": 0.2304,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.075973510742188,
+      "rewards/margins": 5.222441673278809,
+      "rewards/rejected": 11.853530883789062,
+      "step": 530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1405210494995117,
+      "learning_rate": 3.1069799215509847e-06,
+      "logits/chosen": -0.05050881579518318,
+      "logits/rejected": -0.061149902641773224,
+      "logps/chosen": -0.954046368598938,
+      "logps/rejected": -52.70227813720703,
+      "loss": 0.0384,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.378849029541016,
+      "rewards/margins": 5.253483295440674,
+      "rewards/rejected": 12.125364303588867,
+      "step": 540
+    },
+    {
+      "epoch": 0.48,
+      "eval_logits/chosen": -0.05120665580034256,
+      "eval_logits/rejected": -0.06123337894678116,
+      "eval_logps/chosen": -2.9872913360595703,
+      "eval_logps/rejected": -52.27314758300781,
+      "eval_loss": 0.17387841641902924,
+      "eval_rewards/accuracies": 0.9490000605583191,
+      "eval_rewards/chosen": 17.204164505004883,
+      "eval_rewards/margins": 5.070757865905762,
+      "eval_rewards/rejected": 12.133406639099121,
+      "eval_runtime": 361.5449,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 540
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.07270358502864838,
+      "learning_rate": 3.0314132238824416e-06,
+      "logits/chosen": -0.05125313252210617,
+      "logits/rejected": -0.06174170970916748,
+      "logps/chosen": -3.2163877487182617,
+      "logps/rejected": -47.79279327392578,
+      "loss": 0.2087,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.408517837524414,
+      "rewards/margins": 5.061524391174316,
+      "rewards/rejected": 12.346992492675781,
+      "step": 550
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.10005924850702286,
+      "learning_rate": 2.955334448116915e-06,
+      "logits/chosen": -0.041773442178964615,
+      "logits/rejected": -0.05364570394158363,
+      "logps/chosen": -0.363404780626297,
+      "logps/rejected": -56.32415008544922,
+      "loss": 0.0969,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.3824462890625,
+      "rewards/margins": 5.5404510498046875,
+      "rewards/rejected": 11.841995239257812,
+      "step": 560
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 7.818356990814209,
+      "learning_rate": 2.8788169049530533e-06,
+      "logits/chosen": -0.04309462010860443,
+      "logits/rejected": -0.05494442582130432,
+      "logps/chosen": -2.2242724895477295,
+      "logps/rejected": -56.444740295410156,
+      "loss": 0.1447,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.275390625,
+      "rewards/margins": 5.55007791519165,
+      "rewards/rejected": 11.725313186645508,
+      "step": 570
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 0.03519747406244278,
+      "learning_rate": 2.8019343278926397e-06,
+      "logits/chosen": -0.03506368771195412,
+      "logits/rejected": -0.046854715794324875,
+      "logps/chosen": -0.5200096964836121,
+      "logps/rejected": -59.05330276489258,
+      "loss": 0.0998,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.320554733276367,
+      "rewards/margins": 5.727260589599609,
+      "rewards/rejected": 11.593294143676758,
+      "step": 580
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 0.04519123584032059,
+      "learning_rate": 2.7247608021898265e-06,
+      "logits/chosen": -0.03204537555575371,
+      "logits/rejected": -0.04383891448378563,
+      "logps/chosen": -1.1271060705184937,
+      "logps/rejected": -59.308895111083984,
+      "loss": 0.1162,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.256351470947266,
+      "rewards/margins": 5.6881890296936035,
+      "rewards/rejected": 11.568161010742188,
+      "step": 590
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.028689857572317123,
+      "learning_rate": 2.647370693461432e-06,
+      "logits/chosen": -0.02834726870059967,
+      "logits/rejected": -0.03827046602964401,
+      "logps/chosen": -5.673943519592285,
+      "logps/rejected": -55.72624588012695,
+      "loss": 0.4008,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.761056900024414,
+      "rewards/margins": 4.804043769836426,
+      "rewards/rejected": 11.957012176513672,
+      "step": 600
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "eval_logits/chosen": -0.035840023308992386,
+      "eval_logits/rejected": -0.04687971621751785,
+      "eval_logps/chosen": -2.1760435104370117,
+      "eval_logps/rejected": -56.62664031982422,
+      "eval_loss": 0.17061151564121246,
+      "eval_rewards/accuracies": 0.9470000267028809,
+      "eval_rewards/chosen": 17.285289764404297,
+      "eval_rewards/margins": 5.587231636047363,
+      "eval_rewards/rejected": 11.698057174682617,
+      "eval_runtime": 361.5056,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 600
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 1.1994622945785522,
+      "learning_rate": 2.569838576027068e-06,
+      "logits/chosen": -0.03231767192482948,
+      "logits/rejected": -0.04397805407643318,
+      "logps/chosen": -1.5904741287231445,
+      "logps/rejected": -62.10520553588867,
+      "loss": 0.1265,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.11206817626953,
+      "rewards/margins": 5.734784126281738,
+      "rewards/rejected": 11.37728500366211,
+      "step": 610
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 0.14163845777511597,
+      "learning_rate": 2.4922391610481544e-06,
+      "logits/chosen": -0.03293662518262863,
+      "logits/rejected": -0.04335154965519905,
+      "logps/chosen": -1.3568997383117676,
+      "logps/rejected": -58.347434997558594,
+      "loss": 0.097,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.345399856567383,
+      "rewards/margins": 5.823373794555664,
+      "rewards/rejected": 11.522026062011719,
+      "step": 620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.125027060508728,
+      "learning_rate": 2.4146472245350804e-06,
+      "logits/chosen": -0.02864963933825493,
+      "logits/rejected": -0.03835710883140564,
+      "logps/chosen": -2.494175434112549,
+      "logps/rejected": -55.33067321777344,
+      "loss": 0.3115,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.262205123901367,
+      "rewards/margins": 5.4287261962890625,
+      "rewards/rejected": 11.833479881286621,
+      "step": 630
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.03514016419649124,
+      "learning_rate": 2.337137535291868e-06,
+      "logits/chosen": -0.02757749892771244,
+      "logits/rejected": -0.0376611053943634,
+      "logps/chosen": -2.3163387775421143,
+      "logps/rejected": -53.54579162597656,
+      "loss": 0.1268,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.542863845825195,
+      "rewards/margins": 5.799897193908691,
+      "rewards/rejected": 11.742965698242188,
+      "step": 640
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 34.20791244506836,
+      "learning_rate": 2.259784782867782e-06,
+      "logits/chosen": -0.02032250165939331,
+      "logits/rejected": -0.031542714685201645,
+      "logps/chosen": -1.7910137176513672,
+      "logps/rejected": -62.980018615722656,
+      "loss": 0.1831,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.982879638671875,
+      "rewards/margins": 5.589078903198242,
+      "rewards/rejected": 11.393800735473633,
+      "step": 650
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 32.356956481933594,
+      "learning_rate": 2.182663505585314e-06,
+      "logits/chosen": -0.01711965538561344,
+      "logits/rejected": -0.028038471937179565,
+      "logps/chosen": -2.662904739379883,
+      "logps/rejected": -63.326297760009766,
+      "loss": 0.1678,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 16.997777938842773,
+      "rewards/margins": 5.741157531738281,
+      "rewards/rejected": 11.256620407104492,
+      "step": 660
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "eval_logits/chosen": -0.016018809750676155,
+      "eval_logits/rejected": -0.027020033448934555,
+      "eval_logps/chosen": -3.00822114944458,
+      "eval_logps/rejected": -57.951629638671875,
+      "eval_loss": 0.2050127536058426,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.202072143554688,
+      "eval_rewards/margins": 5.636512756347656,
+      "eval_rewards/rejected": 11.565557479858398,
+      "eval_runtime": 361.5073,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 660
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 86.53874969482422,
+      "learning_rate": 2.1058480187138863e-06,
+      "logits/chosen": -0.012352555990219116,
+      "logits/rejected": -0.023966707289218903,
+      "logps/chosen": -2.129209041595459,
+      "logps/rejected": -62.1393928527832,
+      "loss": 0.1025,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.069538116455078,
+      "rewards/margins": 5.678930759429932,
+      "rewards/rejected": 11.390605926513672,
+      "step": 670
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 2.892672300338745,
+      "learning_rate": 2.0294123428584985e-06,
+      "logits/chosen": -0.011562807485461235,
+      "logits/rejected": -0.020859256386756897,
+      "logps/chosen": -3.2486608028411865,
+      "logps/rejected": -59.56721496582031,
+      "loss": 0.1961,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.051376342773438,
+      "rewards/margins": 5.515361785888672,
+      "rewards/rejected": 11.536016464233398,
+      "step": 680
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 30.26588249206543,
+      "learning_rate": 1.953430132632311e-06,
+      "logits/chosen": -0.011488726362586021,
+      "logits/rejected": -0.021527227014303207,
+      "logps/chosen": -1.794357180595398,
+      "logps/rejected": -60.649505615234375,
+      "loss": 0.2821,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.114843368530273,
+      "rewards/margins": 5.612217903137207,
+      "rewards/rejected": 11.50262451171875,
+      "step": 690
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 0.030314341187477112,
+      "learning_rate": 1.8779746056819104e-06,
+      "logits/chosen": -0.014436552301049232,
+      "logits/rejected": -0.026044374331831932,
+      "logps/chosen": -3.1617178916931152,
+      "logps/rejected": -60.94548797607422,
+      "loss": 0.3141,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.08698272705078,
+      "rewards/margins": 5.719264984130859,
+      "rewards/rejected": 11.367716789245605,
+      "step": 700
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 0.06711317598819733,
+      "learning_rate": 1.8031184721336364e-06,
+      "logits/chosen": -0.017312290146946907,
+      "logits/rejected": -0.026554957032203674,
+      "logps/chosen": -5.120705604553223,
+      "logps/rejected": -54.33483123779297,
+      "loss": 0.3827,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 17.1247501373291,
+      "rewards/margins": 5.327882766723633,
+      "rewards/rejected": 11.796867370605469,
+      "step": 710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 60.36637496948242,
+      "learning_rate": 1.7289338645289711e-06,
+      "logits/chosen": -0.01987219974398613,
+      "logits/rejected": -0.02766304835677147,
+      "logps/chosen": -2.1757986545562744,
+      "logps/rejected": -54.02531433105469,
+      "loss": 0.2272,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.239776611328125,
+      "rewards/margins": 5.250313758850098,
+      "rewards/rejected": 11.989462852478027,
+      "step": 720
+    },
+    {
+      "epoch": 0.64,
+      "eval_logits/chosen": -0.022905193269252777,
+      "eval_logits/rejected": -0.0321992002427578,
+      "eval_logps/chosen": -1.100506067276001,
+      "eval_logps/rejected": -55.91169738769531,
+      "eval_loss": 0.14018221199512482,
+      "eval_rewards/accuracies": 0.9520000219345093,
+      "eval_rewards/chosen": 17.392841339111328,
+      "eval_rewards/margins": 5.623291492462158,
+      "eval_rewards/rejected": 11.769551277160645,
+      "eval_runtime": 361.416,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 720
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 0.029294608160853386,
+      "learning_rate": 1.6554922683164875e-06,
+      "logits/chosen": -0.016416028141975403,
+      "logits/rejected": -0.025527067482471466,
+      "logps/chosen": -0.906692385673523,
+      "logps/rejected": -59.142173767089844,
+      "loss": 0.1171,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.282169342041016,
+      "rewards/margins": 5.6978583335876465,
+      "rewards/rejected": 11.584310531616211,
+      "step": 730
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 33.89070510864258,
+      "learning_rate": 1.5828644529673592e-06,
+      "logits/chosen": -0.016044551506638527,
+      "logits/rejected": -0.025811903178691864,
+      "logps/chosen": -1.1167538166046143,
+      "logps/rejected": -59.84492492675781,
+      "loss": 0.2001,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.266983032226562,
+      "rewards/margins": 5.749438285827637,
+      "rewards/rejected": 11.51754379272461,
+      "step": 740
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.53726726770401,
+      "learning_rate": 1.5111204037807844e-06,
+      "logits/chosen": -0.012529855594038963,
+      "logits/rejected": -0.023765765130519867,
+      "logps/chosen": -0.3461765646934509,
+      "logps/rejected": -55.841102600097656,
+      "loss": 0.1692,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.61865997314453,
+      "rewards/margins": 6.019055366516113,
+      "rewards/rejected": 11.599604606628418,
+      "step": 750
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.7173987627029419,
+      "learning_rate": 1.4403292544450625e-06,
+      "logits/chosen": -0.013380522839725018,
+      "logits/rejected": -0.022107835859060287,
+      "logps/chosen": -3.2759666442871094,
+      "logps/rejected": -53.836822509765625,
+      "loss": 0.2418,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.383333206176758,
+      "rewards/margins": 5.590358734130859,
+      "rewards/rejected": 11.792974472045898,
+      "step": 760
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 2.744900941848755,
+      "learning_rate": 1.3705592204192853e-06,
+      "logits/chosen": -0.011318420059978962,
+      "logits/rejected": -0.021840626373887062,
+      "logps/chosen": -2.592001438140869,
+      "logps/rejected": -57.06926345825195,
+      "loss": 0.1749,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.286128997802734,
+      "rewards/margins": 5.662715911865234,
+      "rewards/rejected": 11.623414993286133,
+      "step": 770
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 70.76551818847656,
+      "learning_rate": 1.301877533199859e-06,
+      "logits/chosen": -0.013944407925009727,
+      "logits/rejected": -0.025667501613497734,
+      "logps/chosen": -2.1070234775543213,
+      "logps/rejected": -57.72039794921875,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.28545379638672,
+      "rewards/margins": 5.708344459533691,
+      "rewards/rejected": 11.577108383178711,
+      "step": 780
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "eval_logits/chosen": -0.01664295792579651,
+      "eval_logits/rejected": -0.026626665145158768,
+      "eval_logps/chosen": -1.082255244255066,
+      "eval_logps/rejected": -55.95073699951172,
+      "eval_loss": 0.24412688612937927,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.394668579101562,
+      "eval_rewards/margins": 5.629020690917969,
+      "eval_rewards/rejected": 11.765647888183594,
+      "eval_runtime": 361.4602,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 780
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 0.41359376907348633,
+      "learning_rate": 1.2343503755351729e-06,
+      "logits/chosen": -0.012772129848599434,
+      "logits/rejected": -0.023186586797237396,
+      "logps/chosen": -0.8030359148979187,
+      "logps/rejected": -57.84947967529297,
+      "loss": 0.2143,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.32242202758789,
+      "rewards/margins": 5.645486831665039,
+      "rewards/rejected": 11.676933288574219,
+      "step": 790
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.41748157143592834,
+      "learning_rate": 1.168042817650881e-06,
+      "logits/chosen": -0.012502101249992847,
+      "logits/rejected": -0.023272844031453133,
+      "logps/chosen": -1.0377256870269775,
+      "logps/rejected": -57.05615997314453,
+      "loss": 0.0792,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.60501480102539,
+      "rewards/margins": 6.159370422363281,
+      "rewards/rejected": 11.445646286010742,
+      "step": 800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.039210401475429535,
+      "learning_rate": 1.1030187545472012e-06,
+      "logits/chosen": -0.008177272044122219,
+      "logits/rejected": -0.01832464337348938,
+      "logps/chosen": -3.8701748847961426,
+      "logps/rejected": -57.661476135253906,
+      "loss": 0.3177,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.14336395263672,
+      "rewards/margins": 5.583393573760986,
+      "rewards/rejected": 11.55997085571289,
+      "step": 810
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.6096951961517334,
+      "learning_rate": 1.0393408444287048e-06,
+      "logits/chosen": -0.006832236424088478,
+      "logits/rejected": -0.01682097464799881,
+      "logps/chosen": -2.1741790771484375,
+      "logps/rejected": -57.273414611816406,
+      "loss": 0.2128,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.426712036132812,
+      "rewards/margins": 5.911205291748047,
+      "rewards/rejected": 11.515506744384766,
+      "step": 820
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 3.7820959091186523,
+      "learning_rate": 9.770704483258782e-07,
+      "logits/chosen": -0.009998206980526447,
