Training in progress, step 100, checkpoint

.gitattributes

@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text
+last-checkpoint/tokenizer.json filter=lfs diff=lfs merge=lfs -text

last-checkpoint/README.md

@@ -0,0 +1,202 @@
+---
+base_model: Qwen/Qwen1.5-7B
+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 0.13.2

last-checkpoint/adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen1.5-7B",
+  "bias": "none",
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 128,
+  "lora_dropout": 0.1,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "o_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

last-checkpoint/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ce02f701af666e8574f34bbece43cf678e6c43828c6401013c7adaee7f3e44c
+size 639691872

last-checkpoint/added_tokens.json

@@ -0,0 +1,5 @@
+{
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644
+}

last-checkpoint/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8575e904c1cf22935656eef06768f9eada06da0066a043880265c8ccc56d8ae8
+size 1279529146

last-checkpoint/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:386fcc8cc1089aade9450d86fb239ea3483f455fd2d78d8378645feecfec9d69
+size 14244

last-checkpoint/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46c96054ea697f2f09206927300764c4f5f88c9d18ac53664cf1661a0b6f2665
+size 1064

last-checkpoint/special_tokens_map.json

@@ -0,0 +1,20 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

last-checkpoint/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcfe42da0a4497e8b2b172c1f9f4ec423a46dc12907f4349c55025f670422ba9
+size 11418266