+      "logits/rejected": -0.0204261876642704,
+      "logps/chosen": -1.9802953004837036,
+      "logps/rejected": -56.901512145996094,
+      "loss": 0.1296,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.56071662902832,
+      "rewards/margins": 6.169583320617676,
+      "rewards/rejected": 11.391134262084961,
+      "step": 830
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.126626968383789,
+      "learning_rate": 9.162675709666865e-07,
+      "logits/chosen": -0.00826224498450756,
+      "logits/rejected": -0.018977787345647812,
+      "logps/chosen": -1.4256607294082642,
+      "logps/rejected": -61.65986633300781,
+      "loss": 0.0635,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.324975967407227,
+      "rewards/margins": 6.09440803527832,
+      "rewards/rejected": 11.230567932128906,
+      "step": 840
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "eval_logits/chosen": -0.01111944392323494,
+      "eval_logits/rejected": -0.021697774529457092,
+      "eval_logps/chosen": -1.2168633937835693,
+      "eval_logps/rejected": -58.2642822265625,
+      "eval_loss": 0.1689341962337494,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.381206512451172,
+      "eval_rewards/margins": 5.846914291381836,
+      "eval_rewards/rejected": 11.534292221069336,
+      "eval_runtime": 361.6192,
+      "eval_samples_per_second": 2.765,
+      "eval_steps_per_second": 0.346,
+      "step": 840
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 5.250723838806152,
+      "learning_rate": 8.569908029550686e-07,
+      "logits/chosen": -0.006854387000203133,
+      "logits/rejected": -0.018336206674575806,
+      "logps/chosen": -0.6238930821418762,
+      "logps/rejected": -60.925689697265625,
+      "loss": 0.1157,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.443281173706055,
+      "rewards/margins": 6.164813995361328,
+      "rewards/rejected": 11.278467178344727,
+      "step": 850
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 3.1401162147521973,
+      "learning_rate": 7.992972643121227e-07,
+      "logits/chosen": -0.0037835021503269672,
+      "logits/rejected": -0.013135241344571114,
+      "logps/chosen": -0.8492221832275391,
+      "logps/rejected": -55.516075134277344,
+      "loss": 0.2252,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.488588333129883,
+      "rewards/margins": 5.736725807189941,
+      "rewards/rejected": 11.751862525939941,
+      "step": 860
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 55.528812408447266,
+      "learning_rate": 7.432425494343509e-07,
+      "logits/chosen": -0.0033687639515846968,
+      "logits/rejected": -0.013152632862329483,
+      "logps/chosen": -1.3188884258270264,
+      "logps/rejected": -57.9510498046875,
+      "loss": 0.1398,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.412578582763672,
+      "rewards/margins": 5.868515968322754,
+      "rewards/rejected": 11.544061660766602,
+      "step": 870
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.039824869483709335,
+      "learning_rate": 6.888806735220396e-07,
+      "logits/chosen": -0.0010406378423795104,
+      "logits/rejected": -0.012095071375370026,
+      "logps/chosen": -2.0619027614593506,
+      "logps/rejected": -59.65806579589844,
+      "loss": 0.2966,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.192535400390625,
+      "rewards/margins": 5.7006731033325195,
+      "rewards/rejected": 11.491861343383789,
+      "step": 880
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 0.7536466717720032,
+      "learning_rate": 6.362640205293583e-07,
+      "logits/chosen": -0.0016857212176546454,
+      "logits/rejected": -0.010936147533357143,
+      "logps/chosen": -2.1478958129882812,
+      "logps/rejected": -58.20386505126953,
+      "loss": 0.1641,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.243247985839844,
+      "rewards/margins": 5.678771018981934,
+      "rewards/rejected": 11.564477920532227,
+      "step": 890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3857377767562866,
+      "learning_rate": 5.854432926863684e-07,
+      "logits/chosen": 0.00038508616853505373,
+      "logits/rejected": -0.011034643277525902,
+      "logps/chosen": -1.606274127960205,
+      "logps/rejected": -62.58662414550781,
+      "loss": 0.1703,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.211261749267578,
+      "rewards/margins": 5.966868877410889,
+      "rewards/rejected": 11.244392395019531,
+      "step": 900
+    },
+    {
+      "epoch": 0.8,
+      "eval_logits/chosen": 0.00021816430671606213,
+      "eval_logits/rejected": -0.010477552190423012,
+      "eval_logps/chosen": -1.7576563358306885,
+      "eval_logps/rejected": -59.79063415527344,
+      "eval_loss": 0.13995186984539032,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.327129364013672,
+      "eval_rewards/margins": 5.945469856262207,
+      "eval_rewards/rejected": 11.381658554077148,
+      "eval_runtime": 361.4592,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 900
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 0.13143697381019592,
+      "learning_rate": 5.364674616415547e-07,
+      "logits/chosen": 0.0005570838693529367,
+      "logits/rejected": -0.011198626831173897,
+      "logps/chosen": -0.12791283428668976,
+      "logps/rejected": -65.7052993774414,
+      "loss": 0.0507,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.224443435668945,
+      "rewards/margins": 6.163690567016602,
+      "rewards/rejected": 11.060752868652344,
+      "step": 910
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 102.9654541015625,
+      "learning_rate": 4.893837212719859e-07,
+      "logits/chosen": -0.0008557128603570163,
+      "logits/rejected": -0.01163212489336729,
+      "logps/chosen": -1.3292646408081055,
+      "logps/rejected": -61.644893646240234,
+      "loss": 0.0881,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.276538848876953,
+      "rewards/margins": 5.96080207824707,
+      "rewards/rejected": 11.3157377243042,
+      "step": 920
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 146.04498291015625,
+      "learning_rate": 4.442374422065493e-07,
+      "logits/chosen": 0.002922601066529751,
+      "logits/rejected": -0.007130052894353867,
+      "logps/chosen": -1.6069023609161377,
+      "logps/rejected": -61.52588653564453,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.184974670410156,
+      "rewards/margins": 5.7995734214782715,
+      "rewards/rejected": 11.38540267944336,
+      "step": 930
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 23.539485931396484,
+      "learning_rate": 4.0107212810610974e-07,
+      "logits/chosen": 0.0018056132830679417,
+      "logits/rejected": -0.007847340777516365,
+      "logps/chosen": -3.637047290802002,
+      "logps/rejected": -61.21245574951172,
+      "loss": 0.2763,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.960758209228516,
+      "rewards/margins": 5.551811695098877,
+      "rewards/rejected": 11.40894603729248,
+      "step": 940
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 31.52926254272461,
+      "learning_rate": 3.599293737426932e-07,
+      "logits/chosen": 0.0032081177923828363,
+      "logits/rejected": -0.007756482809782028,
+      "logps/chosen": -1.172515869140625,
+      "logps/rejected": -66.5853271484375,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 16.992876052856445,
+      "rewards/margins": 5.886469841003418,
+      "rewards/rejected": 11.106407165527344,
+      "step": 950
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.07474468648433685,
+      "learning_rate": 3.208488249181216e-07,
+      "logits/chosen": 0.0022508346009999514,
+      "logits/rejected": -0.009156409651041031,
+      "logps/chosen": -0.7930470108985901,
+      "logps/rejected": -60.068790435791016,
+      "loss": 0.1138,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.401771545410156,
+      "rewards/margins": 6.0026960372924805,
+      "rewards/rejected": 11.399076461791992,
+      "step": 960
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "eval_logits/chosen": 0.001469604205340147,
+      "eval_logits/rejected": -0.009397665038704872,
+      "eval_logps/chosen": -1.8795456886291504,
+      "eval_logps/rejected": -60.17564010620117,
+      "eval_loss": 0.1441129744052887,
+      "eval_rewards/accuracies": 0.9630000591278076,
+      "eval_rewards/chosen": 17.314937591552734,
+      "eval_rewards/margins": 5.9717817306518555,
+      "eval_rewards/rejected": 11.343156814575195,
+      "eval_runtime": 361.5344,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 960
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 68.90747833251953,
+      "learning_rate": 2.838681402606952e-07,
+      "logits/chosen": 0.004552370868623257,
+      "logits/rejected": -0.005488495342433453,
+      "logps/chosen": -3.7298974990844727,
+      "logps/rejected": -64.72488403320312,
+      "loss": 0.2425,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.832754135131836,
+      "rewards/margins": 5.644216537475586,
+      "rewards/rejected": 11.18853759765625,
+      "step": 970
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 0.028206102550029755,
+      "learning_rate": 2.490229549367443e-07,
+      "logits/chosen": 0.0025807656347751617,
+      "logits/rejected": -0.008657123893499374,
+      "logps/chosen": -1.1825838088989258,
+      "logps/rejected": -63.57493591308594,
+      "loss": 0.0609,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.227130889892578,
+      "rewards/margins": 6.053717136383057,
+      "rewards/rejected": 11.173412322998047,
+      "step": 980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39742231369018555,
+      "learning_rate": 2.1634684631203412e-07,
+      "logits/chosen": 0.0048486413434147835,
+      "logits/rejected": -0.006056814920157194,
+      "logps/chosen": -2.919680595397949,
+      "logps/rejected": -57.8403205871582,
+      "loss": 0.3464,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.175275802612305,
+      "rewards/margins": 5.559727668762207,
+      "rewards/rejected": 11.615548133850098,
+      "step": 990
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 30.22509002685547,
+      "learning_rate": 1.8587130159608196e-07,
+      "logits/chosen": 0.0050649940967559814,
+      "logits/rejected": -0.0071399761363863945,
+      "logps/chosen": -0.16131475567817688,
+      "logps/rejected": -66.25190734863281,
+      "loss": 0.0063,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 17.313472747802734,
+      "rewards/margins": 6.412895679473877,
+      "rewards/rejected": 10.9005765914917,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 53.02751922607422,
+      "learning_rate": 1.5762568750059604e-07,
+      "logits/chosen": 0.005312003195285797,
+      "logits/rejected": -0.0038596936501562595,
+      "logps/chosen": -4.00323486328125,
+      "logps/rejected": -57.656890869140625,
+      "loss": 0.2741,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.167781829833984,
+      "rewards/margins": 5.632095813751221,
+      "rewards/rejected": 11.535685539245605,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 119.56330108642578,
+      "learning_rate": 1.316372219412454e-07,
+      "logits/chosen": 0.004333779215812683,
+      "logits/rejected": -0.007937717251479626,
+      "logps/chosen": -0.10221245884895325,
+      "logps/rejected": -67.09977722167969,
+      "loss": 0.0513,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.151248931884766,
+      "rewards/margins": 6.162137985229492,
+      "rewards/rejected": 10.98911190032959,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "eval_logits/chosen": 0.004528110846877098,
+      "eval_logits/rejected": -0.006501312367618084,
+      "eval_logps/chosen": -1.817779302597046,
+      "eval_logps/rejected": -60.34454345703125,
+      "eval_loss": 0.14118175208568573,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.321117401123047,
+      "eval_rewards/margins": 5.994848251342773,
+      "eval_rewards/rejected": 11.32626724243164,
+      "eval_runtime": 361.545,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 0.9030271768569946,
+      "learning_rate": 1.0793094781005792e-07,
+      "logits/chosen": 0.0036234352737665176,
+      "logits/rejected": -0.008390933275222778,
+      "logps/chosen": -0.09899584949016571,
+      "logps/rejected": -60.80555725097656,
+      "loss": 0.0349,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.619266510009766,
+      "rewards/margins": 6.4814043045043945,
+      "rewards/rejected": 11.137863159179688,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 157.070068359375,
+      "learning_rate": 8.652970884369255e-08,
+      "logits/chosen": 0.005321115255355835,
+      "logits/rejected": -0.004976513795554638,
+      "logps/chosen": -1.3535833358764648,
+      "logps/rejected": -61.2861328125,
+      "loss": 0.1927,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.209800720214844,
+      "rewards/margins": 5.795952796936035,
+      "rewards/rejected": 11.413846015930176,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 24.539953231811523,
+      "learning_rate": 6.745412761086007e-08,
+      "logits/chosen": 0.005845514126121998,
+      "logits/rejected": -0.003921338357031345,
+      "logps/chosen": -1.7223193645477295,
+      "logps/rejected": -58.172142028808594,
+      "loss": 0.1317,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.360435485839844,
+      "rewards/margins": 5.838218688964844,
+      "rewards/rejected": 11.522216796875,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 0.061458222568035126,
+      "learning_rate": 5.0722585640090305e-08,
+      "logits/chosen": 0.003759522922337055,
+      "logits/rejected": -0.0068417866714298725,
+      "logps/chosen": -1.351855754852295,
+      "logps/rejected": -59.56566619873047,
+      "loss": 0.2279,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.245838165283203,
+      "rewards/margins": 5.7038679122924805,
+      "rewards/rejected": 11.541970252990723,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 101.87332916259766,
+      "learning_rate": 3.635120570700784e-08,
+      "logits/chosen": 0.007607857696712017,
+      "logits/rejected": -0.0035320711322128773,
+      "logps/chosen": -1.4004669189453125,
+      "logps/rejected": -63.46752166748047,
+      "loss": 0.1021,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.240036010742188,
+      "rewards/margins": 6.078882694244385,
+      "rewards/rejected": 11.161155700683594,
+      "step": 1070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9229660034179688,
+      "learning_rate": 2.4353836298169343e-08,
+      "logits/chosen": 0.006209026090800762,
+      "logits/rejected": -0.0033816141076385975,
+      "logps/chosen": -2.0281822681427,
+      "logps/rejected": -61.460670471191406,
+      "loss": 0.1189,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.19972801208496,
+      "rewards/margins": 5.881450653076172,
+      "rewards/rejected": 11.318277359008789,
+      "step": 1080
+    },
+    {
+      "epoch": 0.96,
+      "eval_logits/chosen": 0.00737445754930377,
+      "eval_logits/rejected": -0.0035832570865750313,
+      "eval_logps/chosen": -2.1419789791107178,
+      "eval_logps/rejected": -60.60612106323242,
+      "eval_loss": 0.15081512928009033,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.288694381713867,
+      "eval_rewards/margins": 5.98858642578125,
+      "eval_rewards/rejected": 11.300108909606934,
+      "eval_runtime": 361.4165,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 0.11888863146305084,
+      "learning_rate": 1.4742038266447046e-08,
+      "logits/chosen": 0.006043245084583759,
+      "logits/rejected": -0.004712546244263649,
+      "logps/chosen": -1.4127472639083862,
+      "logps/rejected": -64.81529998779297,
+      "loss": 0.0728,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.181049346923828,
+      "rewards/margins": 6.14100456237793,
+      "rewards/rejected": 11.040044784545898,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 2.001819372177124,
+      "learning_rate": 7.525073690809737e-09,
+      "logits/chosen": 0.0055408780463039875,
+      "logits/rejected": -0.005598037503659725,
+      "logps/chosen": -1.017110824584961,
+      "logps/rejected": -60.28044891357422,
+      "loss": 0.0357,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.4679012298584,
+      "rewards/margins": 6.193048477172852,
+      "rewards/rejected": 11.274852752685547,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 0.027056939899921417,
+      "learning_rate": 2.709896951238744e-09,
+      "logits/chosen": 0.00636716466397047,
+      "logits/rejected": -0.004360577557235956,
+      "logps/chosen": -2.8028905391693115,
+      "logps/rejected": -60.86518859863281,
+      "loss": 0.1973,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.210033416748047,
+      "rewards/margins": 5.936794281005859,
+      "rewards/rejected": 11.273238182067871,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 0.15637506544589996,
+      "learning_rate": 3.0114802737818415e-10,
+      "logits/chosen": 0.0038183885626494884,
+      "logits/rejected": -0.007296917960047722,
+      "logps/chosen": -1.1477452516555786,
+      "logps/rejected": -59.37739181518555,
+      "loss": 0.0538,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.553251266479492,
+      "rewards/margins": 6.2790141105651855,
+      "rewards/rejected": 11.274236679077148,
+      "step": 1120
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1125,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 1.4338459346927616e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1125/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5286bf380083ef64133f25e161ddfb8549ff44790e432f478753f221ef89e695
+size 5368

checkpoint-500/README.md

@@ -0,0 +1,202 @@
+---
+base_model: mistralai/Mistral-Nemo-Instruct-2407
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.12.0

checkpoint-500/adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-Nemo-Instruct-2407",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "gate_proj",
+    "k_proj",
+    "o_proj",
+    "v_proj",
+    "down_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-500/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2405da5efa3fa58305eb789eb90dbe77348a677c94c781b71f5a7828902865ef
+size 114106856

checkpoint-500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83c427609b64317bf25a71f520f1925991aac70912f290b39b6953fa9c67daa4
+size 228536930

checkpoint-500/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91bd7f619e4cd37883f469c08e90105c4d218fd82ffc43ae58fa9fdbcc37fce5
+size 14512

checkpoint-500/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b0a7593f9ab52bf47328c6d50954dce1fcd69866aa6f5f35851aef7f7af3899
+size 14512

checkpoint-500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fd9d73796ba0c0da5d08a8e5aa955713fdd4d633eb58f1fea83f6389e8837c8
+size 1064

checkpoint-500/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-500/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

checkpoint-500/trainer_state.json

@@ -0,0 +1,911 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4444444444444444,
+  "eval_steps": 60,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 8.760091781616211,
+      "learning_rate": 4.4247787610619474e-07,
+      "logits/chosen": -0.8248252868652344,
+      "logits/rejected": -0.8263720273971558,
+      "logps/chosen": -0.36086463928222656,
+      "logps/rejected": -5.696224689483643,
+      "loss": 1.1038,
+      "rewards/accuracies": 0.5125000476837158,
+      "rewards/chosen": 17.43745994567871,
+      "rewards/margins": 0.5984855890274048,
+      "rewards/rejected": 16.838973999023438,
+      "step": 10
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 8.855981826782227,
+      "learning_rate": 8.849557522123895e-07,
+      "logits/chosen": -0.8169006109237671,
+      "logits/rejected": -0.819770872592926,
+      "logps/chosen": -0.12464660406112671,
+      "logps/rejected": -7.139842987060547,
+      "loss": 1.1887,
+      "rewards/accuracies": 0.4000000059604645,
+      "rewards/chosen": 17.17649269104004,
+      "rewards/margins": 0.19107049703598022,
+      "rewards/rejected": 16.98542022705078,
+      "step": 20
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 16.764184951782227,
+      "learning_rate": 1.3274336283185843e-06,
+      "logits/chosen": -0.8003113865852356,
+      "logits/rejected": -0.8030117750167847,
+      "logps/chosen": -0.34651467204093933,
+      "logps/rejected": -6.967917442321777,
+      "loss": 1.0563,
+      "rewards/accuracies": 0.44999998807907104,
+      "rewards/chosen": 17.280975341796875,
+      "rewards/margins": 0.40005987882614136,
+      "rewards/rejected": 16.88091468811035,
+      "step": 30
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 8.33682918548584,
+      "learning_rate": 1.769911504424779e-06,
+      "logits/chosen": -0.7695047855377197,
+      "logits/rejected": -0.7739207148551941,
+      "logps/chosen": -1.5993006229400635,
+      "logps/rejected": -8.504932403564453,
+      "loss": 0.7596,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 17.283912658691406,
+      "rewards/margins": 0.6976072192192078,
+      "rewards/rejected": 16.5863037109375,
+      "step": 40
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 4.494723320007324,
+      "learning_rate": 2.212389380530974e-06,
+      "logits/chosen": -0.7154140472412109,
+      "logits/rejected": -0.7225576043128967,
+      "logps/chosen": -3.112199068069458,
+      "logps/rejected": -12.212080001831055,
+      "loss": 0.6083,
+      "rewards/accuracies": 0.4625000059604645,
+      "rewards/chosen": 17.03064727783203,
+      "rewards/margins": 0.7148451805114746,
+      "rewards/rejected": 16.3158016204834,
+      "step": 50
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 5.110287666320801,
+      "learning_rate": 2.6548672566371687e-06,
+      "logits/chosen": -0.6322453022003174,
+      "logits/rejected": -0.6387485265731812,
+      "logps/chosen": -5.650620460510254,
+      "logps/rejected": -12.759811401367188,
+      "loss": 0.3835,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.101289749145508,
+      "rewards/margins": 1.1824612617492676,
+      "rewards/rejected": 15.918828964233398,
+      "step": 60
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "eval_logits/chosen": -0.5826543569564819,
+      "eval_logits/rejected": -0.5914276838302612,
+      "eval_logps/chosen": -3.5471787452697754,
+      "eval_logps/rejected": -16.51181983947754,
+      "eval_loss": 0.3286525011062622,
+      "eval_rewards/accuracies": 0.9280000925064087,
+      "eval_rewards/chosen": 17.148174285888672,
+      "eval_rewards/margins": 1.4386365413665771,
+      "eval_rewards/rejected": 15.709539413452148,
+      "eval_runtime": 372.0227,
+      "eval_samples_per_second": 2.688,
+      "eval_steps_per_second": 0.336,
+      "step": 60
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 5.098133563995361,
+      "learning_rate": 3.097345132743363e-06,
+      "logits/chosen": -0.5378152132034302,
+      "logits/rejected": -0.5494933724403381,
+      "logps/chosen": -1.5099802017211914,
+      "logps/rejected": -21.206321716308594,
+      "loss": 0.2931,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.083791732788086,
+      "rewards/margins": 1.5844331979751587,
+      "rewards/rejected": 15.499359130859375,
+      "step": 70
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 29.787437438964844,
+      "learning_rate": 3.539823008849558e-06,
+      "logits/chosen": -0.443774938583374,
+      "logits/rejected": -0.45571577548980713,
+      "logps/chosen": -1.5804342031478882,
+      "logps/rejected": -22.606929779052734,
+      "loss": 0.202,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.302125930786133,
+      "rewards/margins": 2.174014091491699,
+      "rewards/rejected": 15.128110885620117,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.14398193359375,
+      "learning_rate": 3.982300884955752e-06,
+      "logits/chosen": -0.3626072406768799,
+      "logits/rejected": -0.3787815570831299,
+      "logps/chosen": -2.203828811645508,
+      "logps/rejected": -29.433551788330078,
+      "loss": 0.2123,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.00284194946289,
+      "rewards/margins": 2.320391893386841,
+      "rewards/rejected": 14.682450294494629,
+      "step": 90
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 29.672739028930664,
+      "learning_rate": 4.424778761061948e-06,
+      "logits/chosen": -0.3035663962364197,
+      "logits/rejected": -0.31762221455574036,
+      "logps/chosen": -3.433589458465576,
+      "logps/rejected": -29.9322509765625,
+      "loss": 0.2592,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.929956436157227,
+      "rewards/margins": 2.31272029876709,
+      "rewards/rejected": 14.617237091064453,
+      "step": 100
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.873722791671753,
+      "learning_rate": 4.867256637168142e-06,
+      "logits/chosen": -0.2679600715637207,
+      "logits/rejected": -0.2826440930366516,
+      "logps/chosen": -0.9653514623641968,
+      "logps/rejected": -30.235322952270508,
+      "loss": 0.1336,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.462385177612305,
+      "rewards/margins": 3.1994175910949707,
+      "rewards/rejected": 14.26296615600586,
+      "step": 110
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.6913721561431885,
+      "learning_rate": 4.999409761242696e-06,
+      "logits/chosen": -0.22222033143043518,
+      "logits/rejected": -0.23720571398735046,
+      "logps/chosen": -4.4953508377075195,
+      "logps/rejected": -34.074745178222656,
+      "loss": 0.2552,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.04866600036621,
+      "rewards/margins": 3.1014418601989746,
+      "rewards/rejected": 13.947224617004395,
+      "step": 120
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "eval_logits/chosen": -0.206527978181839,
+      "eval_logits/rejected": -0.22178640961647034,
+      "eval_logps/chosen": -3.69442081451416,
+      "eval_logps/rejected": -36.072166442871094,
+      "eval_loss": 0.18996010720729828,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.133451461791992,
+      "eval_rewards/margins": 3.379946708679199,
+      "eval_rewards/rejected": 13.753504753112793,
+      "eval_runtime": 361.5279,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 120
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 61.80262756347656,
+      "learning_rate": 4.996519466816778e-06,
+      "logits/chosen": -0.18473535776138306,
+      "logits/rejected": -0.1988501250743866,
+      "logps/chosen": -3.7009687423706055,
+      "logps/rejected": -39.289939880371094,
+      "loss": 0.1394,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.106964111328125,
+      "rewards/margins": 3.633338212966919,
+      "rewards/rejected": 13.473625183105469,
+      "step": 130
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.6732702255249023,
+      "learning_rate": 4.9912234871722805e-06,
+      "logits/chosen": -0.16134041547775269,
+      "logits/rejected": -0.17547868192195892,
+      "logps/chosen": -3.0637736320495605,
+      "logps/rejected": -40.07548522949219,
+      "loss": 0.1408,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.392223358154297,
+      "rewards/margins": 4.242353439331055,
+      "rewards/rejected": 13.149867057800293,
+      "step": 140
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.346453994512558,
+      "learning_rate": 4.98352692559805e-06,
+      "logits/chosen": -0.13797929883003235,
+      "logits/rejected": -0.15283086895942688,
+      "logps/chosen": -5.14492130279541,
+      "logps/rejected": -47.97212219238281,
+      "loss": 0.2153,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.896778106689453,
+      "rewards/margins": 4.227695465087891,
+      "rewards/rejected": 12.669081687927246,
+      "step": 150
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.21871662139892578,
+      "learning_rate": 4.973437198621237e-06,
+      "logits/chosen": -0.12396670132875443,
+      "logits/rejected": -0.13780555129051208,
+      "logps/chosen": -6.108860015869141,
+      "logps/rejected": -54.90739440917969,
+      "loss": 0.0388,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 16.75935935974121,
+      "rewards/margins": 4.755282878875732,
+      "rewards/rejected": 12.004077911376953,
+      "step": 160
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 235.12429809570312,
+      "learning_rate": 4.960964028860621e-06,
+      "logits/chosen": -0.1140839159488678,
+      "logits/rejected": -0.1263057291507721,
+      "logps/chosen": -12.605452537536621,
+      "logps/rejected": -53.81230926513672,
+      "loss": 0.4651,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.101238250732422,
+      "rewards/margins": 3.9864249229431152,
+      "rewards/rejected": 12.114812850952148,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 190.97048950195312,
+      "learning_rate": 4.946119435657738e-06,
+      "logits/chosen": -0.10746976733207703,
+      "logits/rejected": -0.11878640949726105,
+      "logps/chosen": -8.5105562210083,
+      "logps/rejected": -51.314781188964844,
+      "loss": 0.2362,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.719980239868164,
+      "rewards/margins": 4.549674034118652,
+      "rewards/rejected": 12.170306205749512,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "eval_logits/chosen": -0.10870806127786636,
+      "eval_logits/rejected": -0.12223993986845016,
+      "eval_logps/chosen": -4.414996147155762,
+      "eval_logps/rejected": -53.885032653808594,
+      "eval_loss": 0.20236633718013763,
+      "eval_rewards/accuracies": 0.9510000944137573,
+      "eval_rewards/chosen": 17.06139373779297,
+      "eval_rewards/margins": 5.089176177978516,
+      "eval_rewards/rejected": 11.97221851348877,
+      "eval_runtime": 361.4355,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 180
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 56.81266784667969,
+      "learning_rate": 4.928917723494854e-06,
+      "logits/chosen": -0.10682469606399536,
+      "logits/rejected": -0.12124393880367279,
+      "logps/chosen": -3.058413028717041,
+      "logps/rejected": -55.052528381347656,
+      "loss": 0.2442,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.058589935302734,
+      "rewards/margins": 5.056097984313965,
+      "rewards/rejected": 12.002490043640137,
+      "step": 190
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 175.06552124023438,
+      "learning_rate": 4.909375468210947e-06,
+      "logits/chosen": -0.10520349442958832,
+      "logits/rejected": -0.12018950283527374,
+      "logps/chosen": -4.114959716796875,
+      "logps/rejected": -55.9394645690918,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.98603057861328,
+      "rewards/margins": 5.105838775634766,
+      "rewards/rejected": 11.880191802978516,
+      "step": 200
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 78.06558990478516,
+      "learning_rate": 4.8875115010289655e-06,
+      "logits/chosen": -0.10475558042526245,
+      "logits/rejected": -0.11949175596237183,
+      "logps/chosen": -6.760301113128662,
+      "logps/rejected": -53.91607666015625,
+      "loss": 0.2843,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.857545852661133,
+      "rewards/margins": 4.917357921600342,
+      "rewards/rejected": 11.94018840789795,
+      "step": 210
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 15.880486488342285,
+      "learning_rate": 4.863346890409768e-06,
+      "logits/chosen": -0.11213523149490356,
+      "logits/rejected": -0.12581588327884674,
+      "logps/chosen": -6.759585380554199,
+      "logps/rejected": -51.10936737060547,
+      "loss": 0.5104,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.859071731567383,
+      "rewards/margins": 4.638372898101807,
+      "rewards/rejected": 12.220698356628418,
+      "step": 220
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 46.97845458984375,
+      "learning_rate": 4.836904921750224e-06,
+      "logits/chosen": -0.11947059631347656,
+      "logits/rejected": -0.1329912692308426,
+      "logps/chosen": -3.608184814453125,
+      "logps/rejected": -48.794761657714844,
+      "loss": 0.2134,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.235904693603516,
+      "rewards/margins": 4.859888076782227,
+      "rewards/rejected": 12.376014709472656,
+      "step": 230
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 24.032859802246094,
+      "learning_rate": 4.808211074945042e-06,
+      "logits/chosen": -0.1200513243675232,
+      "logits/rejected": -0.1333036869764328,
+      "logps/chosen": -3.7552154064178467,
+      "logps/rejected": -49.87453079223633,
+      "loss": 0.1781,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.094650268554688,
+      "rewards/margins": 4.68077278137207,
+      "rewards/rejected": 12.41387939453125,
+      "step": 240
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "eval_logits/chosen": -0.12433278560638428,
+      "eval_logits/rejected": -0.13808581233024597,
+      "eval_logps/chosen": -4.408891201019287,
+      "eval_logps/rejected": -50.744781494140625,
+      "eval_loss": 0.1546352356672287,
+      "eval_rewards/accuracies": 0.9500000476837158,
+      "eval_rewards/chosen": 17.06200408935547,
+      "eval_rewards/margins": 4.775761604309082,
+      "eval_rewards/rejected": 12.286243438720703,
+      "eval_runtime": 361.4974,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 240
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.25737640261650085,
+      "learning_rate": 4.7772929998339485e-06,
+      "logits/chosen": -0.12348780035972595,
+      "logits/rejected": -0.13704943656921387,
+      "logps/chosen": -4.4299187660217285,
+      "logps/rejected": -53.074607849121094,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.087068557739258,
+      "rewards/margins": 5.06691837310791,
+      "rewards/rejected": 12.020149230957031,
+      "step": 250
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 0.1839389204978943,
+      "learning_rate": 4.744180489557859e-06,
+      "logits/chosen": -0.12177034467458725,
+      "logits/rejected": -0.1342695653438568,
+      "logps/chosen": -3.775188446044922,
+      "logps/rejected": -53.98720932006836,
+      "loss": 0.1896,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.12021255493164,
+      "rewards/margins": 5.148064613342285,
+      "rewards/rejected": 11.972146987915039,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.258485794067383,
+      "learning_rate": 4.708905451849754e-06,
+      "logits/chosen": -0.11067859083414078,
+      "logits/rejected": -0.12377731502056122,
+      "logps/chosen": -6.418317794799805,
+      "logps/rejected": -56.57402801513672,
+      "loss": 0.2315,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.738832473754883,
+      "rewards/margins": 4.884931564331055,
+      "rewards/rejected": 11.853900909423828,
+      "step": 270
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 77.56194305419922,
+      "learning_rate": 4.671501878287879e-06,
+      "logits/chosen": -0.1184445172548294,
+      "logits/rejected": -0.1339874565601349,
+      "logps/chosen": -10.12116527557373,
+      "logps/rejected": -53.403907775878906,
+      "loss": 0.5343,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 16.458633422851562,
+      "rewards/margins": 4.402472496032715,
+      "rewards/rejected": 12.056160926818848,
+      "step": 280
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 67.53883361816406,
+      "learning_rate": 4.6320058115409295e-06,
+      "logits/chosen": -0.1448262631893158,
+      "logits/rejected": -0.15793387591838837,
+      "logps/chosen": -3.4666190147399902,
+      "logps/rejected": -48.79213333129883,
+      "loss": 0.5017,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 16.945899963378906,
+      "rewards/margins": 4.2686333656311035,
+      "rewards/rejected": 12.677268028259277,
+      "step": 290
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.17521341145038605,
+      "learning_rate": 4.590455310636778e-06,
+      "logits/chosen": -0.16128253936767578,
+      "logits/rejected": -0.17375555634498596,
+      "logps/chosen": -2.9032950401306152,
+      "logps/rejected": -47.69734191894531,
+      "loss": 0.265,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.18383026123047,
+      "rewards/margins": 4.541309356689453,
+      "rewards/rejected": 12.642518997192383,
+      "step": 300
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "eval_logits/chosen": -0.17444846034049988,
+      "eval_logits/rejected": -0.18559777736663818,
+      "eval_logps/chosen": -2.535512924194336,
+      "eval_logps/rejected": -47.16367721557617,
+      "eval_loss": 0.15360687673091888,
+      "eval_rewards/accuracies": 0.9440000653266907,
+      "eval_rewards/chosen": 17.249343872070312,
+      "eval_rewards/margins": 4.604989051818848,
+      "eval_rewards/rejected": 12.644353866577148,
+      "eval_runtime": 361.4575,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 300
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 0.5040452480316162,
+      "learning_rate": 4.54689041428819e-06,
+      "logits/chosen": -0.16974106431007385,
+      "logits/rejected": -0.1810058057308197,
+      "logps/chosen": -1.233938217163086,
+      "logps/rejected": -49.907745361328125,
+      "loss": 0.1132,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.34117889404297,
+      "rewards/margins": 4.934173583984375,
+      "rewards/rejected": 12.407005310058594,
+      "step": 310
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 100.02949523925781,
+      "learning_rate": 4.501353102310901e-06,
+      "logits/chosen": -0.15705889463424683,
+      "logits/rejected": -0.1695334017276764,
+      "logps/chosen": -1.0820492506027222,
+      "logps/rejected": -52.577110290527344,
+      "loss": 0.1194,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.33388900756836,
+      "rewards/margins": 5.154760837554932,
+      "rewards/rejected": 12.179126739501953,
+      "step": 320
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 0.2689219117164612,
+      "learning_rate": 4.453887255171206e-06,
+      "logits/chosen": -0.13849371671676636,
+      "logits/rejected": -0.14990833401679993,
+      "logps/chosen": -1.8435032367706299,
+      "logps/rejected": -54.79044723510742,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.2423095703125,
+      "rewards/margins": 5.28987979888916,
+      "rewards/rejected": 11.952428817749023,
+      "step": 330
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 0.09305431693792343,
+      "learning_rate": 4.404538611702055e-06,
+      "logits/chosen": -0.12299702316522598,
+      "logits/rejected": -0.13453055918216705,
+      "logps/chosen": -2.9897143840789795,
+      "logps/rejected": -52.954498291015625,
+      "loss": 0.2873,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.17474365234375,
+      "rewards/margins": 5.071004867553711,
+      "rewards/rejected": 12.103739738464355,
+      "step": 340
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 59.282073974609375,
+      "learning_rate": 4.3533547250284015e-06,
+      "logits/chosen": -0.11913029849529266,
+      "logits/rejected": -0.12785324454307556,
+      "logps/chosen": -3.9456872940063477,
+      "logps/rejected": -48.68487548828125,
+      "loss": 0.4332,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 17.12805938720703,
+      "rewards/margins": 4.669450283050537,
+      "rewards/rejected": 12.458610534667969,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31101909279823303,
+      "learning_rate": 4.300384916744261e-06,
+      "logits/chosen": -0.11280188709497452,
+      "logits/rejected": -0.12300585210323334,
+      "logps/chosen": -2.1714723110198975,
+      "logps/rejected": -54.74174118041992,
+      "loss": 0.1605,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.326162338256836,
+      "rewards/margins": 5.467062473297119,
+      "rewards/rejected": 11.859098434448242,
+      "step": 360
+    },
+    {
+      "epoch": 0.32,
+      "eval_logits/chosen": -0.10620756447315216,
+      "eval_logits/rejected": -0.11727114766836166,
+      "eval_logps/chosen": -1.4165427684783936,
+      "eval_logps/rejected": -50.9525146484375,
+      "eval_loss": 0.3194349706172943,
+      "eval_rewards/accuracies": 0.9210000038146973,
+      "eval_rewards/chosen": 17.36124038696289,
+      "eval_rewards/margins": 5.095769882202148,
+      "eval_rewards/rejected": 12.26546859741211,
+      "eval_runtime": 361.5072,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 360
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 6.1126532554626465,
+      "learning_rate": 4.24568022938566e-06,
+      "logits/chosen": -0.10354311764240265,
+      "logits/rejected": -0.11526636779308319,
+      "logps/chosen": -1.2935255765914917,
+      "logps/rejected": -55.57566833496094,
+      "loss": 0.1711,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.439346313476562,
+      "rewards/margins": 5.700921058654785,
+      "rewards/rejected": 11.738424301147461,
+      "step": 370
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 34.15927505493164,
+      "learning_rate": 4.189293377245241e-06,
+      "logits/chosen": -0.1029932051897049,
+      "logits/rejected": -0.11382515728473663,
+      "logps/chosen": -2.5132687091827393,
+      "logps/rejected": -55.50346374511719,
+      "loss": 0.4359,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.731037139892578,
+      "rewards/margins": 4.368172645568848,
+      "rewards/rejected": 12.362865447998047,
+      "step": 380
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 2.8422904014587402,
+      "learning_rate": 4.131278695575952e-06,
+      "logits/chosen": -0.10793520510196686,
+      "logits/rejected": -0.12109285593032837,
+      "logps/chosen": -3.014652729034424,
+      "logps/rejected": -53.98411560058594,
+      "loss": 0.2161,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.137393951416016,
+      "rewards/margins": 5.105995178222656,
+      "rewards/rejected": 12.03139877319336,
+      "step": 390
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 54.0329475402832,
+      "learning_rate": 4.071692088232743e-06,
+      "logits/chosen": -0.10393750667572021,
+      "logits/rejected": -0.11834606528282166,
+      "logps/chosen": -2.1508543491363525,
+      "logps/rejected": -45.60733413696289,
+      "loss": 0.2077,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.586124420166016,
+      "rewards/margins": 5.077212333679199,
+      "rewards/rejected": 12.5089111328125,
+      "step": 400
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 81.61144256591797,
+      "learning_rate": 4.010590973802737e-06,
+      "logits/chosen": -0.09564584493637085,
+      "logits/rejected": -0.10617707669734955,
+      "logps/chosen": -3.4572842121124268,
+      "logps/rejected": -50.92162322998047,
+      "loss": 0.2478,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 17.010910034179688,
+      "rewards/margins": 4.556198596954346,
+      "rewards/rejected": 12.454713821411133,
+      "step": 410
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 0.30974289774894714,
+      "learning_rate": 3.948034230275781e-06,
+      "logits/chosen": -0.09134417027235031,
+      "logits/rejected": -0.1020016297698021,
+      "logps/chosen": -5.046698570251465,
+      "logps/rejected": -48.908958435058594,
+      "loss": 0.2894,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.007888793945312,
+      "rewards/margins": 4.53641414642334,
+      "rewards/rejected": 12.471475601196289,
+      "step": 420
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "eval_logits/chosen": -0.09054450690746307,
+      "eval_logits/rejected": -0.10264354199171066,
+      "eval_logps/chosen": -1.913105845451355,
+      "eval_logps/rejected": -51.11127471923828,
+      "eval_loss": 0.16789735853672028,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.311582565307617,
+      "eval_rewards/margins": 5.061989784240723,
+      "eval_rewards/rejected": 12.249593734741211,
+      "eval_runtime": 361.5337,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 420
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 12.824393272399902,
+      "learning_rate": 3.884082138308699e-06,
+      "logits/chosen": -0.08666776865720749,
+      "logits/rejected": -0.0997733399271965,
+      "logps/chosen": -1.7306327819824219,
+      "logps/rejected": -54.273292541503906,
+      "loss": 0.2298,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.167621612548828,
+      "rewards/margins": 5.065673351287842,
+      "rewards/rejected": 12.101947784423828,
+      "step": 430
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.30713599920272827,
+      "learning_rate": 3.818796323137896e-06,
+      "logits/chosen": -0.09174907952547073,
+      "logits/rejected": -0.10376611351966858,
+      "logps/chosen": -1.489154577255249,
+      "logps/rejected": -54.580726623535156,
+      "loss": 0.2513,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.22280502319336,
+      "rewards/margins": 5.175349235534668,
+      "rewards/rejected": 12.047454833984375,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 87.4791488647461,
+      "learning_rate": 3.7522396951963303e-06,
+      "logits/chosen": -0.09688778221607208,
+      "logits/rejected": -0.10897806286811829,
+      "logps/chosen": -3.157695770263672,
+      "logps/rejected": -50.96417236328125,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.345651626586914,
+      "rewards/margins": 5.245656967163086,
+      "rewards/rejected": 12.099993705749512,
+      "step": 450
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 146.2008056640625,
+      "learning_rate": 3.684476389492026e-06,
+      "logits/chosen": -0.09378582239151001,
+      "logits/rejected": -0.10475654900074005,
+      "logps/chosen": -0.5611928701400757,
+      "logps/rejected": -56.518890380859375,
+      "loss": 0.1981,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.113712310791016,
+      "rewards/margins": 5.068872928619385,
+      "rewards/rejected": 12.044839859008789,
+      "step": 460
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.9137721061706543,
+      "learning_rate": 3.6155717038065783e-06,
+      "logits/chosen": -0.08695463836193085,
+      "logits/rejected": -0.09596743434667587,
+      "logps/chosen": -1.5298550128936768,
+      "logps/rejected": -50.27445983886719,
+      "loss": 0.2066,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.35186004638672,
+      "rewards/margins": 5.014693260192871,
+      "rewards/rejected": 12.337167739868164,
+      "step": 470
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 84.80391693115234,
+      "learning_rate": 3.545592035773192e-06,
+      "logits/chosen": -0.0746893435716629,
+      "logits/rejected": -0.08653923869132996,
+      "logps/chosen": -2.0052125453948975,
+      "logps/rejected": -57.502811431884766,
+      "loss": 0.1149,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.14373016357422,
+      "rewards/margins": 5.360415935516357,
+      "rewards/rejected": 11.783313751220703,
+      "step": 480
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "eval_logits/chosen": -0.07700399309396744,
+      "eval_logits/rejected": -0.08828537166118622,
+      "eval_logps/chosen": -4.48896598815918,
+      "eval_logps/rejected": -53.76282501220703,
+      "eval_loss": 0.29511645436286926,
+      "eval_rewards/accuracies": 0.9230000376701355,
+      "eval_rewards/chosen": 17.053997039794922,
+      "eval_rewards/margins": 5.069558143615723,
+      "eval_rewards/rejected": 11.984437942504883,
+      "eval_runtime": 361.5035,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 480
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 82.9616470336914,
+      "learning_rate": 3.4746048188948806e-06,
+      "logits/chosen": -0.06675051152706146,
+      "logits/rejected": -0.07860895991325378,
+      "logps/chosen": -4.162237167358398,
+      "logps/rejected": -54.77789306640625,
+      "loss": 0.2979,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.047603607177734,
+      "rewards/margins": 5.138361930847168,
+      "rewards/rejected": 11.909242630004883,
+      "step": 490
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.04293210059404373,
+      "learning_rate": 3.4026784575644887e-06,
+      "logits/chosen": -0.06424491107463837,
+      "logits/rejected": -0.07567107677459717,
+      "logps/chosen": -2.05729603767395,
+      "logps/rejected": -56.646087646484375,
+      "loss": 0.4378,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.947803497314453,
+      "rewards/margins": 4.919981956481934,
+      "rewards/rejected": 12.02782154083252,
+      "step": 500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1125,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.372648598634496e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5286bf380083ef64133f25e161ddfb8549ff44790e432f478753f221ef89e695
+size 5368

eval_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": 0.0034948738757520914,
+    "eval_logits/rejected": -0.007565807551145554,
+    "eval_logps/chosen": -2.20627498626709,
+    "eval_logps/rejected": -60.60332489013672,
+    "eval_loss": 0.15350937843322754,
+    "eval_rewards/accuracies": 0.9610000848770142,
+    "eval_rewards/chosen": 17.28226661682129,
+    "eval_rewards/margins": 5.981877326965332,
+    "eval_rewards/rejected": 11.30038833618164,
+    "eval_runtime": 361.5652,
+    "eval_samples_per_second": 2.766,
+    "eval_steps_per_second": 0.346
+}

runs/Dec18_22-22-59_yadi/events.out.tfevents.1734560709.yadi.436386.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:697250b13dc063f122f3cccdacb9ae4ca8a13cbea14f3d9f6c0083f67e62277a
+size 95792

runs/Dec18_22-22-59_yadi/events.out.tfevents.1734574921.yadi.436386.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77f0ff4dbe534e3fef2d0e8a554d703d5b9903029ef4e11f5a8c6830e79a5bfe
+size 828