last-checkpoint/tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

last-checkpoint/trainer_state.json

@@ -0,0 +1,1533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.02502111156288118,
+  "eval_steps": 500,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00025021111562881183,
+      "grad_norm": 170.68319702148438,
+      "learning_rate": 2.5000000000000002e-08,
+      "logits/chosen": 0.7628318667411804,
+      "logits/rejected": 0.6978355646133423,
+      "logps/chosen": -799.0330200195312,
+      "logps/rejected": -820.4984741210938,
+      "loss": 5.5452,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005004222312576237,
+      "grad_norm": 165.87217712402344,
+      "learning_rate": 5.0000000000000004e-08,
+      "logits/chosen": 0.5875029563903809,
+      "logits/rejected": 0.4390100836753845,
+      "logps/chosen": -864.8237915039062,
+      "logps/rejected": -952.4483032226562,
+      "loss": 5.5452,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007506333468864354,
+      "grad_norm": 180.83462524414062,
+      "learning_rate": 7.500000000000001e-08,
+      "logits/chosen": 1.1111327409744263,
+      "logits/rejected": 1.1068625450134277,
+      "logps/chosen": -790.4888305664062,
+      "logps/rejected": -749.4112548828125,
+      "loss": 5.7013,
+      "rewards/accuracies": 0.25,
+      "rewards/chosen": -0.035940174013376236,
+      "rewards/margins": -0.035558320581912994,
+      "rewards/rejected": -0.00038185110315680504,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010008444625152473,
+      "grad_norm": 181.13485717773438,
+      "learning_rate": 1.0000000000000001e-07,
+      "logits/chosen": 0.5954170823097229,
+      "logits/rejected": 0.7202694416046143,
+      "logps/chosen": -881.8717041015625,
+      "logps/rejected": -991.069580078125,
+      "loss": 5.4935,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.0166747085750103,
+      "rewards/margins": 0.017112253233790398,
+      "rewards/rejected": -0.00043754721991717815,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012510555781440592,
+      "grad_norm": 190.6446990966797,
+      "learning_rate": 1.2500000000000002e-07,
+      "logits/chosen": 0.6433669328689575,
+      "logits/rejected": 0.6477421522140503,
+      "logps/chosen": -873.276123046875,
+      "logps/rejected": -860.74609375,
+      "loss": 5.5934,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.004312766715884209,
+      "rewards/margins": -0.008207356557250023,
+      "rewards/rejected": 0.012520124204456806,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015012666937728708,
+      "grad_norm": 165.24261474609375,
+      "learning_rate": 1.5000000000000002e-07,
+      "logits/chosen": 0.3789970576763153,
+      "logits/rejected": 0.44303324818611145,
+      "logps/chosen": -779.4014282226562,
+      "logps/rejected": -820.912109375,
+      "loss": 5.6292,
+      "rewards/accuracies": 0.3125,
+      "rewards/chosen": 0.00436730682849884,
+      "rewards/margins": -0.01688399538397789,
+      "rewards/rejected": 0.021251298487186432,
+      "step": 6
+    },
+    {
+      "epoch": 0.0017514778094016826,
+      "grad_norm": 167.0378875732422,
+      "learning_rate": 1.7500000000000002e-07,
+      "logits/chosen": 0.5348235368728638,
+      "logits/rejected": 0.42912158370018005,
+      "logps/chosen": -819.9292602539062,
+      "logps/rejected": -866.7522583007812,
+      "loss": 5.7262,
+      "rewards/accuracies": 0.375,
+      "rewards/chosen": -0.04971008747816086,
+      "rewards/margins": -0.04176750034093857,
+      "rewards/rejected": -0.007942580617964268,
+      "step": 7
+    },
+    {
+      "epoch": 0.0020016889250304946,
+      "grad_norm": 202.10984802246094,
+      "learning_rate": 2.0000000000000002e-07,
+      "logits/chosen": 0.6080639362335205,
+      "logits/rejected": 0.6230067014694214,
+      "logps/chosen": -845.4388427734375,
+      "logps/rejected": -896.5678100585938,
+      "loss": 5.7521,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": -0.041010383516550064,
+      "rewards/margins": -0.045076847076416016,
+      "rewards/rejected": 0.004066466819494963,
+      "step": 8
+    },
+    {
+      "epoch": 0.0022519000406593065,
+      "grad_norm": 189.44422912597656,
+      "learning_rate": 2.2500000000000002e-07,
+      "logits/chosen": 0.6168553829193115,
+      "logits/rejected": 0.8411935567855835,
+      "logps/chosen": -890.399658203125,
+      "logps/rejected": -847.189697265625,
+      "loss": 5.7522,
+      "rewards/accuracies": 0.3125,
+      "rewards/chosen": -0.056749649345874786,
+      "rewards/margins": -0.04940682649612427,
+      "rewards/rejected": -0.0073428284376859665,
+      "step": 9
+    },
+    {
+      "epoch": 0.0025021111562881183,
+      "grad_norm": 198.04713439941406,
+      "learning_rate": 2.5000000000000004e-07,
+      "logits/chosen": 0.27231016755104065,
+      "logits/rejected": 0.21679289638996124,
+      "logps/chosen": -975.139892578125,
+      "logps/rejected": -1047.917236328125,
+      "loss": 5.4884,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.018453503027558327,