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0240ce510f08e6c2041724e9043e33be9d251d1e4a4d94eb68cd47b954b61d2
+size 17078292

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "total_flos": 1.4338459346927616e+18,
+    "train_loss": 0.23597783709896936,
+    "train_runtime": 13850.7044,
+    "train_samples_per_second": 0.65,
+    "train_steps_per_second": 0.081
+}

trainer_log.jsonl

@@ -0,0 +1,131 @@
+{"current_steps": 10, "total_steps": 1125, "loss": 1.1038, "accuracy": 0.5125000476837158, "lr": 4.4247787610619474e-07, "epoch": 0.008888888888888889, "percentage": 0.89, "elapsed_time": "0:01:11", "remaining_time": "2:12:28"}
+{"current_steps": 20, "total_steps": 1125, "loss": 1.1887, "accuracy": 0.4000000059604645, "lr": 8.849557522123895e-07, "epoch": 0.017777777777777778, "percentage": 1.78, "elapsed_time": "0:02:19", "remaining_time": "2:08:48"}
+{"current_steps": 30, "total_steps": 1125, "loss": 1.0563, "accuracy": 0.44999998807907104, "lr": 1.3274336283185843e-06, "epoch": 0.02666666666666667, "percentage": 2.67, "elapsed_time": "0:03:27", "remaining_time": "2:06:15"}
+{"current_steps": 40, "total_steps": 1125, "loss": 0.7596, "accuracy": 0.5, "lr": 1.769911504424779e-06, "epoch": 0.035555555555555556, "percentage": 3.56, "elapsed_time": "0:04:36", "remaining_time": "2:04:51"}
+{"current_steps": 50, "total_steps": 1125, "loss": 0.6083, "accuracy": 0.4625000059604645, "lr": 2.212389380530974e-06, "epoch": 0.044444444444444446, "percentage": 4.44, "elapsed_time": "0:05:44", "remaining_time": "2:03:19"}
+{"current_steps": 60, "total_steps": 1125, "loss": 0.3835, "accuracy": 0.9125000238418579, "lr": 2.6548672566371687e-06, "epoch": 0.05333333333333334, "percentage": 5.33, "elapsed_time": "0:06:51", "remaining_time": "2:01:46"}
+{"current_steps": 60, "total_steps": 1125, "eval_loss": 0.3286525011062622, "epoch": 0.05333333333333334, "percentage": 5.33, "elapsed_time": "0:13:03", "remaining_time": "3:51:50"}
+{"current_steps": 70, "total_steps": 1125, "loss": 0.2931, "accuracy": 0.9375, "lr": 3.097345132743363e-06, "epoch": 0.06222222222222222, "percentage": 6.22, "elapsed_time": "0:14:08", "remaining_time": "3:33:12"}
+{"current_steps": 80, "total_steps": 1125, "loss": 0.202, "accuracy": 0.9750000238418579, "lr": 3.539823008849558e-06, "epoch": 0.07111111111111111, "percentage": 7.11, "elapsed_time": "0:15:13", "remaining_time": "3:18:57"}
+{"current_steps": 90, "total_steps": 1125, "loss": 0.2123, "accuracy": 0.925000011920929, "lr": 3.982300884955752e-06, "epoch": 0.08, "percentage": 8.0, "elapsed_time": "0:16:19", "remaining_time": "3:07:39"}
+{"current_steps": 100, "total_steps": 1125, "loss": 0.2592, "accuracy": 0.9375, "lr": 4.424778761061948e-06, "epoch": 0.08888888888888889, "percentage": 8.89, "elapsed_time": "0:17:24", "remaining_time": "2:58:22"}
+{"current_steps": 110, "total_steps": 1125, "loss": 0.1336, "accuracy": 0.949999988079071, "lr": 4.867256637168142e-06, "epoch": 0.09777777777777778, "percentage": 9.78, "elapsed_time": "0:18:29", "remaining_time": "2:50:35"}
+{"current_steps": 120, "total_steps": 1125, "loss": 0.2552, "accuracy": 0.8999999761581421, "lr": 4.999409761242696e-06, "epoch": 0.10666666666666667, "percentage": 10.67, "elapsed_time": "0:19:34", "remaining_time": "2:43:55"}
+{"current_steps": 120, "total_steps": 1125, "eval_loss": 0.18996010720729828, "epoch": 0.10666666666666667, "percentage": 10.67, "elapsed_time": "0:25:35", "remaining_time": "3:34:23"}
+{"current_steps": 130, "total_steps": 1125, "loss": 0.1394, "accuracy": 0.9624999761581421, "lr": 4.996519466816778e-06, "epoch": 0.11555555555555555, "percentage": 11.56, "elapsed_time": "0:26:41", "remaining_time": "3:24:16"}
+{"current_steps": 140, "total_steps": 1125, "loss": 0.1408, "accuracy": 0.9750000238418579, "lr": 4.9912234871722805e-06, "epoch": 0.12444444444444444, "percentage": 12.44, "elapsed_time": "0:27:46", "remaining_time": "3:15:25"}
+{"current_steps": 150, "total_steps": 1125, "loss": 0.2153, "accuracy": 0.9375, "lr": 4.98352692559805e-06, "epoch": 0.13333333333333333, "percentage": 13.33, "elapsed_time": "0:28:51", "remaining_time": "3:07:35"}
+{"current_steps": 160, "total_steps": 1125, "loss": 0.0388, "accuracy": 0.9750000238418579, "lr": 4.973437198621237e-06, "epoch": 0.14222222222222222, "percentage": 14.22, "elapsed_time": "0:29:56", "remaining_time": "3:00:36"}
+{"current_steps": 170, "total_steps": 1125, "loss": 0.4651, "accuracy": 0.875, "lr": 4.960964028860621e-06, "epoch": 0.1511111111111111, "percentage": 15.11, "elapsed_time": "0:31:01", "remaining_time": "2:54:18"}
+{"current_steps": 180, "total_steps": 1125, "loss": 0.2362, "accuracy": 0.925000011920929, "lr": 4.946119435657738e-06, "epoch": 0.16, "percentage": 16.0, "elapsed_time": "0:32:06", "remaining_time": "2:48:34"}
+{"current_steps": 180, "total_steps": 1125, "eval_loss": 0.20236633718013763, "epoch": 0.16, "percentage": 16.0, "elapsed_time": "0:38:08", "remaining_time": "3:20:12"}
+{"current_steps": 190, "total_steps": 1125, "loss": 0.2442, "accuracy": 0.9500000476837158, "lr": 4.928917723494854e-06, "epoch": 0.1688888888888889, "percentage": 16.89, "elapsed_time": "0:39:13", "remaining_time": "3:12:59"}
+{"current_steps": 200, "total_steps": 1125, "loss": 0.1915, "accuracy": 0.9500000476837158, "lr": 4.909375468210947e-06, "epoch": 0.17777777777777778, "percentage": 17.78, "elapsed_time": "0:40:18", "remaining_time": "3:06:24"}
+{"current_steps": 210, "total_steps": 1125, "loss": 0.2843, "accuracy": 0.9375, "lr": 4.8875115010289655e-06, "epoch": 0.18666666666666668, "percentage": 18.67, "elapsed_time": "0:41:23", "remaining_time": "3:00:19"}
+{"current_steps": 220, "total_steps": 1125, "loss": 0.5104, "accuracy": 0.875, "lr": 4.863346890409768e-06, "epoch": 0.19555555555555557, "percentage": 19.56, "elapsed_time": "0:42:28", "remaining_time": "2:54:41"}
+{"current_steps": 230, "total_steps": 1125, "loss": 0.2134, "accuracy": 0.925000011920929, "lr": 4.836904921750224e-06, "epoch": 0.20444444444444446, "percentage": 20.44, "elapsed_time": "0:43:33", "remaining_time": "2:49:28"}
+{"current_steps": 240, "total_steps": 1125, "loss": 0.1781, "accuracy": 0.9500000476837158, "lr": 4.808211074945042e-06, "epoch": 0.21333333333333335, "percentage": 21.33, "elapsed_time": "0:44:38", "remaining_time": "2:44:35"}
+{"current_steps": 240, "total_steps": 1125, "eval_loss": 0.1546352356672287, "epoch": 0.21333333333333335, "percentage": 21.33, "elapsed_time": "0:50:39", "remaining_time": "3:06:48"}
+{"current_steps": 250, "total_steps": 1125, "loss": 0.1373, "accuracy": 0.9375, "lr": 4.7772929998339485e-06, "epoch": 0.2222222222222222, "percentage": 22.22, "elapsed_time": "0:51:44", "remaining_time": "3:01:05"}
+{"current_steps": 260, "total_steps": 1125, "loss": 0.1896, "accuracy": 0.949999988079071, "lr": 4.744180489557859e-06, "epoch": 0.2311111111111111, "percentage": 23.11, "elapsed_time": "0:52:49", "remaining_time": "2:55:44"}
+{"current_steps": 270, "total_steps": 1125, "loss": 0.2315, "accuracy": 0.9375, "lr": 4.708905451849754e-06, "epoch": 0.24, "percentage": 24.0, "elapsed_time": "0:53:54", "remaining_time": "2:50:42"}
+{"current_steps": 280, "total_steps": 1125, "loss": 0.5343, "accuracy": 0.862500011920929, "lr": 4.671501878287879e-06, "epoch": 0.24888888888888888, "percentage": 24.89, "elapsed_time": "0:54:59", "remaining_time": "2:45:56"}
+{"current_steps": 290, "total_steps": 1125, "loss": 0.5017, "accuracy": 0.887499988079071, "lr": 4.6320058115409295e-06, "epoch": 0.2577777777777778, "percentage": 25.78, "elapsed_time": "0:56:04", "remaining_time": "2:41:26"}
+{"current_steps": 300, "total_steps": 1125, "loss": 0.265, "accuracy": 0.925000011920929, "lr": 4.590455310636778e-06, "epoch": 0.26666666666666666, "percentage": 26.67, "elapsed_time": "0:57:09", "remaining_time": "2:37:09"}
+{"current_steps": 300, "total_steps": 1125, "eval_loss": 0.15360687673091888, "epoch": 0.26666666666666666, "percentage": 26.67, "elapsed_time": "1:03:10", "remaining_time": "2:53:44"}
+{"current_steps": 310, "total_steps": 1125, "loss": 0.1132, "accuracy": 0.9500000476837158, "lr": 4.54689041428819e-06, "epoch": 0.27555555555555555, "percentage": 27.56, "elapsed_time": "1:04:15", "remaining_time": "2:48:56"}
+{"current_steps": 320, "total_steps": 1125, "loss": 0.1194, "accuracy": 0.9500000476837158, "lr": 4.501353102310901e-06, "epoch": 0.28444444444444444, "percentage": 28.44, "elapsed_time": "1:05:21", "remaining_time": "2:44:25"}
+{"current_steps": 330, "total_steps": 1125, "loss": 0.0926, "accuracy": 0.9500000476837158, "lr": 4.453887255171206e-06, "epoch": 0.29333333333333333, "percentage": 29.33, "elapsed_time": "1:06:26", "remaining_time": "2:40:04"}
+{"current_steps": 340, "total_steps": 1125, "loss": 0.2873, "accuracy": 0.925000011920929, "lr": 4.404538611702055e-06, "epoch": 0.3022222222222222, "percentage": 30.22, "elapsed_time": "1:07:31", "remaining_time": "2:35:54"}
+{"current_steps": 350, "total_steps": 1125, "loss": 0.4332, "accuracy": 0.875, "lr": 4.3533547250284015e-06, "epoch": 0.3111111111111111, "percentage": 31.11, "elapsed_time": "1:08:36", "remaining_time": "2:31:54"}
+{"current_steps": 360, "total_steps": 1125, "loss": 0.1605, "accuracy": 0.9500000476837158, "lr": 4.300384916744261e-06, "epoch": 0.32, "percentage": 32.0, "elapsed_time": "1:09:41", "remaining_time": "2:28:05"}
+{"current_steps": 360, "total_steps": 1125, "eval_loss": 0.3194349706172943, "epoch": 0.32, "percentage": 32.0, "elapsed_time": "1:15:42", "remaining_time": "2:40:53"}
+{"current_steps": 370, "total_steps": 1125, "loss": 0.1711, "accuracy": 0.9500000476837158, "lr": 4.24568022938566e-06, "epoch": 0.3288888888888889, "percentage": 32.89, "elapsed_time": "1:16:47", "remaining_time": "2:36:42"}
+{"current_steps": 380, "total_steps": 1125, "loss": 0.4359, "accuracy": 0.8875000476837158, "lr": 4.189293377245241e-06, "epoch": 0.3377777777777778, "percentage": 33.78, "elapsed_time": "1:17:52", "remaining_time": "2:32:40"}
+{"current_steps": 390, "total_steps": 1125, "loss": 0.2161, "accuracy": 0.949999988079071, "lr": 4.131278695575952e-06, "epoch": 0.3466666666666667, "percentage": 34.67, "elapsed_time": "1:18:57", "remaining_time": "2:28:48"}
+{"current_steps": 400, "total_steps": 1125, "loss": 0.2077, "accuracy": 0.925000011920929, "lr": 4.071692088232743e-06, "epoch": 0.35555555555555557, "percentage": 35.56, "elapsed_time": "1:20:02", "remaining_time": "2:25:04"}
+{"current_steps": 410, "total_steps": 1125, "loss": 0.2478, "accuracy": 0.8875000476837158, "lr": 4.010590973802737e-06, "epoch": 0.36444444444444446, "percentage": 36.44, "elapsed_time": "1:21:07", "remaining_time": "2:21:27"}
+{"current_steps": 420, "total_steps": 1125, "loss": 0.2894, "accuracy": 0.8999999761581421, "lr": 3.948034230275781e-06, "epoch": 0.37333333333333335, "percentage": 37.33, "elapsed_time": "1:22:12", "remaining_time": "2:17:58"}
+{"current_steps": 420, "total_steps": 1125, "eval_loss": 0.16789735853672028, "epoch": 0.37333333333333335, "percentage": 37.33, "elapsed_time": "1:28:13", "remaining_time": "2:28:05"}
+{"current_steps": 430, "total_steps": 1125, "loss": 0.2298, "accuracy": 0.9500000476837158, "lr": 3.884082138308699e-06, "epoch": 0.38222222222222224, "percentage": 38.22, "elapsed_time": "1:29:18", "remaining_time": "2:24:20"}
+{"current_steps": 440, "total_steps": 1125, "loss": 0.2513, "accuracy": 0.9375, "lr": 3.818796323137896e-06, "epoch": 0.39111111111111113, "percentage": 39.11, "elapsed_time": "1:30:23", "remaining_time": "2:20:43"}
+{"current_steps": 450, "total_steps": 1125, "loss": 0.1758, "accuracy": 0.9500000476837158, "lr": 3.7522396951963303e-06, "epoch": 0.4, "percentage": 40.0, "elapsed_time": "1:31:28", "remaining_time": "2:17:12"}
+{"current_steps": 460, "total_steps": 1125, "loss": 0.1981, "accuracy": 0.9500000476837158, "lr": 3.684476389492026e-06, "epoch": 0.4088888888888889, "percentage": 40.89, "elapsed_time": "1:32:33", "remaining_time": "2:13:48"}
+{"current_steps": 470, "total_steps": 1125, "loss": 0.2066, "accuracy": 0.9375, "lr": 3.6155717038065783e-06, "epoch": 0.4177777777777778, "percentage": 41.78, "elapsed_time": "1:33:38", "remaining_time": "2:10:29"}
+{"current_steps": 480, "total_steps": 1125, "loss": 0.1149, "accuracy": 0.9500000476837158, "lr": 3.545592035773192e-06, "epoch": 0.4266666666666667, "percentage": 42.67, "elapsed_time": "1:34:43", "remaining_time": "2:07:16"}
+{"current_steps": 480, "total_steps": 1125, "eval_loss": 0.29511645436286926, "epoch": 0.4266666666666667, "percentage": 42.67, "elapsed_time": "1:40:44", "remaining_time": "2:15:22"}
+{"current_steps": 490, "total_steps": 1125, "loss": 0.2979, "accuracy": 0.9125000238418579, "lr": 3.4746048188948806e-06, "epoch": 0.43555555555555553, "percentage": 43.56, "elapsed_time": "1:41:49", "remaining_time": "2:11:57"}
+{"current_steps": 500, "total_steps": 1125, "loss": 0.4378, "accuracy": 0.8875000476837158, "lr": 3.4026784575644887e-06, "epoch": 0.4444444444444444, "percentage": 44.44, "elapsed_time": "1:42:54", "remaining_time": "2:08:37"}
+{"current_steps": 510, "total_steps": 1125, "loss": 0.2613, "accuracy": 0.9375, "lr": 3.329882261149148e-06, "epoch": 0.4533333333333333, "percentage": 45.33, "elapsed_time": "1:44:00", "remaining_time": "2:05:25"}
+{"current_steps": 520, "total_steps": 1125, "loss": 0.1517, "accuracy": 0.9375, "lr": 3.25628637720269e-06, "epoch": 0.4622222222222222, "percentage": 46.22, "elapsed_time": "1:45:05", "remaining_time": "2:02:16"}
+{"current_steps": 530, "total_steps": 1125, "loss": 0.2304, "accuracy": 0.9375, "lr": 3.181961723870359e-06, "epoch": 0.4711111111111111, "percentage": 47.11, "elapsed_time": "1:46:10", "remaining_time": "1:59:11"}
+{"current_steps": 540, "total_steps": 1125, "loss": 0.0384, "accuracy": 0.987500011920929, "lr": 3.1069799215509847e-06, "epoch": 0.48, "percentage": 48.0, "elapsed_time": "1:47:15", "remaining_time": "1:56:11"}
+{"current_steps": 540, "total_steps": 1125, "eval_loss": 0.17387841641902924, "epoch": 0.48, "percentage": 48.0, "elapsed_time": "1:53:16", "remaining_time": "2:02:43"}
+{"current_steps": 550, "total_steps": 1125, "loss": 0.2087, "accuracy": 0.9125000238418579, "lr": 3.0314132238824416e-06, "epoch": 0.4888888888888889, "percentage": 48.89, "elapsed_time": "1:54:21", "remaining_time": "1:59:33"}
+{"current_steps": 560, "total_steps": 1125, "loss": 0.0969, "accuracy": 0.9624999761581421, "lr": 2.955334448116915e-06, "epoch": 0.49777777777777776, "percentage": 49.78, "elapsed_time": "1:55:26", "remaining_time": "1:56:28"}
+{"current_steps": 570, "total_steps": 1125, "loss": 0.1447, "accuracy": 0.949999988079071, "lr": 2.8788169049530533e-06, "epoch": 0.5066666666666667, "percentage": 50.67, "elapsed_time": "1:56:31", "remaining_time": "1:53:27"}
+{"current_steps": 580, "total_steps": 1125, "loss": 0.0998, "accuracy": 0.949999988079071, "lr": 2.8019343278926397e-06, "epoch": 0.5155555555555555, "percentage": 51.56, "elapsed_time": "1:57:36", "remaining_time": "1:50:30"}
+{"current_steps": 590, "total_steps": 1125, "loss": 0.1162, "accuracy": 0.9750000238418579, "lr": 2.7247608021898265e-06, "epoch": 0.5244444444444445, "percentage": 52.44, "elapsed_time": "1:58:41", "remaining_time": "1:47:37"}
+{"current_steps": 600, "total_steps": 1125, "loss": 0.4008, "accuracy": 0.8875000476837158, "lr": 2.647370693461432e-06, "epoch": 0.5333333333333333, "percentage": 53.33, "elapsed_time": "1:59:46", "remaining_time": "1:44:47"}
+{"current_steps": 600, "total_steps": 1125, "eval_loss": 0.17061151564121246, "epoch": 0.5333333333333333, "percentage": 53.33, "elapsed_time": "2:05:47", "remaining_time": "1:50:04"}
+{"current_steps": 610, "total_steps": 1125, "loss": 0.1265, "accuracy": 0.9624999761581421, "lr": 2.569838576027068e-06, "epoch": 0.5422222222222223, "percentage": 54.22, "elapsed_time": "2:06:52", "remaining_time": "1:47:06"}
+{"current_steps": 620, "total_steps": 1125, "loss": 0.097, "accuracy": 0.9624999761581421, "lr": 2.4922391610481544e-06, "epoch": 0.5511111111111111, "percentage": 55.11, "elapsed_time": "2:07:57", "remaining_time": "1:44:13"}
+{"current_steps": 630, "total_steps": 1125, "loss": 0.3115, "accuracy": 0.925000011920929, "lr": 2.4146472245350804e-06, "epoch": 0.56, "percentage": 56.0, "elapsed_time": "2:09:02", "remaining_time": "1:41:23"}
+{"current_steps": 640, "total_steps": 1125, "loss": 0.1268, "accuracy": 0.9500000476837158, "lr": 2.337137535291868e-06, "epoch": 0.5688888888888889, "percentage": 56.89, "elapsed_time": "2:10:07", "remaining_time": "1:38:36"}
+{"current_steps": 650, "total_steps": 1125, "loss": 0.1831, "accuracy": 0.9500000476837158, "lr": 2.259784782867782e-06, "epoch": 0.5777777777777777, "percentage": 57.78, "elapsed_time": "2:11:12", "remaining_time": "1:35:52"}
+{"current_steps": 660, "total_steps": 1125, "loss": 0.1678, "accuracy": 0.9624999761581421, "lr": 2.182663505585314e-06, "epoch": 0.5866666666666667, "percentage": 58.67, "elapsed_time": "2:12:17", "remaining_time": "1:33:11"}
+{"current_steps": 660, "total_steps": 1125, "eval_loss": 0.2050127536058426, "epoch": 0.5866666666666667, "percentage": 58.67, "elapsed_time": "2:18:18", "remaining_time": "1:37:26"}
+{"current_steps": 670, "total_steps": 1125, "loss": 0.1025, "accuracy": 0.9624999761581421, "lr": 2.1058480187138863e-06, "epoch": 0.5955555555555555, "percentage": 59.56, "elapsed_time": "2:19:23", "remaining_time": "1:34:39"}
+{"current_steps": 680, "total_steps": 1125, "loss": 0.1961, "accuracy": 0.9500000476837158, "lr": 2.0294123428584985e-06, "epoch": 0.6044444444444445, "percentage": 60.44, "elapsed_time": "2:20:28", "remaining_time": "1:31:55"}
+{"current_steps": 690, "total_steps": 1125, "loss": 0.2821, "accuracy": 0.9375, "lr": 1.953430132632311e-06, "epoch": 0.6133333333333333, "percentage": 61.33, "elapsed_time": "2:21:33", "remaining_time": "1:29:14"}
+{"current_steps": 700, "total_steps": 1125, "loss": 0.3141, "accuracy": 0.9375, "lr": 1.8779746056819104e-06, "epoch": 0.6222222222222222, "percentage": 62.22, "elapsed_time": "2:22:38", "remaining_time": "1:26:36"}
+{"current_steps": 710, "total_steps": 1125, "loss": 0.3827, "accuracy": 0.887499988079071, "lr": 1.8031184721336364e-06, "epoch": 0.6311111111111111, "percentage": 63.11, "elapsed_time": "2:23:43", "remaining_time": "1:24:00"}
+{"current_steps": 720, "total_steps": 1125, "loss": 0.2272, "accuracy": 0.949999988079071, "lr": 1.7289338645289711e-06, "epoch": 0.64, "percentage": 64.0, "elapsed_time": "2:24:48", "remaining_time": "1:21:27"}
+{"current_steps": 720, "total_steps": 1125, "eval_loss": 0.14018221199512482, "epoch": 0.64, "percentage": 64.0, "elapsed_time": "2:30:49", "remaining_time": "1:24:50"}
+{"current_steps": 730, "total_steps": 1125, "loss": 0.1171, "accuracy": 0.9624999761581421, "lr": 1.6554922683164875e-06, "epoch": 0.6488888888888888, "percentage": 64.89, "elapsed_time": "2:31:54", "remaining_time": "1:22:11"}
+{"current_steps": 740, "total_steps": 1125, "loss": 0.2001, "accuracy": 0.9624999761581421, "lr": 1.5828644529673592e-06, "epoch": 0.6577777777777778, "percentage": 65.78, "elapsed_time": "2:32:59", "remaining_time": "1:19:35"}
+{"current_steps": 750, "total_steps": 1125, "loss": 0.1692, "accuracy": 0.9375, "lr": 1.5111204037807844e-06, "epoch": 0.6666666666666666, "percentage": 66.67, "elapsed_time": "2:34:04", "remaining_time": "1:17:02"}
+{"current_steps": 760, "total_steps": 1125, "loss": 0.2418, "accuracy": 0.949999988079071, "lr": 1.4403292544450625e-06, "epoch": 0.6755555555555556, "percentage": 67.56, "elapsed_time": "2:35:09", "remaining_time": "1:14:30"}
+{"current_steps": 770, "total_steps": 1125, "loss": 0.1749, "accuracy": 0.9500000476837158, "lr": 1.3705592204192853e-06, "epoch": 0.6844444444444444, "percentage": 68.44, "elapsed_time": "2:36:14", "remaining_time": "1:12:01"}
+{"current_steps": 780, "total_steps": 1125, "loss": 0.1915, "accuracy": 0.925000011920929, "lr": 1.301877533199859e-06, "epoch": 0.6933333333333334, "percentage": 69.33, "elapsed_time": "2:37:19", "remaining_time": "1:09:34"}
+{"current_steps": 780, "total_steps": 1125, "eval_loss": 0.24412688612937927, "epoch": 0.6933333333333334, "percentage": 69.33, "elapsed_time": "2:43:20", "remaining_time": "1:12:14"}
+{"current_steps": 790, "total_steps": 1125, "loss": 0.2143, "accuracy": 0.9375, "lr": 1.2343503755351729e-06, "epoch": 0.7022222222222222, "percentage": 70.22, "elapsed_time": "2:44:25", "remaining_time": "1:09:43"}
+{"current_steps": 800, "total_steps": 1125, "loss": 0.0792, "accuracy": 0.9750000238418579, "lr": 1.168042817650881e-06, "epoch": 0.7111111111111111, "percentage": 71.11, "elapsed_time": "2:45:30", "remaining_time": "1:07:14"}
+{"current_steps": 810, "total_steps": 1125, "loss": 0.3177, "accuracy": 0.925000011920929, "lr": 1.1030187545472012e-06, "epoch": 0.72, "percentage": 72.0, "elapsed_time": "2:46:35", "remaining_time": "1:04:47"}
+{"current_steps": 820, "total_steps": 1125, "loss": 0.2128, "accuracy": 0.949999988079071, "lr": 1.0393408444287048e-06, "epoch": 0.7288888888888889, "percentage": 72.89, "elapsed_time": "2:47:40", "remaining_time": "1:02:21"}
+{"current_steps": 830, "total_steps": 1125, "loss": 0.1296, "accuracy": 0.9750000238418579, "lr": 9.770704483258782e-07, "epoch": 0.7377777777777778, "percentage": 73.78, "elapsed_time": "2:48:44", "remaining_time": "0:59:58"}
+{"current_steps": 840, "total_steps": 1125, "loss": 0.0635, "accuracy": 0.987500011920929, "lr": 9.162675709666865e-07, "epoch": 0.7466666666666667, "percentage": 74.67, "elapsed_time": "2:49:49", "remaining_time": "0:57:37"}
+{"current_steps": 840, "total_steps": 1125, "eval_loss": 0.1689341962337494, "epoch": 0.7466666666666667, "percentage": 74.67, "elapsed_time": "2:55:51", "remaining_time": "0:59:39"}
+{"current_steps": 850, "total_steps": 1125, "loss": 0.1157, "accuracy": 0.9750000238418579, "lr": 8.569908029550686e-07, "epoch": 0.7555555555555555, "percentage": 75.56, "elapsed_time": "2:56:56", "remaining_time": "0:57:14"}
+{"current_steps": 860, "total_steps": 1125, "loss": 0.2252, "accuracy": 0.9375, "lr": 7.992972643121227e-07, "epoch": 0.7644444444444445, "percentage": 76.44, "elapsed_time": "2:58:01", "remaining_time": "0:54:51"}
+{"current_steps": 870, "total_steps": 1125, "loss": 0.1398, "accuracy": 0.9624999761581421, "lr": 7.432425494343509e-07, "epoch": 0.7733333333333333, "percentage": 77.33, "elapsed_time": "2:59:06", "remaining_time": "0:52:29"}
+{"current_steps": 880, "total_steps": 1125, "loss": 0.2966, "accuracy": 0.9375, "lr": 6.888806735220396e-07, "epoch": 0.7822222222222223, "percentage": 78.22, "elapsed_time": "3:00:11", "remaining_time": "0:50:09"}
+{"current_steps": 890, "total_steps": 1125, "loss": 0.1641, "accuracy": 0.9375, "lr": 6.362640205293583e-07, "epoch": 0.7911111111111111, "percentage": 79.11, "elapsed_time": "3:01:15", "remaining_time": "0:47:51"}
+{"current_steps": 900, "total_steps": 1125, "loss": 0.1703, "accuracy": 0.9624999761581421, "lr": 5.854432926863684e-07, "epoch": 0.8, "percentage": 80.0, "elapsed_time": "3:02:20", "remaining_time": "0:45:35"}
+{"current_steps": 900, "total_steps": 1125, "eval_loss": 0.13995186984539032, "epoch": 0.8, "percentage": 80.0, "elapsed_time": "3:08:22", "remaining_time": "0:47:05"}
+{"current_steps": 910, "total_steps": 1125, "loss": 0.0507, "accuracy": 0.9750000238418579, "lr": 5.364674616415547e-07, "epoch": 0.8088888888888889, "percentage": 80.89, "elapsed_time": "3:09:27", "remaining_time": "0:44:45"}
+{"current_steps": 920, "total_steps": 1125, "loss": 0.0881, "accuracy": 0.9500000476837158, "lr": 4.893837212719859e-07, "epoch": 0.8177777777777778, "percentage": 81.78, "elapsed_time": "3:10:31", "remaining_time": "0:42:27"}
+{"current_steps": 930, "total_steps": 1125, "loss": 0.1373, "accuracy": 0.949999988079071, "lr": 4.442374422065493e-07, "epoch": 0.8266666666666667, "percentage": 82.67, "elapsed_time": "3:11:36", "remaining_time": "0:40:10"}
+{"current_steps": 940, "total_steps": 1125, "loss": 0.2763, "accuracy": 0.925000011920929, "lr": 4.0107212810610974e-07, "epoch": 0.8355555555555556, "percentage": 83.56, "elapsed_time": "3:12:41", "remaining_time": "0:37:55"}
+{"current_steps": 950, "total_steps": 1125, "loss": 0.1758, "accuracy": 0.949999988079071, "lr": 3.599293737426932e-07, "epoch": 0.8444444444444444, "percentage": 84.44, "elapsed_time": "3:13:46", "remaining_time": "0:35:41"}
+{"current_steps": 960, "total_steps": 1125, "loss": 0.1138, "accuracy": 0.9624999761581421, "lr": 3.208488249181216e-07, "epoch": 0.8533333333333334, "percentage": 85.33, "elapsed_time": "3:14:51", "remaining_time": "0:33:29"}
+{"current_steps": 960, "total_steps": 1125, "eval_loss": 0.1441129744052887, "epoch": 0.8533333333333334, "percentage": 85.33, "elapsed_time": "3:20:53", "remaining_time": "0:34:31"}
+{"current_steps": 970, "total_steps": 1125, "loss": 0.2425, "accuracy": 0.925000011920929, "lr": 2.838681402606952e-07, "epoch": 0.8622222222222222, "percentage": 86.22, "elapsed_time": "3:21:57", "remaining_time": "0:32:16"}
+{"current_steps": 980, "total_steps": 1125, "loss": 0.0609, "accuracy": 0.9750000238418579, "lr": 2.490229549367443e-07, "epoch": 0.8711111111111111, "percentage": 87.11, "elapsed_time": "3:23:02", "remaining_time": "0:30:02"}
+{"current_steps": 990, "total_steps": 1125, "loss": 0.3464, "accuracy": 0.8999999761581421, "lr": 2.1634684631203412e-07, "epoch": 0.88, "percentage": 88.0, "elapsed_time": "3:24:07", "remaining_time": "0:27:50"}
+{"current_steps": 1000, "total_steps": 1125, "loss": 0.0063, "accuracy": 1.0, "lr": 1.8587130159608196e-07, "epoch": 0.8888888888888888, "percentage": 88.89, "elapsed_time": "3:25:12", "remaining_time": "0:25:39"}
+{"current_steps": 1010, "total_steps": 1125, "loss": 0.2741, "accuracy": 0.925000011920929, "lr": 1.5762568750059604e-07, "epoch": 0.8977777777777778, "percentage": 89.78, "elapsed_time": "3:26:18", "remaining_time": "0:23:29"}
+{"current_steps": 1020, "total_steps": 1125, "loss": 0.0513, "accuracy": 0.9750000238418579, "lr": 1.316372219412454e-07, "epoch": 0.9066666666666666, "percentage": 90.67, "elapsed_time": "3:27:23", "remaining_time": "0:21:20"}
+{"current_steps": 1020, "total_steps": 1125, "eval_loss": 0.14118175208568573, "epoch": 0.9066666666666666, "percentage": 90.67, "elapsed_time": "3:33:25", "remaining_time": "0:21:58"}
+{"current_steps": 1030, "total_steps": 1125, "loss": 0.0349, "accuracy": 0.987500011920929, "lr": 1.0793094781005792e-07, "epoch": 0.9155555555555556, "percentage": 91.56, "elapsed_time": "3:34:29", "remaining_time": "0:19:47"}
+{"current_steps": 1040, "total_steps": 1125, "loss": 0.1927, "accuracy": 0.9500000476837158, "lr": 8.652970884369255e-08, "epoch": 0.9244444444444444, "percentage": 92.44, "elapsed_time": "3:35:34", "remaining_time": "0:17:37"}
+{"current_steps": 1050, "total_steps": 1125, "loss": 0.1317, "accuracy": 0.9750000238418579, "lr": 6.745412761086007e-08, "epoch": 0.9333333333333333, "percentage": 93.33, "elapsed_time": "3:36:39", "remaining_time": "0:15:28"}
+{"current_steps": 1060, "total_steps": 1125, "loss": 0.2279, "accuracy": 0.9375, "lr": 5.0722585640090305e-08, "epoch": 0.9422222222222222, "percentage": 94.22, "elapsed_time": "3:37:44", "remaining_time": "0:13:21"}
+{"current_steps": 1070, "total_steps": 1125, "loss": 0.1021, "accuracy": 0.9624999761581421, "lr": 3.635120570700784e-08, "epoch": 0.9511111111111111, "percentage": 95.11, "elapsed_time": "3:38:49", "remaining_time": "0:11:14"}
+{"current_steps": 1080, "total_steps": 1125, "loss": 0.1189, "accuracy": 0.9750000238418579, "lr": 2.4353836298169343e-08, "epoch": 0.96, "percentage": 96.0, "elapsed_time": "3:39:54", "remaining_time": "0:09:09"}
+{"current_steps": 1080, "total_steps": 1125, "eval_loss": 0.15081512928009033, "epoch": 0.96, "percentage": 96.0, "elapsed_time": "3:45:55", "remaining_time": "0:09:24"}
+{"current_steps": 1090, "total_steps": 1125, "loss": 0.0728, "accuracy": 0.9750000238418579, "lr": 1.4742038266447046e-08, "epoch": 0.9688888888888889, "percentage": 96.89, "elapsed_time": "3:47:00", "remaining_time": "0:07:17"}
+{"current_steps": 1100, "total_steps": 1125, "loss": 0.0357, "accuracy": 0.9750000238418579, "lr": 7.525073690809737e-09, "epoch": 0.9777777777777777, "percentage": 97.78, "elapsed_time": "3:48:05", "remaining_time": "0:05:11"}
+{"current_steps": 1110, "total_steps": 1125, "loss": 0.1973, "accuracy": 0.949999988079071, "lr": 2.709896951238744e-09, "epoch": 0.9866666666666667, "percentage": 98.67, "elapsed_time": "3:49:10", "remaining_time": "0:03:05"}
+{"current_steps": 1120, "total_steps": 1125, "loss": 0.0538, "accuracy": 0.987500011920929, "lr": 3.0114802737818415e-10, "epoch": 0.9955555555555555, "percentage": 99.56, "elapsed_time": "3:50:15", "remaining_time": "0:01:01"}
+{"current_steps": 1125, "total_steps": 1125, "epoch": 1.0, "percentage": 100.0, "elapsed_time": "3:50:48", "remaining_time": "0:00:00"}