+      "rewards/margins": 0.020519159734249115,
+      "rewards/rejected": -0.002065659500658512,
+      "step": 10
+    },
+    {
+      "epoch": 0.0027523222719169297,
+      "grad_norm": 162.54795837402344,
+      "learning_rate": 2.75e-07,
+      "logits/chosen": 0.4516029357910156,
+      "logits/rejected": 0.48896458745002747,
+      "logps/chosen": -911.2274780273438,
+      "logps/rejected": -832.1043090820312,
+      "loss": 5.4267,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.013639260083436966,
+      "rewards/margins": 0.03403634950518608,
+      "rewards/rejected": -0.020397093147039413,
+      "step": 11
+    },
+    {
+      "epoch": 0.0030025333875457415,
+      "grad_norm": 168.70565795898438,
+      "learning_rate": 3.0000000000000004e-07,
+      "logits/chosen": 0.4096769690513611,
+      "logits/rejected": 0.3834303617477417,
+      "logps/chosen": -761.2208862304688,
+      "logps/rejected": -731.2702026367188,
+      "loss": 5.7227,
+      "rewards/accuracies": 0.25,
+      "rewards/chosen": -0.032328225672245026,
+      "rewards/margins": -0.04112980514764786,
+      "rewards/rejected": 0.008801579475402832,
+      "step": 12
+    },
+    {
+      "epoch": 0.0032527445031745534,
+      "grad_norm": 211.1651611328125,
+      "learning_rate": 3.25e-07,
+      "logits/chosen": 0.7261509299278259,
+      "logits/rejected": 0.712794840335846,
+      "logps/chosen": -875.9140625,
+      "logps/rejected": -837.6126708984375,
+      "loss": 5.7022,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": -0.04412355646491051,
+      "rewards/margins": -0.034722425043582916,
+      "rewards/rejected": -0.009401128627359867,
+      "step": 13
+    },
+    {
+      "epoch": 0.003502955618803365,
+      "grad_norm": 158.79537963867188,
+      "learning_rate": 3.5000000000000004e-07,
+      "logits/chosen": 0.721327006816864,
+      "logits/rejected": 0.882017970085144,
+      "logps/chosen": -719.6497802734375,
+      "logps/rejected": -699.1884765625,
+      "loss": 5.7616,
+      "rewards/accuracies": 0.3125,
+      "rewards/chosen": -0.003787088207900524,
+      "rewards/margins": -0.05063605308532715,
+      "rewards/rejected": 0.04684896767139435,
+      "step": 14
+    },
+    {
+      "epoch": 0.003753166734432177,
+      "grad_norm": 205.3097686767578,
+      "learning_rate": 3.75e-07,
+      "logits/chosen": 0.5128509998321533,
+      "logits/rejected": 0.4282974600791931,
+      "logps/chosen": -1061.45751953125,
+      "logps/rejected": -1026.2188720703125,
+      "loss": 5.4322,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.019785691052675247,
+      "rewards/margins": 0.03162631765007973,
+      "rewards/rejected": -0.011840627528727055,
+      "step": 15
+    },
+    {
+      "epoch": 0.004003377850060989,
+      "grad_norm": 184.5326690673828,
+      "learning_rate": 4.0000000000000003e-07,
+      "logits/chosen": 0.6119356751441956,
+      "logits/rejected": 0.6239423751831055,
+      "logps/chosen": -995.9874877929688,
+      "logps/rejected": -1025.8966064453125,
+      "loss": 5.4171,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.00705108605325222,
+      "rewards/margins": 0.039296723902225494,
+      "rewards/rejected": -0.032245635986328125,
+      "step": 16
+    },
+    {
+      "epoch": 0.004253588965689801,
+      "grad_norm": 179.45584106445312,
+      "learning_rate": 4.2500000000000006e-07,
+      "logits/chosen": 0.8823148012161255,
+      "logits/rejected": 0.8341612219810486,
+      "logps/chosen": -890.3267211914062,
+      "logps/rejected": -970.2156982421875,
+      "loss": 5.4777,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.03143415227532387,
+      "rewards/margins": 0.02045259438455105,
+      "rewards/rejected": 0.01098155789077282,
+      "step": 17
+    },
+    {
+      "epoch": 0.004503800081318613,
+      "grad_norm": 191.9487762451172,
+      "learning_rate": 4.5000000000000003e-07,
+      "logits/chosen": 0.6515500545501709,
+      "logits/rejected": 0.5901349782943726,
+      "logps/chosen": -1176.125732421875,
+      "logps/rejected": -1147.20068359375,
+      "loss": 5.4751,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.008896443992853165,
+      "rewards/margins": 0.02563781850039959,
+      "rewards/rejected": -0.016741372644901276,
+      "step": 18
+    },
+    {
+      "epoch": 0.004754011196947424,
+      "grad_norm": 155.9108428955078,
+      "learning_rate": 4.7500000000000006e-07,
+      "logits/chosen": 0.9351687431335449,
+      "logits/rejected": 0.9427083134651184,
+      "logps/chosen": -660.8029174804688,
+      "logps/rejected": -643.7144775390625,
+      "loss": 5.489,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.016555000096559525,
+      "rewards/margins": 0.018336273729801178,
+      "rewards/rejected": -0.0017812740989029408,
+      "step": 19
+    },
+    {
+      "epoch": 0.005004222312576237,
+      "grad_norm": 177.9388885498047,
+      "learning_rate": 5.000000000000001e-07,
+      "logits/chosen": 0.4800458550453186,
+      "logits/rejected": 0.5687589645385742,
+      "logps/chosen": -842.927490234375,