trainer_state.json

@@ -0,0 +1,2010 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 60,
+  "global_step": 1125,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 8.760091781616211,
+      "learning_rate": 4.4247787610619474e-07,
+      "logits/chosen": -0.8248252868652344,
+      "logits/rejected": -0.8263720273971558,
+      "logps/chosen": -0.36086463928222656,
+      "logps/rejected": -5.696224689483643,
+      "loss": 1.1038,
+      "rewards/accuracies": 0.5125000476837158,
+      "rewards/chosen": 17.43745994567871,
+      "rewards/margins": 0.5984855890274048,
+      "rewards/rejected": 16.838973999023438,
+      "step": 10
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 8.855981826782227,
+      "learning_rate": 8.849557522123895e-07,
+      "logits/chosen": -0.8169006109237671,
+      "logits/rejected": -0.819770872592926,
+      "logps/chosen": -0.12464660406112671,
+      "logps/rejected": -7.139842987060547,
+      "loss": 1.1887,
+      "rewards/accuracies": 0.4000000059604645,
+      "rewards/chosen": 17.17649269104004,
+      "rewards/margins": 0.19107049703598022,
+      "rewards/rejected": 16.98542022705078,
+      "step": 20
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 16.764184951782227,
+      "learning_rate": 1.3274336283185843e-06,
+      "logits/chosen": -0.8003113865852356,
+      "logits/rejected": -0.8030117750167847,
+      "logps/chosen": -0.34651467204093933,
+      "logps/rejected": -6.967917442321777,
+      "loss": 1.0563,
+      "rewards/accuracies": 0.44999998807907104,
+      "rewards/chosen": 17.280975341796875,
+      "rewards/margins": 0.40005987882614136,
+      "rewards/rejected": 16.88091468811035,
+      "step": 30
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 8.33682918548584,
+      "learning_rate": 1.769911504424779e-06,
+      "logits/chosen": -0.7695047855377197,
+      "logits/rejected": -0.7739207148551941,
+      "logps/chosen": -1.5993006229400635,
+      "logps/rejected": -8.504932403564453,
+      "loss": 0.7596,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 17.283912658691406,
+      "rewards/margins": 0.6976072192192078,
+      "rewards/rejected": 16.5863037109375,
+      "step": 40
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 4.494723320007324,
+      "learning_rate": 2.212389380530974e-06,
+      "logits/chosen": -0.7154140472412109,
+      "logits/rejected": -0.7225576043128967,
+      "logps/chosen": -3.112199068069458,
+      "logps/rejected": -12.212080001831055,
+      "loss": 0.6083,
+      "rewards/accuracies": 0.4625000059604645,
+      "rewards/chosen": 17.03064727783203,
+      "rewards/margins": 0.7148451805114746,
+      "rewards/rejected": 16.3158016204834,
+      "step": 50
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 5.110287666320801,
+      "learning_rate": 2.6548672566371687e-06,
+      "logits/chosen": -0.6322453022003174,
+      "logits/rejected": -0.6387485265731812,
+      "logps/chosen": -5.650620460510254,
+      "logps/rejected": -12.759811401367188,
+      "loss": 0.3835,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.101289749145508,
+      "rewards/margins": 1.1824612617492676,
+      "rewards/rejected": 15.918828964233398,
+      "step": 60
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "eval_logits/chosen": -0.5826543569564819,
+      "eval_logits/rejected": -0.5914276838302612,
+      "eval_logps/chosen": -3.5471787452697754,
+      "eval_logps/rejected": -16.51181983947754,
+      "eval_loss": 0.3286525011062622,
+      "eval_rewards/accuracies": 0.9280000925064087,
+      "eval_rewards/chosen": 17.148174285888672,
+      "eval_rewards/margins": 1.4386365413665771,
+      "eval_rewards/rejected": 15.709539413452148,
+      "eval_runtime": 372.0227,
+      "eval_samples_per_second": 2.688,
+      "eval_steps_per_second": 0.336,
+      "step": 60
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 5.098133563995361,
+      "learning_rate": 3.097345132743363e-06,
+      "logits/chosen": -0.5378152132034302,
+      "logits/rejected": -0.5494933724403381,
+      "logps/chosen": -1.5099802017211914,
+      "logps/rejected": -21.206321716308594,
+      "loss": 0.2931,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.083791732788086,
+      "rewards/margins": 1.5844331979751587,
+      "rewards/rejected": 15.499359130859375,
+      "step": 70
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 29.787437438964844,
+      "learning_rate": 3.539823008849558e-06,
+      "logits/chosen": -0.443774938583374,
+      "logits/rejected": -0.45571577548980713,
+      "logps/chosen": -1.5804342031478882,
+      "logps/rejected": -22.606929779052734,
+      "loss": 0.202,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.302125930786133,
+      "rewards/margins": 2.174014091491699,
+      "rewards/rejected": 15.128110885620117,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.14398193359375,
+      "learning_rate": 3.982300884955752e-06,
+      "logits/chosen": -0.3626072406768799,
+      "logits/rejected": -0.3787815570831299,
+      "logps/chosen": -2.203828811645508,
+      "logps/rejected": -29.433551788330078,
+      "loss": 0.2123,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.00284194946289,
+      "rewards/margins": 2.320391893386841,
+      "rewards/rejected": 14.682450294494629,
+      "step": 90
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 29.672739028930664,
+      "learning_rate": 4.424778761061948e-06,
+      "logits/chosen": -0.3035663962364197,
+      "logits/rejected": -0.31762221455574036,
+      "logps/chosen": -3.433589458465576,
+      "logps/rejected": -29.9322509765625,
+      "loss": 0.2592,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.929956436157227,
+      "rewards/margins": 2.31272029876709,
+      "rewards/rejected": 14.617237091064453,
+      "step": 100
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 1.873722791671753,
+      "learning_rate": 4.867256637168142e-06,
+      "logits/chosen": -0.2679600715637207,
+      "logits/rejected": -0.2826440930366516,
+      "logps/chosen": -0.9653514623641968,
+      "logps/rejected": -30.235322952270508,
+      "loss": 0.1336,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.462385177612305,
+      "rewards/margins": 3.1994175910949707,
+      "rewards/rejected": 14.26296615600586,
+      "step": 110
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 1.6913721561431885,
+      "learning_rate": 4.999409761242696e-06,
+      "logits/chosen": -0.22222033143043518,
+      "logits/rejected": -0.23720571398735046,
+      "logps/chosen": -4.4953508377075195,
+      "logps/rejected": -34.074745178222656,
+      "loss": 0.2552,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.04866600036621,
+      "rewards/margins": 3.1014418601989746,
+      "rewards/rejected": 13.947224617004395,
+      "step": 120
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "eval_logits/chosen": -0.206527978181839,
+      "eval_logits/rejected": -0.22178640961647034,
+      "eval_logps/chosen": -3.69442081451416,
+      "eval_logps/rejected": -36.072166442871094,
+      "eval_loss": 0.18996010720729828,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.133451461791992,
+      "eval_rewards/margins": 3.379946708679199,
+      "eval_rewards/rejected": 13.753504753112793,
+      "eval_runtime": 361.5279,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 120
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 61.80262756347656,
+      "learning_rate": 4.996519466816778e-06,
+      "logits/chosen": -0.18473535776138306,
+      "logits/rejected": -0.1988501250743866,
+      "logps/chosen": -3.7009687423706055,
+      "logps/rejected": -39.289939880371094,
+      "loss": 0.1394,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.106964111328125,
+      "rewards/margins": 3.633338212966919,
+      "rewards/rejected": 13.473625183105469,
+      "step": 130
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 1.6732702255249023,
+      "learning_rate": 4.9912234871722805e-06,
+      "logits/chosen": -0.16134041547775269,
+      "logits/rejected": -0.17547868192195892,
+      "logps/chosen": -3.0637736320495605,
+      "logps/rejected": -40.07548522949219,
+      "loss": 0.1408,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.392223358154297,
+      "rewards/margins": 4.242353439331055,
+      "rewards/rejected": 13.149867057800293,
+      "step": 140
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.346453994512558,
+      "learning_rate": 4.98352692559805e-06,
+      "logits/chosen": -0.13797929883003235,
+      "logits/rejected": -0.15283086895942688,
+      "logps/chosen": -5.14492130279541,
+      "logps/rejected": -47.97212219238281,
+      "loss": 0.2153,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.896778106689453,
+      "rewards/margins": 4.227695465087891,
+      "rewards/rejected": 12.669081687927246,
+      "step": 150
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.21871662139892578,
+      "learning_rate": 4.973437198621237e-06,
+      "logits/chosen": -0.12396670132875443,
+      "logits/rejected": -0.13780555129051208,
+      "logps/chosen": -6.108860015869141,
+      "logps/rejected": -54.90739440917969,
+      "loss": 0.0388,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 16.75935935974121,
+      "rewards/margins": 4.755282878875732,
+      "rewards/rejected": 12.004077911376953,
+      "step": 160
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 235.12429809570312,
+      "learning_rate": 4.960964028860621e-06,
+      "logits/chosen": -0.1140839159488678,
+      "logits/rejected": -0.1263057291507721,
+      "logps/chosen": -12.605452537536621,
+      "logps/rejected": -53.81230926513672,
+      "loss": 0.4651,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.101238250732422,
+      "rewards/margins": 3.9864249229431152,
+      "rewards/rejected": 12.114812850952148,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 190.97048950195312,
+      "learning_rate": 4.946119435657738e-06,
+      "logits/chosen": -0.10746976733207703,
+      "logits/rejected": -0.11878640949726105,
+      "logps/chosen": -8.5105562210083,
+      "logps/rejected": -51.314781188964844,
+      "loss": 0.2362,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.719980239868164,
+      "rewards/margins": 4.549674034118652,
+      "rewards/rejected": 12.170306205749512,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "eval_logits/chosen": -0.10870806127786636,
+      "eval_logits/rejected": -0.12223993986845016,
+      "eval_logps/chosen": -4.414996147155762,
+      "eval_logps/rejected": -53.885032653808594,
+      "eval_loss": 0.20236633718013763,
+      "eval_rewards/accuracies": 0.9510000944137573,
+      "eval_rewards/chosen": 17.06139373779297,
+      "eval_rewards/margins": 5.089176177978516,
+      "eval_rewards/rejected": 11.97221851348877,
+      "eval_runtime": 361.4355,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 180
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 56.81266784667969,
+      "learning_rate": 4.928917723494854e-06,
+      "logits/chosen": -0.10682469606399536,
+      "logits/rejected": -0.12124393880367279,
+      "logps/chosen": -3.058413028717041,
+      "logps/rejected": -55.052528381347656,
+      "loss": 0.2442,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.058589935302734,
+      "rewards/margins": 5.056097984313965,
+      "rewards/rejected": 12.002490043640137,
+      "step": 190
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 175.06552124023438,
+      "learning_rate": 4.909375468210947e-06,
+      "logits/chosen": -0.10520349442958832,
+      "logits/rejected": -0.12018950283527374,
+      "logps/chosen": -4.114959716796875,
+      "logps/rejected": -55.9394645690918,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.98603057861328,
+      "rewards/margins": 5.105838775634766,
+      "rewards/rejected": 11.880191802978516,
+      "step": 200
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 78.06558990478516,
+      "learning_rate": 4.8875115010289655e-06,
+      "logits/chosen": -0.10475558042526245,
+      "logits/rejected": -0.11949175596237183,
+      "logps/chosen": -6.760301113128662,
+      "logps/rejected": -53.91607666015625,
+      "loss": 0.2843,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.857545852661133,
+      "rewards/margins": 4.917357921600342,
+      "rewards/rejected": 11.94018840789795,
+      "step": 210
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 15.880486488342285,
+      "learning_rate": 4.863346890409768e-06,
+      "logits/chosen": -0.11213523149490356,
+      "logits/rejected": -0.12581588327884674,
+      "logps/chosen": -6.759585380554199,
+      "logps/rejected": -51.10936737060547,
+      "loss": 0.5104,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 16.859071731567383,
+      "rewards/margins": 4.638372898101807,
+      "rewards/rejected": 12.220698356628418,
+      "step": 220
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 46.97845458984375,
+      "learning_rate": 4.836904921750224e-06,
+      "logits/chosen": -0.11947059631347656,
+      "logits/rejected": -0.1329912692308426,
+      "logps/chosen": -3.608184814453125,
+      "logps/rejected": -48.794761657714844,
+      "loss": 0.2134,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.235904693603516,
+      "rewards/margins": 4.859888076782227,
+      "rewards/rejected": 12.376014709472656,
+      "step": 230
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 24.032859802246094,
+      "learning_rate": 4.808211074945042e-06,
+      "logits/chosen": -0.1200513243675232,
+      "logits/rejected": -0.1333036869764328,
+      "logps/chosen": -3.7552154064178467,
+      "logps/rejected": -49.87453079223633,
+      "loss": 0.1781,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.094650268554688,
+      "rewards/margins": 4.68077278137207,
+      "rewards/rejected": 12.41387939453125,
+      "step": 240
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "eval_logits/chosen": -0.12433278560638428,
+      "eval_logits/rejected": -0.13808581233024597,
+      "eval_logps/chosen": -4.408891201019287,
+      "eval_logps/rejected": -50.744781494140625,
+      "eval_loss": 0.1546352356672287,
+      "eval_rewards/accuracies": 0.9500000476837158,
+      "eval_rewards/chosen": 17.06200408935547,
+      "eval_rewards/margins": 4.775761604309082,
+      "eval_rewards/rejected": 12.286243438720703,
+      "eval_runtime": 361.4974,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 240
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.25737640261650085,
+      "learning_rate": 4.7772929998339485e-06,
+      "logits/chosen": -0.12348780035972595,
+      "logits/rejected": -0.13704943656921387,
+      "logps/chosen": -4.4299187660217285,
+      "logps/rejected": -53.074607849121094,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.087068557739258,
+      "rewards/margins": 5.06691837310791,
+      "rewards/rejected": 12.020149230957031,
+      "step": 250
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 0.1839389204978943,
+      "learning_rate": 4.744180489557859e-06,
+      "logits/chosen": -0.12177034467458725,
+      "logits/rejected": -0.1342695653438568,
+      "logps/chosen": -3.775188446044922,
+      "logps/rejected": -53.98720932006836,
+      "loss": 0.1896,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.12021255493164,
+      "rewards/margins": 5.148064613342285,
+      "rewards/rejected": 11.972146987915039,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.258485794067383,
+      "learning_rate": 4.708905451849754e-06,
+      "logits/chosen": -0.11067859083414078,
+      "logits/rejected": -0.12377731502056122,
+      "logps/chosen": -6.418317794799805,
+      "logps/rejected": -56.57402801513672,
+      "loss": 0.2315,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 16.738832473754883,
+      "rewards/margins": 4.884931564331055,
+      "rewards/rejected": 11.853900909423828,
+      "step": 270
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 77.56194305419922,
+      "learning_rate": 4.671501878287879e-06,
+      "logits/chosen": -0.1184445172548294,
+      "logits/rejected": -0.1339874565601349,
+      "logps/chosen": -10.12116527557373,
+      "logps/rejected": -53.403907775878906,
+      "loss": 0.5343,
+      "rewards/accuracies": 0.862500011920929,
+      "rewards/chosen": 16.458633422851562,
+      "rewards/margins": 4.402472496032715,
+      "rewards/rejected": 12.056160926818848,
+      "step": 280
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 67.53883361816406,
+      "learning_rate": 4.6320058115409295e-06,
+      "logits/chosen": -0.1448262631893158,
+      "logits/rejected": -0.15793387591838837,
+      "logps/chosen": -3.4666190147399902,
+      "logps/rejected": -48.79213333129883,
+      "loss": 0.5017,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 16.945899963378906,
+      "rewards/margins": 4.2686333656311035,
+      "rewards/rejected": 12.677268028259277,
+      "step": 290
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.17521341145038605,
+      "learning_rate": 4.590455310636778e-06,
+      "logits/chosen": -0.16128253936767578,
+      "logits/rejected": -0.17375555634498596,
+      "logps/chosen": -2.9032950401306152,
+      "logps/rejected": -47.69734191894531,
+      "loss": 0.265,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.18383026123047,
+      "rewards/margins": 4.541309356689453,
+      "rewards/rejected": 12.642518997192383,
+      "step": 300
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "eval_logits/chosen": -0.17444846034049988,
+      "eval_logits/rejected": -0.18559777736663818,
+      "eval_logps/chosen": -2.535512924194336,
+      "eval_logps/rejected": -47.16367721557617,
+      "eval_loss": 0.15360687673091888,
+      "eval_rewards/accuracies": 0.9440000653266907,
+      "eval_rewards/chosen": 17.249343872070312,
+      "eval_rewards/margins": 4.604989051818848,
+      "eval_rewards/rejected": 12.644353866577148,
+      "eval_runtime": 361.4575,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 300
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 0.5040452480316162,
+      "learning_rate": 4.54689041428819e-06,
+      "logits/chosen": -0.16974106431007385,
+      "logits/rejected": -0.1810058057308197,
+      "logps/chosen": -1.233938217163086,
+      "logps/rejected": -49.907745361328125,
+      "loss": 0.1132,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.34117889404297,
+      "rewards/margins": 4.934173583984375,
+      "rewards/rejected": 12.407005310058594,
+      "step": 310
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 100.02949523925781,
+      "learning_rate": 4.501353102310901e-06,
+      "logits/chosen": -0.15705889463424683,
+      "logits/rejected": -0.1695334017276764,
+      "logps/chosen": -1.0820492506027222,
+      "logps/rejected": -52.577110290527344,
+      "loss": 0.1194,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.33388900756836,
+      "rewards/margins": 5.154760837554932,
+      "rewards/rejected": 12.179126739501953,
+      "step": 320
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 0.2689219117164612,
+      "learning_rate": 4.453887255171206e-06,
+      "logits/chosen": -0.13849371671676636,
+      "logits/rejected": -0.14990833401679993,
+      "logps/chosen": -1.8435032367706299,
+      "logps/rejected": -54.79044723510742,
+      "loss": 0.0926,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.2423095703125,
+      "rewards/margins": 5.28987979888916,
+      "rewards/rejected": 11.952428817749023,
+      "step": 330
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 0.09305431693792343,
+      "learning_rate": 4.404538611702055e-06,
+      "logits/chosen": -0.12299702316522598,
+      "logits/rejected": -0.13453055918216705,
+      "logps/chosen": -2.9897143840789795,
+      "logps/rejected": -52.954498291015625,
+      "loss": 0.2873,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.17474365234375,
+      "rewards/margins": 5.071004867553711,
+      "rewards/rejected": 12.103739738464355,
+      "step": 340
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 59.282073974609375,
+      "learning_rate": 4.3533547250284015e-06,
+      "logits/chosen": -0.11913029849529266,
+      "logits/rejected": -0.12785324454307556,
+      "logps/chosen": -3.9456872940063477,
+      "logps/rejected": -48.68487548828125,
+      "loss": 0.4332,
+      "rewards/accuracies": 0.875,
+      "rewards/chosen": 17.12805938720703,
+      "rewards/margins": 4.669450283050537,
+      "rewards/rejected": 12.458610534667969,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31101909279823303,
+      "learning_rate": 4.300384916744261e-06,
+      "logits/chosen": -0.11280188709497452,
+      "logits/rejected": -0.12300585210323334,
+      "logps/chosen": -2.1714723110198975,
+      "logps/rejected": -54.74174118041992,
+      "loss": 0.1605,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.326162338256836,
+      "rewards/margins": 5.467062473297119,
+      "rewards/rejected": 11.859098434448242,
+      "step": 360
+    },
+    {
+      "epoch": 0.32,
+      "eval_logits/chosen": -0.10620756447315216,
+      "eval_logits/rejected": -0.11727114766836166,
+      "eval_logps/chosen": -1.4165427684783936,
+      "eval_logps/rejected": -50.9525146484375,
+      "eval_loss": 0.3194349706172943,
+      "eval_rewards/accuracies": 0.9210000038146973,
+      "eval_rewards/chosen": 17.36124038696289,
+      "eval_rewards/margins": 5.095769882202148,
+      "eval_rewards/rejected": 12.26546859741211,
+      "eval_runtime": 361.5072,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 360
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 6.1126532554626465,
+      "learning_rate": 4.24568022938566e-06,
+      "logits/chosen": -0.10354311764240265,
+      "logits/rejected": -0.11526636779308319,
+      "logps/chosen": -1.2935255765914917,
+      "logps/rejected": -55.57566833496094,
+      "loss": 0.1711,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.439346313476562,
+      "rewards/margins": 5.700921058654785,
+      "rewards/rejected": 11.738424301147461,
+      "step": 370
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 34.15927505493164,
+      "learning_rate": 4.189293377245241e-06,
+      "logits/chosen": -0.1029932051897049,
+      "logits/rejected": -0.11382515728473663,
+      "logps/chosen": -2.5132687091827393,
+      "logps/rejected": -55.50346374511719,
+      "loss": 0.4359,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.731037139892578,
+      "rewards/margins": 4.368172645568848,
+      "rewards/rejected": 12.362865447998047,
+      "step": 380
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 2.8422904014587402,
+      "learning_rate": 4.131278695575952e-06,
+      "logits/chosen": -0.10793520510196686,
+      "logits/rejected": -0.12109285593032837,
+      "logps/chosen": -3.014652729034424,
+      "logps/rejected": -53.98411560058594,
+      "loss": 0.2161,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.137393951416016,
+      "rewards/margins": 5.105995178222656,
+      "rewards/rejected": 12.03139877319336,
+      "step": 390
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 54.0329475402832,
+      "learning_rate": 4.071692088232743e-06,
+      "logits/chosen": -0.10393750667572021,
+      "logits/rejected": -0.11834606528282166,
+      "logps/chosen": -2.1508543491363525,
+      "logps/rejected": -45.60733413696289,
+      "loss": 0.2077,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.586124420166016,
+      "rewards/margins": 5.077212333679199,
+      "rewards/rejected": 12.5089111328125,
+      "step": 400
+    },
+    {
+      "epoch": 0.36444444444444446,
+      "grad_norm": 81.61144256591797,
+      "learning_rate": 4.010590973802737e-06,
+      "logits/chosen": -0.09564584493637085,
+      "logits/rejected": -0.10617707669734955,
+      "logps/chosen": -3.4572842121124268,
+      "logps/rejected": -50.92162322998047,
+      "loss": 0.2478,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 17.010910034179688,
+      "rewards/margins": 4.556198596954346,
+      "rewards/rejected": 12.454713821411133,
+      "step": 410
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "grad_norm": 0.30974289774894714,
+      "learning_rate": 3.948034230275781e-06,
+      "logits/chosen": -0.09134417027235031,
+      "logits/rejected": -0.1020016297698021,
+      "logps/chosen": -5.046698570251465,
+      "logps/rejected": -48.908958435058594,
+      "loss": 0.2894,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.007888793945312,
+      "rewards/margins": 4.53641414642334,
+      "rewards/rejected": 12.471475601196289,
+      "step": 420
+    },
+    {
+      "epoch": 0.37333333333333335,
+      "eval_logits/chosen": -0.09054450690746307,
+      "eval_logits/rejected": -0.10264354199171066,
+      "eval_logps/chosen": -1.913105845451355,
+      "eval_logps/rejected": -51.11127471923828,
+      "eval_loss": 0.16789735853672028,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.311582565307617,
+      "eval_rewards/margins": 5.061989784240723,
+      "eval_rewards/rejected": 12.249593734741211,
+      "eval_runtime": 361.5337,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 420
+    },
+    {
+      "epoch": 0.38222222222222224,
+      "grad_norm": 12.824393272399902,
+      "learning_rate": 3.884082138308699e-06,
+      "logits/chosen": -0.08666776865720749,
+      "logits/rejected": -0.0997733399271965,
+      "logps/chosen": -1.7306327819824219,
+      "logps/rejected": -54.273292541503906,
+      "loss": 0.2298,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.167621612548828,
+      "rewards/margins": 5.065673351287842,
+      "rewards/rejected": 12.101947784423828,
+      "step": 430
+    },
+    {
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.30713599920272827,
+      "learning_rate": 3.818796323137896e-06,
+      "logits/chosen": -0.09174907952547073,
+      "logits/rejected": -0.10376611351966858,
+      "logps/chosen": -1.489154577255249,
+      "logps/rejected": -54.580726623535156,
+      "loss": 0.2513,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.22280502319336,
+      "rewards/margins": 5.175349235534668,
+      "rewards/rejected": 12.047454833984375,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 87.4791488647461,
+      "learning_rate": 3.7522396951963303e-06,
+      "logits/chosen": -0.09688778221607208,
+      "logits/rejected": -0.10897806286811829,
+      "logps/chosen": -3.157695770263672,
+      "logps/rejected": -50.96417236328125,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.345651626586914,
+      "rewards/margins": 5.245656967163086,
+      "rewards/rejected": 12.099993705749512,
+      "step": 450
+    },
+    {
+      "epoch": 0.4088888888888889,
+      "grad_norm": 146.2008056640625,
+      "learning_rate": 3.684476389492026e-06,
+      "logits/chosen": -0.09378582239151001,
+      "logits/rejected": -0.10475654900074005,
+      "logps/chosen": -0.5611928701400757,
+      "logps/rejected": -56.518890380859375,
+      "loss": 0.1981,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.113712310791016,
+      "rewards/margins": 5.068872928619385,
+      "rewards/rejected": 12.044839859008789,
+      "step": 460
+    },
+    {
+      "epoch": 0.4177777777777778,
+      "grad_norm": 1.9137721061706543,
+      "learning_rate": 3.6155717038065783e-06,
+      "logits/chosen": -0.08695463836193085,
+      "logits/rejected": -0.09596743434667587,
+      "logps/chosen": -1.5298550128936768,
+      "logps/rejected": -50.27445983886719,
+      "loss": 0.2066,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.35186004638672,
+      "rewards/margins": 5.014693260192871,
+      "rewards/rejected": 12.337167739868164,
+      "step": 470
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "grad_norm": 84.80391693115234,
+      "learning_rate": 3.545592035773192e-06,
+      "logits/chosen": -0.0746893435716629,
+      "logits/rejected": -0.08653923869132996,
+      "logps/chosen": -2.0052125453948975,
+      "logps/rejected": -57.502811431884766,
+      "loss": 0.1149,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.14373016357422,
+      "rewards/margins": 5.360415935516357,
+      "rewards/rejected": 11.783313751220703,
+      "step": 480
+    },
+    {
+      "epoch": 0.4266666666666667,
+      "eval_logits/chosen": -0.07700399309396744,
+      "eval_logits/rejected": -0.08828537166118622,
+      "eval_logps/chosen": -4.48896598815918,
+      "eval_logps/rejected": -53.76282501220703,
+      "eval_loss": 0.29511645436286926,
+      "eval_rewards/accuracies": 0.9230000376701355,
+      "eval_rewards/chosen": 17.053997039794922,
+      "eval_rewards/margins": 5.069558143615723,
+      "eval_rewards/rejected": 11.984437942504883,
+      "eval_runtime": 361.5035,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 480
+    },
+    {
+      "epoch": 0.43555555555555553,
+      "grad_norm": 82.9616470336914,
+      "learning_rate": 3.4746048188948806e-06,
+      "logits/chosen": -0.06675051152706146,
+      "logits/rejected": -0.07860895991325378,
+      "logps/chosen": -4.162237167358398,
+      "logps/rejected": -54.77789306640625,
+      "loss": 0.2979,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.047603607177734,
+      "rewards/margins": 5.138361930847168,
+      "rewards/rejected": 11.909242630004883,
+      "step": 490
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.04293210059404373,
+      "learning_rate": 3.4026784575644887e-06,
+      "logits/chosen": -0.06424491107463837,
+      "logits/rejected": -0.07567107677459717,
+      "logps/chosen": -2.05729603767395,
+      "logps/rejected": -56.646087646484375,
+      "loss": 0.4378,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.947803497314453,
+      "rewards/margins": 4.919981956481934,
+      "rewards/rejected": 12.02782154083252,
+      "step": 500
+    },
+    {
+      "epoch": 0.4533333333333333,
+      "grad_norm": 0.07136644423007965,
+      "learning_rate": 3.329882261149148e-06,
+      "logits/chosen": -0.06423303484916687,
+      "logits/rejected": -0.07512776553630829,
+      "logps/chosen": -3.1519265174865723,
+      "logps/rejected": -53.53008270263672,
+      "loss": 0.2613,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.112144470214844,
+      "rewards/margins": 5.046430587768555,
+      "rewards/rejected": 12.065712928771973,
+      "step": 510
+    },
+    {
+      "epoch": 0.4622222222222222,
+      "grad_norm": 0.29279613494873047,
+      "learning_rate": 3.25628637720269e-06,
+      "logits/chosen": -0.060233693569898605,
+      "logits/rejected": -0.07076811790466309,
+      "logps/chosen": -1.2358124256134033,
+      "logps/rejected": -50.61806106567383,
+      "loss": 0.1517,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.457351684570312,
+      "rewards/margins": 5.2062835693359375,
+      "rewards/rejected": 12.251070022583008,
+      "step": 520
+    },
+    {
+      "epoch": 0.4711111111111111,
+      "grad_norm": 0.7946074604988098,
+      "learning_rate": 3.181961723870359e-06,
+      "logits/chosen": -0.054482050240039825,
+      "logits/rejected": -0.0657092034816742,
+      "logps/chosen": -0.7682158946990967,
+      "logps/rejected": -58.81409454345703,
+      "loss": 0.2304,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.075973510742188,
+      "rewards/margins": 5.222441673278809,
+      "rewards/rejected": 11.853530883789062,
+      "step": 530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1405210494995117,
+      "learning_rate": 3.1069799215509847e-06,
+      "logits/chosen": -0.05050881579518318,
+      "logits/rejected": -0.061149902641773224,
+      "logps/chosen": -0.954046368598938,
+      "logps/rejected": -52.70227813720703,
+      "loss": 0.0384,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.378849029541016,
+      "rewards/margins": 5.253483295440674,
+      "rewards/rejected": 12.125364303588867,
+      "step": 540
+    },
+    {
+      "epoch": 0.48,
+      "eval_logits/chosen": -0.05120665580034256,
+      "eval_logits/rejected": -0.06123337894678116,
+      "eval_logps/chosen": -2.9872913360595703,
+      "eval_logps/rejected": -52.27314758300781,
+      "eval_loss": 0.17387841641902924,
+      "eval_rewards/accuracies": 0.9490000605583191,
+      "eval_rewards/chosen": 17.204164505004883,
+      "eval_rewards/margins": 5.070757865905762,
+      "eval_rewards/rejected": 12.133406639099121,
+      "eval_runtime": 361.5449,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 540
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.07270358502864838,
+      "learning_rate": 3.0314132238824416e-06,
+      "logits/chosen": -0.05125313252210617,
+      "logits/rejected": -0.06174170970916748,
+      "logps/chosen": -3.2163877487182617,
+      "logps/rejected": -47.79279327392578,
+      "loss": 0.2087,
+      "rewards/accuracies": 0.9125000238418579,
+      "rewards/chosen": 17.408517837524414,
+      "rewards/margins": 5.061524391174316,
+      "rewards/rejected": 12.346992492675781,
+      "step": 550
+    },
+    {
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.10005924850702286,
+      "learning_rate": 2.955334448116915e-06,
+      "logits/chosen": -0.041773442178964615,
+      "logits/rejected": -0.05364570394158363,
+      "logps/chosen": -0.363404780626297,
+      "logps/rejected": -56.32415008544922,
+      "loss": 0.0969,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.3824462890625,
+      "rewards/margins": 5.5404510498046875,
+      "rewards/rejected": 11.841995239257812,
+      "step": 560
+    },
+    {
+      "epoch": 0.5066666666666667,
+      "grad_norm": 7.818356990814209,
+      "learning_rate": 2.8788169049530533e-06,
+      "logits/chosen": -0.04309462010860443,
+      "logits/rejected": -0.05494442582130432,
+      "logps/chosen": -2.2242724895477295,
+      "logps/rejected": -56.444740295410156,
+      "loss": 0.1447,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.275390625,
+      "rewards/margins": 5.55007791519165,
+      "rewards/rejected": 11.725313186645508,
+      "step": 570
+    },
+    {
+      "epoch": 0.5155555555555555,
+      "grad_norm": 0.03519747406244278,
+      "learning_rate": 2.8019343278926397e-06,
+      "logits/chosen": -0.03506368771195412,
+      "logits/rejected": -0.046854715794324875,
+      "logps/chosen": -0.5200096964836121,
+      "logps/rejected": -59.05330276489258,
+      "loss": 0.0998,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.320554733276367,
+      "rewards/margins": 5.727260589599609,
+      "rewards/rejected": 11.593294143676758,
+      "step": 580
+    },
+    {
+      "epoch": 0.5244444444444445,
+      "grad_norm": 0.04519123584032059,
+      "learning_rate": 2.7247608021898265e-06,
+      "logits/chosen": -0.03204537555575371,
+      "logits/rejected": -0.04383891448378563,
+      "logps/chosen": -1.1271060705184937,
+      "logps/rejected": -59.308895111083984,
+      "loss": 0.1162,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.256351470947266,
+      "rewards/margins": 5.6881890296936035,
+      "rewards/rejected": 11.568161010742188,
+      "step": 590
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.028689857572317123,
+      "learning_rate": 2.647370693461432e-06,
+      "logits/chosen": -0.02834726870059967,
+      "logits/rejected": -0.03827046602964401,
+      "logps/chosen": -5.673943519592285,
+      "logps/rejected": -55.72624588012695,
+      "loss": 0.4008,
+      "rewards/accuracies": 0.8875000476837158,
+      "rewards/chosen": 16.761056900024414,
+      "rewards/margins": 4.804043769836426,
+      "rewards/rejected": 11.957012176513672,
+      "step": 600
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "eval_logits/chosen": -0.035840023308992386,
+      "eval_logits/rejected": -0.04687971621751785,
+      "eval_logps/chosen": -2.1760435104370117,
+      "eval_logps/rejected": -56.62664031982422,
+      "eval_loss": 0.17061151564121246,
+      "eval_rewards/accuracies": 0.9470000267028809,
+      "eval_rewards/chosen": 17.285289764404297,
+      "eval_rewards/margins": 5.587231636047363,
+      "eval_rewards/rejected": 11.698057174682617,
+      "eval_runtime": 361.5056,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 600
+    },
+    {
+      "epoch": 0.5422222222222223,
+      "grad_norm": 1.1994622945785522,
+      "learning_rate": 2.569838576027068e-06,
+      "logits/chosen": -0.03231767192482948,
+      "logits/rejected": -0.04397805407643318,
+      "logps/chosen": -1.5904741287231445,
+      "logps/rejected": -62.10520553588867,
+      "loss": 0.1265,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.11206817626953,
+      "rewards/margins": 5.734784126281738,
+      "rewards/rejected": 11.37728500366211,
+      "step": 610
+    },
+    {
+      "epoch": 0.5511111111111111,
+      "grad_norm": 0.14163845777511597,
+      "learning_rate": 2.4922391610481544e-06,
+      "logits/chosen": -0.03293662518262863,
+      "logits/rejected": -0.04335154965519905,
+      "logps/chosen": -1.3568997383117676,
+      "logps/rejected": -58.347434997558594,
+      "loss": 0.097,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.345399856567383,
+      "rewards/margins": 5.823373794555664,
+      "rewards/rejected": 11.522026062011719,
+      "step": 620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.125027060508728,
+      "learning_rate": 2.4146472245350804e-06,
+      "logits/chosen": -0.02864963933825493,
+      "logits/rejected": -0.03835710883140564,
+      "logps/chosen": -2.494175434112549,
+      "logps/rejected": -55.33067321777344,
+      "loss": 0.3115,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.262205123901367,
+      "rewards/margins": 5.4287261962890625,
+      "rewards/rejected": 11.833479881286621,
+      "step": 630
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.03514016419649124,
+      "learning_rate": 2.337137535291868e-06,
+      "logits/chosen": -0.02757749892771244,
+      "logits/rejected": -0.0376611053943634,
+      "logps/chosen": -2.3163387775421143,
+      "logps/rejected": -53.54579162597656,
+      "loss": 0.1268,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.542863845825195,
+      "rewards/margins": 5.799897193908691,
+      "rewards/rejected": 11.742965698242188,
+      "step": 640
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 34.20791244506836,
+      "learning_rate": 2.259784782867782e-06,
+      "logits/chosen": -0.02032250165939331,
+      "logits/rejected": -0.031542714685201645,
+      "logps/chosen": -1.7910137176513672,
+      "logps/rejected": -62.980018615722656,
+      "loss": 0.1831,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 16.982879638671875,
+      "rewards/margins": 5.589078903198242,
+      "rewards/rejected": 11.393800735473633,
+      "step": 650
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 32.356956481933594,
+      "learning_rate": 2.182663505585314e-06,
+      "logits/chosen": -0.01711965538561344,
+      "logits/rejected": -0.028038471937179565,
+      "logps/chosen": -2.662904739379883,
+      "logps/rejected": -63.326297760009766,
+      "loss": 0.1678,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 16.997777938842773,
+      "rewards/margins": 5.741157531738281,
+      "rewards/rejected": 11.256620407104492,
+      "step": 660
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "eval_logits/chosen": -0.016018809750676155,
+      "eval_logits/rejected": -0.027020033448934555,
+      "eval_logps/chosen": -3.00822114944458,
+      "eval_logps/rejected": -57.951629638671875,
+      "eval_loss": 0.2050127536058426,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.202072143554688,
+      "eval_rewards/margins": 5.636512756347656,
+      "eval_rewards/rejected": 11.565557479858398,
+      "eval_runtime": 361.5073,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 660
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 86.53874969482422,
+      "learning_rate": 2.1058480187138863e-06,
+      "logits/chosen": -0.012352555990219116,
+      "logits/rejected": -0.023966707289218903,
+      "logps/chosen": -2.129209041595459,
+      "logps/rejected": -62.1393928527832,
+      "loss": 0.1025,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.069538116455078,
+      "rewards/margins": 5.678930759429932,
+      "rewards/rejected": 11.390605926513672,
+      "step": 670
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 2.892672300338745,
+      "learning_rate": 2.0294123428584985e-06,
+      "logits/chosen": -0.011562807485461235,
+      "logits/rejected": -0.020859256386756897,
+      "logps/chosen": -3.2486608028411865,
+      "logps/rejected": -59.56721496582031,
+      "loss": 0.1961,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.051376342773438,
+      "rewards/margins": 5.515361785888672,
+      "rewards/rejected": 11.536016464233398,
+      "step": 680
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 30.26588249206543,
+      "learning_rate": 1.953430132632311e-06,
+      "logits/chosen": -0.011488726362586021,
+      "logits/rejected": -0.021527227014303207,
+      "logps/chosen": -1.794357180595398,
+      "logps/rejected": -60.649505615234375,
+      "loss": 0.2821,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.114843368530273,
+      "rewards/margins": 5.612217903137207,
+      "rewards/rejected": 11.50262451171875,
+      "step": 690
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 0.030314341187477112,
+      "learning_rate": 1.8779746056819104e-06,
+      "logits/chosen": -0.014436552301049232,
+      "logits/rejected": -0.026044374331831932,
+      "logps/chosen": -3.1617178916931152,
+      "logps/rejected": -60.94548797607422,
+      "loss": 0.3141,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.08698272705078,
+      "rewards/margins": 5.719264984130859,
+      "rewards/rejected": 11.367716789245605,
+      "step": 700
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 0.06711317598819733,
+      "learning_rate": 1.8031184721336364e-06,
+      "logits/chosen": -0.017312290146946907,
+      "logits/rejected": -0.026554957032203674,
+      "logps/chosen": -5.120705604553223,
+      "logps/rejected": -54.33483123779297,
+      "loss": 0.3827,
+      "rewards/accuracies": 0.887499988079071,
+      "rewards/chosen": 17.1247501373291,
+      "rewards/margins": 5.327882766723633,
+      "rewards/rejected": 11.796867370605469,
+      "step": 710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 60.36637496948242,
+      "learning_rate": 1.7289338645289711e-06,
+      "logits/chosen": -0.01987219974398613,
+      "logits/rejected": -0.02766304835677147,
+      "logps/chosen": -2.1757986545562744,
+      "logps/rejected": -54.02531433105469,
+      "loss": 0.2272,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.239776611328125,
+      "rewards/margins": 5.250313758850098,
+      "rewards/rejected": 11.989462852478027,
+      "step": 720
+    },
+    {
+      "epoch": 0.64,
+      "eval_logits/chosen": -0.022905193269252777,
+      "eval_logits/rejected": -0.0321992002427578,
+      "eval_logps/chosen": -1.100506067276001,
+      "eval_logps/rejected": -55.91169738769531,
+      "eval_loss": 0.14018221199512482,
+      "eval_rewards/accuracies": 0.9520000219345093,
+      "eval_rewards/chosen": 17.392841339111328,
+      "eval_rewards/margins": 5.623291492462158,
+      "eval_rewards/rejected": 11.769551277160645,
+      "eval_runtime": 361.416,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 720
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 0.029294608160853386,
+      "learning_rate": 1.6554922683164875e-06,
+      "logits/chosen": -0.016416028141975403,
+      "logits/rejected": -0.025527067482471466,
+      "logps/chosen": -0.906692385673523,
+      "logps/rejected": -59.142173767089844,
+      "loss": 0.1171,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.282169342041016,
+      "rewards/margins": 5.6978583335876465,
+      "rewards/rejected": 11.584310531616211,
+      "step": 730
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 33.89070510864258,
+      "learning_rate": 1.5828644529673592e-06,
+      "logits/chosen": -0.016044551506638527,
+      "logits/rejected": -0.025811903178691864,
+      "logps/chosen": -1.1167538166046143,
+      "logps/rejected": -59.84492492675781,
+      "loss": 0.2001,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.266983032226562,
+      "rewards/margins": 5.749438285827637,
+      "rewards/rejected": 11.51754379272461,
+      "step": 740
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.53726726770401,
+      "learning_rate": 1.5111204037807844e-06,
+      "logits/chosen": -0.012529855594038963,
+      "logits/rejected": -0.023765765130519867,
+      "logps/chosen": -0.3461765646934509,
+      "logps/rejected": -55.841102600097656,
+      "loss": 0.1692,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.61865997314453,
+      "rewards/margins": 6.019055366516113,
+      "rewards/rejected": 11.599604606628418,
+      "step": 750
+    },
+    {
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.7173987627029419,
+      "learning_rate": 1.4403292544450625e-06,
+      "logits/chosen": -0.013380522839725018,
+      "logits/rejected": -0.022107835859060287,
+      "logps/chosen": -3.2759666442871094,
+      "logps/rejected": -53.836822509765625,
+      "loss": 0.2418,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.383333206176758,
+      "rewards/margins": 5.590358734130859,
+      "rewards/rejected": 11.792974472045898,
+      "step": 760
+    },
+    {
+      "epoch": 0.6844444444444444,
+      "grad_norm": 2.744900941848755,
+      "learning_rate": 1.3705592204192853e-06,
+      "logits/chosen": -0.011318420059978962,
+      "logits/rejected": -0.021840626373887062,
+      "logps/chosen": -2.592001438140869,
+      "logps/rejected": -57.06926345825195,
+      "loss": 0.1749,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.286128997802734,
+      "rewards/margins": 5.662715911865234,
+      "rewards/rejected": 11.623414993286133,
+      "step": 770
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "grad_norm": 70.76551818847656,
+      "learning_rate": 1.301877533199859e-06,
+      "logits/chosen": -0.013944407925009727,
+      "logits/rejected": -0.025667501613497734,
+      "logps/chosen": -2.1070234775543213,
+      "logps/rejected": -57.72039794921875,
+      "loss": 0.1915,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.28545379638672,
+      "rewards/margins": 5.708344459533691,
+      "rewards/rejected": 11.577108383178711,
+      "step": 780
+    },
+    {
+      "epoch": 0.6933333333333334,
+      "eval_logits/chosen": -0.01664295792579651,
+      "eval_logits/rejected": -0.026626665145158768,
+      "eval_logps/chosen": -1.082255244255066,
+      "eval_logps/rejected": -55.95073699951172,
+      "eval_loss": 0.24412688612937927,
+      "eval_rewards/accuracies": 0.9320000410079956,
+      "eval_rewards/chosen": 17.394668579101562,
+      "eval_rewards/margins": 5.629020690917969,
+      "eval_rewards/rejected": 11.765647888183594,
+      "eval_runtime": 361.4602,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 780
+    },
+    {
+      "epoch": 0.7022222222222222,
+      "grad_norm": 0.41359376907348633,
+      "learning_rate": 1.2343503755351729e-06,
+      "logits/chosen": -0.012772129848599434,
+      "logits/rejected": -0.023186586797237396,
+      "logps/chosen": -0.8030359148979187,
+      "logps/rejected": -57.84947967529297,
+      "loss": 0.2143,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.32242202758789,
+      "rewards/margins": 5.645486831665039,
+      "rewards/rejected": 11.676933288574219,
+      "step": 790
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.41748157143592834,
+      "learning_rate": 1.168042817650881e-06,
+      "logits/chosen": -0.012502101249992847,
+      "logits/rejected": -0.023272844031453133,
+      "logps/chosen": -1.0377256870269775,
+      "logps/rejected": -57.05615997314453,
+      "loss": 0.0792,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.60501480102539,
+      "rewards/margins": 6.159370422363281,
+      "rewards/rejected": 11.445646286010742,
+      "step": 800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.039210401475429535,
+      "learning_rate": 1.1030187545472012e-06,
+      "logits/chosen": -0.008177272044122219,
+      "logits/rejected": -0.01832464337348938,
+      "logps/chosen": -3.8701748847961426,
+      "logps/rejected": -57.661476135253906,
+      "loss": 0.3177,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.14336395263672,
+      "rewards/margins": 5.583393573760986,
+      "rewards/rejected": 11.55997085571289,
+      "step": 810
+    },
+    {
+      "epoch": 0.7288888888888889,
+      "grad_norm": 1.6096951961517334,
+      "learning_rate": 1.0393408444287048e-06,
+      "logits/chosen": -0.006832236424088478,
+      "logits/rejected": -0.01682097464799881,
+      "logps/chosen": -2.1741790771484375,
+      "logps/rejected": -57.273414611816406,
+      "loss": 0.2128,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.426712036132812,
+      "rewards/margins": 5.911205291748047,
+      "rewards/rejected": 11.515506744384766,
+      "step": 820
+    },
+    {
+      "epoch": 0.7377777777777778,
+      "grad_norm": 3.7820959091186523,
+      "learning_rate": 9.770704483258782e-07,
+      "logits/chosen": -0.009998206980526447,
+      "logits/rejected": -0.0204261876642704,
+      "logps/chosen": -1.9802953004837036,
+      "logps/rejected": -56.901512145996094,
+      "loss": 0.1296,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.56071662902832,
+      "rewards/margins": 6.169583320617676,
+      "rewards/rejected": 11.391134262084961,
+      "step": 830
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "grad_norm": 1.126626968383789,
+      "learning_rate": 9.162675709666865e-07,
+      "logits/chosen": -0.00826224498450756,
+      "logits/rejected": -0.018977787345647812,
+      "logps/chosen": -1.4256607294082642,
+      "logps/rejected": -61.65986633300781,
+      "loss": 0.0635,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.324975967407227,
+      "rewards/margins": 6.09440803527832,
+      "rewards/rejected": 11.230567932128906,
+      "step": 840
+    },
+    {
+      "epoch": 0.7466666666666667,
+      "eval_logits/chosen": -0.01111944392323494,
+      "eval_logits/rejected": -0.021697774529457092,
+      "eval_logps/chosen": -1.2168633937835693,
+      "eval_logps/rejected": -58.2642822265625,
+      "eval_loss": 0.1689341962337494,
+      "eval_rewards/accuracies": 0.9450000524520874,
+      "eval_rewards/chosen": 17.381206512451172,
+      "eval_rewards/margins": 5.846914291381836,
+      "eval_rewards/rejected": 11.534292221069336,
+      "eval_runtime": 361.6192,
+      "eval_samples_per_second": 2.765,
+      "eval_steps_per_second": 0.346,
+      "step": 840
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 5.250723838806152,
+      "learning_rate": 8.569908029550686e-07,
+      "logits/chosen": -0.006854387000203133,
+      "logits/rejected": -0.018336206674575806,
+      "logps/chosen": -0.6238930821418762,
+      "logps/rejected": -60.925689697265625,
+      "loss": 0.1157,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.443281173706055,
+      "rewards/margins": 6.164813995361328,
+      "rewards/rejected": 11.278467178344727,
+      "step": 850
+    },
+    {
+      "epoch": 0.7644444444444445,
+      "grad_norm": 3.1401162147521973,
+      "learning_rate": 7.992972643121227e-07,
+      "logits/chosen": -0.0037835021503269672,
+      "logits/rejected": -0.013135241344571114,
+      "logps/chosen": -0.8492221832275391,
+      "logps/rejected": -55.516075134277344,
+      "loss": 0.2252,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.488588333129883,
+      "rewards/margins": 5.736725807189941,
+      "rewards/rejected": 11.751862525939941,
+      "step": 860
+    },
+    {
+      "epoch": 0.7733333333333333,
+      "grad_norm": 55.528812408447266,
+      "learning_rate": 7.432425494343509e-07,
+      "logits/chosen": -0.0033687639515846968,
+      "logits/rejected": -0.013152632862329483,
+      "logps/chosen": -1.3188884258270264,
+      "logps/rejected": -57.9510498046875,
+      "loss": 0.1398,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.412578582763672,
+      "rewards/margins": 5.868515968322754,
+      "rewards/rejected": 11.544061660766602,
+      "step": 870
+    },
+    {
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.039824869483709335,
+      "learning_rate": 6.888806735220396e-07,
+      "logits/chosen": -0.0010406378423795104,
+      "logits/rejected": -0.012095071375370026,
+      "logps/chosen": -2.0619027614593506,
+      "logps/rejected": -59.65806579589844,
+      "loss": 0.2966,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.192535400390625,
+      "rewards/margins": 5.7006731033325195,
+      "rewards/rejected": 11.491861343383789,
+      "step": 880
+    },
+    {
+      "epoch": 0.7911111111111111,
+      "grad_norm": 0.7536466717720032,
+      "learning_rate": 6.362640205293583e-07,
+      "logits/chosen": -0.0016857212176546454,
+      "logits/rejected": -0.010936147533357143,
+      "logps/chosen": -2.1478958129882812,
+      "logps/rejected": -58.20386505126953,
+      "loss": 0.1641,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.243247985839844,
+      "rewards/margins": 5.678771018981934,
+      "rewards/rejected": 11.564477920532227,
+      "step": 890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3857377767562866,
+      "learning_rate": 5.854432926863684e-07,
+      "logits/chosen": 0.00038508616853505373,
+      "logits/rejected": -0.011034643277525902,
+      "logps/chosen": -1.606274127960205,
+      "logps/rejected": -62.58662414550781,
+      "loss": 0.1703,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.211261749267578,
+      "rewards/margins": 5.966868877410889,
+      "rewards/rejected": 11.244392395019531,
+      "step": 900
+    },
+    {
+      "epoch": 0.8,
+      "eval_logits/chosen": 0.00021816430671606213,
+      "eval_logits/rejected": -0.010477552190423012,
+      "eval_logps/chosen": -1.7576563358306885,
+      "eval_logps/rejected": -59.79063415527344,
+      "eval_loss": 0.13995186984539032,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.327129364013672,
+      "eval_rewards/margins": 5.945469856262207,
+      "eval_rewards/rejected": 11.381658554077148,
+      "eval_runtime": 361.4592,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 900
+    },
+    {
+      "epoch": 0.8088888888888889,
+      "grad_norm": 0.13143697381019592,
+      "learning_rate": 5.364674616415547e-07,
+      "logits/chosen": 0.0005570838693529367,
+      "logits/rejected": -0.011198626831173897,
+      "logps/chosen": -0.12791283428668976,
+      "logps/rejected": -65.7052993774414,
+      "loss": 0.0507,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.224443435668945,
+      "rewards/margins": 6.163690567016602,
+      "rewards/rejected": 11.060752868652344,
+      "step": 910
+    },
+    {
+      "epoch": 0.8177777777777778,
+      "grad_norm": 102.9654541015625,
+      "learning_rate": 4.893837212719859e-07,
+      "logits/chosen": -0.0008557128603570163,
+      "logits/rejected": -0.01163212489336729,
+      "logps/chosen": -1.3292646408081055,
+      "logps/rejected": -61.644893646240234,
+      "loss": 0.0881,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.276538848876953,
+      "rewards/margins": 5.96080207824707,
+      "rewards/rejected": 11.3157377243042,
+      "step": 920
+    },
+    {
+      "epoch": 0.8266666666666667,
+      "grad_norm": 146.04498291015625,
+      "learning_rate": 4.442374422065493e-07,
+      "logits/chosen": 0.002922601066529751,
+      "logits/rejected": -0.007130052894353867,
+      "logps/chosen": -1.6069023609161377,
+      "logps/rejected": -61.52588653564453,
+      "loss": 0.1373,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.184974670410156,
+      "rewards/margins": 5.7995734214782715,
+      "rewards/rejected": 11.38540267944336,
+      "step": 930
+    },
+    {
+      "epoch": 0.8355555555555556,
+      "grad_norm": 23.539485931396484,
+      "learning_rate": 4.0107212810610974e-07,
+      "logits/chosen": 0.0018056132830679417,
+      "logits/rejected": -0.007847340777516365,
+      "logps/chosen": -3.637047290802002,
+      "logps/rejected": -61.21245574951172,
+      "loss": 0.2763,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.960758209228516,
+      "rewards/margins": 5.551811695098877,
+      "rewards/rejected": 11.40894603729248,
+      "step": 940
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 31.52926254272461,
+      "learning_rate": 3.599293737426932e-07,
+      "logits/chosen": 0.0032081177923828363,
+      "logits/rejected": -0.007756482809782028,
+      "logps/chosen": -1.172515869140625,
+      "logps/rejected": -66.5853271484375,
+      "loss": 0.1758,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 16.992876052856445,
+      "rewards/margins": 5.886469841003418,
+      "rewards/rejected": 11.106407165527344,
+      "step": 950
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.07474468648433685,
+      "learning_rate": 3.208488249181216e-07,
+      "logits/chosen": 0.0022508346009999514,
+      "logits/rejected": -0.009156409651041031,
+      "logps/chosen": -0.7930470108985901,
+      "logps/rejected": -60.068790435791016,
+      "loss": 0.1138,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.401771545410156,
+      "rewards/margins": 6.0026960372924805,
+      "rewards/rejected": 11.399076461791992,
+      "step": 960
+    },
+    {
+      "epoch": 0.8533333333333334,
+      "eval_logits/chosen": 0.001469604205340147,
+      "eval_logits/rejected": -0.009397665038704872,
+      "eval_logps/chosen": -1.8795456886291504,
+      "eval_logps/rejected": -60.17564010620117,
+      "eval_loss": 0.1441129744052887,
+      "eval_rewards/accuracies": 0.9630000591278076,
+      "eval_rewards/chosen": 17.314937591552734,
+      "eval_rewards/margins": 5.9717817306518555,
+      "eval_rewards/rejected": 11.343156814575195,
+      "eval_runtime": 361.5344,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 960
+    },
+    {
+      "epoch": 0.8622222222222222,
+      "grad_norm": 68.90747833251953,
+      "learning_rate": 2.838681402606952e-07,
+      "logits/chosen": 0.004552370868623257,
+      "logits/rejected": -0.005488495342433453,
+      "logps/chosen": -3.7298974990844727,
+      "logps/rejected": -64.72488403320312,
+      "loss": 0.2425,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 16.832754135131836,
+      "rewards/margins": 5.644216537475586,
+      "rewards/rejected": 11.18853759765625,
+      "step": 970
+    },
+    {
+      "epoch": 0.8711111111111111,
+      "grad_norm": 0.028206102550029755,
+      "learning_rate": 2.490229549367443e-07,
+      "logits/chosen": 0.0025807656347751617,
+      "logits/rejected": -0.008657123893499374,
+      "logps/chosen": -1.1825838088989258,
+      "logps/rejected": -63.57493591308594,
+      "loss": 0.0609,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.227130889892578,
+      "rewards/margins": 6.053717136383057,
+      "rewards/rejected": 11.173412322998047,
+      "step": 980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39742231369018555,
+      "learning_rate": 2.1634684631203412e-07,
+      "logits/chosen": 0.0048486413434147835,
+      "logits/rejected": -0.006056814920157194,
+      "logps/chosen": -2.919680595397949,
+      "logps/rejected": -57.8403205871582,
+      "loss": 0.3464,
+      "rewards/accuracies": 0.8999999761581421,
+      "rewards/chosen": 17.175275802612305,
+      "rewards/margins": 5.559727668762207,
+      "rewards/rejected": 11.615548133850098,
+      "step": 990
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 30.22509002685547,
+      "learning_rate": 1.8587130159608196e-07,
+      "logits/chosen": 0.0050649940967559814,
+      "logits/rejected": -0.0071399761363863945,
+      "logps/chosen": -0.16131475567817688,
+      "logps/rejected": -66.25190734863281,
+      "loss": 0.0063,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": 17.313472747802734,
+      "rewards/margins": 6.412895679473877,
+      "rewards/rejected": 10.9005765914917,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8977777777777778,
+      "grad_norm": 53.02751922607422,
+      "learning_rate": 1.5762568750059604e-07,
+      "logits/chosen": 0.005312003195285797,
+      "logits/rejected": -0.0038596936501562595,
+      "logps/chosen": -4.00323486328125,
+      "logps/rejected": -57.656890869140625,
+      "loss": 0.2741,
+      "rewards/accuracies": 0.925000011920929,
+      "rewards/chosen": 17.167781829833984,
+      "rewards/margins": 5.632095813751221,
+      "rewards/rejected": 11.535685539245605,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "grad_norm": 119.56330108642578,
+      "learning_rate": 1.316372219412454e-07,
+      "logits/chosen": 0.004333779215812683,
+      "logits/rejected": -0.007937717251479626,
+      "logps/chosen": -0.10221245884895325,
+      "logps/rejected": -67.09977722167969,
+      "loss": 0.0513,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.151248931884766,
+      "rewards/margins": 6.162137985229492,
+      "rewards/rejected": 10.98911190032959,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9066666666666666,
+      "eval_logits/chosen": 0.004528110846877098,
+      "eval_logits/rejected": -0.006501312367618084,
+      "eval_logps/chosen": -1.817779302597046,
+      "eval_logps/rejected": -60.34454345703125,
+      "eval_loss": 0.14118175208568573,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.321117401123047,
+      "eval_rewards/margins": 5.994848251342773,
+      "eval_rewards/rejected": 11.32626724243164,
+      "eval_runtime": 361.545,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 0.346,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9155555555555556,
+      "grad_norm": 0.9030271768569946,
+      "learning_rate": 1.0793094781005792e-07,
+      "logits/chosen": 0.0036234352737665176,
+      "logits/rejected": -0.008390933275222778,
+      "logps/chosen": -0.09899584949016571,
+      "logps/rejected": -60.80555725097656,
+      "loss": 0.0349,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.619266510009766,
+      "rewards/margins": 6.4814043045043945,
+      "rewards/rejected": 11.137863159179688,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9244444444444444,
+      "grad_norm": 157.070068359375,
+      "learning_rate": 8.652970884369255e-08,
+      "logits/chosen": 0.005321115255355835,
+      "logits/rejected": -0.004976513795554638,
+      "logps/chosen": -1.3535833358764648,
+      "logps/rejected": -61.2861328125,
+      "loss": 0.1927,
+      "rewards/accuracies": 0.9500000476837158,
+      "rewards/chosen": 17.209800720214844,
+      "rewards/margins": 5.795952796936035,
+      "rewards/rejected": 11.413846015930176,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 24.539953231811523,
+      "learning_rate": 6.745412761086007e-08,
+      "logits/chosen": 0.005845514126121998,
+      "logits/rejected": -0.003921338357031345,
+      "logps/chosen": -1.7223193645477295,
+      "logps/rejected": -58.172142028808594,
+      "loss": 0.1317,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.360435485839844,
+      "rewards/margins": 5.838218688964844,
+      "rewards/rejected": 11.522216796875,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9422222222222222,
+      "grad_norm": 0.061458222568035126,
+      "learning_rate": 5.0722585640090305e-08,
+      "logits/chosen": 0.003759522922337055,
+      "logits/rejected": -0.0068417866714298725,
+      "logps/chosen": -1.351855754852295,
+      "logps/rejected": -59.56566619873047,
+      "loss": 0.2279,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 17.245838165283203,
+      "rewards/margins": 5.7038679122924805,
+      "rewards/rejected": 11.541970252990723,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9511111111111111,
+      "grad_norm": 101.87332916259766,
+      "learning_rate": 3.635120570700784e-08,
+      "logits/chosen": 0.007607857696712017,
+      "logits/rejected": -0.0035320711322128773,
+      "logps/chosen": -1.4004669189453125,
+      "logps/rejected": -63.46752166748047,
+      "loss": 0.1021,
+      "rewards/accuracies": 0.9624999761581421,
+      "rewards/chosen": 17.240036010742188,
+      "rewards/margins": 6.078882694244385,
+      "rewards/rejected": 11.161155700683594,
+      "step": 1070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9229660034179688,
+      "learning_rate": 2.4353836298169343e-08,
+      "logits/chosen": 0.006209026090800762,
+      "logits/rejected": -0.0033816141076385975,
+      "logps/chosen": -2.0281822681427,
+      "logps/rejected": -61.460670471191406,
+      "loss": 0.1189,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.19972801208496,
+      "rewards/margins": 5.881450653076172,
+      "rewards/rejected": 11.318277359008789,
+      "step": 1080
+    },
+    {
+      "epoch": 0.96,
+      "eval_logits/chosen": 0.00737445754930377,
+      "eval_logits/rejected": -0.0035832570865750313,
+      "eval_logps/chosen": -2.1419789791107178,
+      "eval_logps/rejected": -60.60612106323242,
+      "eval_loss": 0.15081512928009033,
+      "eval_rewards/accuracies": 0.9610000848770142,
+      "eval_rewards/chosen": 17.288694381713867,
+      "eval_rewards/margins": 5.98858642578125,
+      "eval_rewards/rejected": 11.300108909606934,
+      "eval_runtime": 361.4165,
+      "eval_samples_per_second": 2.767,
+      "eval_steps_per_second": 0.346,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9688888888888889,
+      "grad_norm": 0.11888863146305084,
+      "learning_rate": 1.4742038266447046e-08,
+      "logits/chosen": 0.006043245084583759,
+      "logits/rejected": -0.004712546244263649,
+      "logps/chosen": -1.4127472639083862,
+      "logps/rejected": -64.81529998779297,
+      "loss": 0.0728,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.181049346923828,
+      "rewards/margins": 6.14100456237793,
+      "rewards/rejected": 11.040044784545898,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 2.001819372177124,
+      "learning_rate": 7.525073690809737e-09,
+      "logits/chosen": 0.0055408780463039875,
+      "logits/rejected": -0.005598037503659725,
+      "logps/chosen": -1.017110824584961,
+      "logps/rejected": -60.28044891357422,
+      "loss": 0.0357,
+      "rewards/accuracies": 0.9750000238418579,
+      "rewards/chosen": 17.4679012298584,
+      "rewards/margins": 6.193048477172852,
+      "rewards/rejected": 11.274852752685547,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9866666666666667,
+      "grad_norm": 0.027056939899921417,
+      "learning_rate": 2.709896951238744e-09,
+      "logits/chosen": 0.00636716466397047,
+      "logits/rejected": -0.004360577557235956,
+      "logps/chosen": -2.8028905391693115,
+      "logps/rejected": -60.86518859863281,
+      "loss": 0.1973,
+      "rewards/accuracies": 0.949999988079071,
+      "rewards/chosen": 17.210033416748047,
+      "rewards/margins": 5.936794281005859,
+      "rewards/rejected": 11.273238182067871,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9955555555555555,
+      "grad_norm": 0.15637506544589996,
+      "learning_rate": 3.0114802737818415e-10,
+      "logits/chosen": 0.0038183885626494884,
+      "logits/rejected": -0.007296917960047722,
+      "logps/chosen": -1.1477452516555786,
+      "logps/rejected": -59.37739181518555,
+      "loss": 0.0538,
+      "rewards/accuracies": 0.987500011920929,
+      "rewards/chosen": 17.553251266479492,
+      "rewards/margins": 6.2790141105651855,
+      "rewards/rejected": 11.274236679077148,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0,
+      "step": 1125,
+      "total_flos": 1.4338459346927616e+18,
+      "train_loss": 0.23597783709896936,
+      "train_runtime": 13850.7044,
+      "train_samples_per_second": 0.65,
+      "train_steps_per_second": 0.081
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1125,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 1.4338459346927616e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}