+      "logps/rejected": -914.377685546875,
+      "loss": 5.3652,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.03765854984521866,
+      "rewards/margins": 0.04844312369823456,
+      "rewards/rejected": -0.010784578509628773,
+      "step": 20
+    },
+    {
+      "epoch": 0.005254433428205048,
+      "grad_norm": 206.02239990234375,
+      "learning_rate": 5.250000000000001e-07,
+      "logits/chosen": 0.37655341625213623,
+      "logits/rejected": 0.4930800199508667,
+      "logps/chosen": -1135.379638671875,
+      "logps/rejected": -1079.0499267578125,
+      "loss": 5.6328,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": -0.011679746210575104,
+      "rewards/margins": -0.01775560900568962,
+      "rewards/rejected": 0.006075860001146793,
+      "step": 21
+    },
+    {
+      "epoch": 0.005504644543833859,
+      "grad_norm": 185.9024658203125,
+      "learning_rate": 5.5e-07,
+      "logits/chosen": 0.6661480069160461,
+      "logits/rejected": 0.7582310438156128,
+      "logps/chosen": -750.8116455078125,
+      "logps/rejected": -776.9088134765625,
+      "loss": 5.5751,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.013671351596713066,
+      "rewards/margins": -0.001869439147412777,
+      "rewards/rejected": 0.015540791675448418,
+      "step": 22
+    },
+    {
+      "epoch": 0.005754855659462672,
+      "grad_norm": 165.102783203125,
+      "learning_rate": 5.750000000000001e-07,
+      "logits/chosen": 0.9347951412200928,
+      "logits/rejected": 1.106078028678894,
+      "logps/chosen": -645.0971069335938,
+      "logps/rejected": -665.1268920898438,
+      "loss": 5.6054,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.01613318920135498,
+      "rewards/margins": -0.010771606117486954,
+      "rewards/rejected": 0.026904797181487083,
+      "step": 23
+    },
+    {
+      "epoch": 0.006005066775091483,
+      "grad_norm": 197.385498046875,
+      "learning_rate": 6.000000000000001e-07,
+      "logits/chosen": 0.48968541622161865,
+      "logits/rejected": 0.23404461145401,
+      "logps/chosen": -975.5427856445312,
+      "logps/rejected": -1082.59716796875,
+      "loss": 5.5153,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.04105112701654434,
+      "rewards/margins": 0.01256849616765976,
+      "rewards/rejected": 0.028482627123594284,
+      "step": 24
+    },
+    {
+      "epoch": 0.006255277890720295,
+      "grad_norm": 173.0858154296875,
+      "learning_rate": 6.25e-07,
+      "logits/chosen": 0.8860877752304077,
+      "logits/rejected": 0.8948439359664917,
+      "logps/chosen": -740.6904296875,
+      "logps/rejected": -742.9269409179688,
+      "loss": 5.5386,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.04100323095917702,
+      "rewards/margins": 0.008243849501013756,
+      "rewards/rejected": 0.03275937959551811,
+      "step": 25
+    },
+    {
+      "epoch": 0.006505489006349107,
+      "grad_norm": 188.4302215576172,
+      "learning_rate": 6.5e-07,
+      "logits/chosen": 0.39670196175575256,
+      "logits/rejected": 0.5695019364356995,
+      "logps/chosen": -1109.718994140625,
+      "logps/rejected": -965.9251708984375,
+      "loss": 5.4071,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": 0.025699710473418236,
+      "rewards/margins": 0.04134707525372505,
+      "rewards/rejected": -0.015647364780306816,
+      "step": 26
+    },
+    {
+      "epoch": 0.006755700121977919,
+      "grad_norm": 176.55296325683594,
+      "learning_rate": 6.750000000000001e-07,
+      "logits/chosen": 0.9913475513458252,
+      "logits/rejected": 0.9675697088241577,
+      "logps/chosen": -826.3568725585938,
+      "logps/rejected": -863.7825927734375,
+      "loss": 5.5437,
+      "rewards/accuracies": 0.375,
+      "rewards/chosen": 0.002540584886446595,
+      "rewards/margins": 0.0020542130805552006,
+      "rewards/rejected": 0.0004863748326897621,
+      "step": 27
+    },
+    {
+      "epoch": 0.00700591123760673,
+      "grad_norm": 171.95623779296875,
+      "learning_rate": 7.000000000000001e-07,
+      "logits/chosen": 0.9001889824867249,
+      "logits/rejected": 0.728625476360321,
+      "logps/chosen": -705.9625854492188,
+      "logps/rejected": -679.670166015625,
+      "loss": 5.3871,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.039113618433475494,
+      "rewards/margins": 0.04316110908985138,
+      "rewards/rejected": -0.004047489259392023,
+      "step": 28
+    },
+    {
+      "epoch": 0.007256122353235543,
+      "grad_norm": 182.76560974121094,
+      "learning_rate": 7.25e-07,
+      "logits/chosen": 0.6934616565704346,
+      "logits/rejected": 0.5715526938438416,
+      "logps/chosen": -958.8621215820312,
+      "logps/rejected": -1024.5343017578125,
+      "loss": 5.4355,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.04835329204797745,
+      "rewards/margins": 0.03557805344462395,
+      "rewards/rejected": 0.012775231152772903,
+      "step": 29
+    },
+    {
+      "epoch": 0.007506333468864354,
+      "grad_norm": 203.4325408935547,
+      "learning_rate": 7.5e-07,
+      "logits/chosen": 0.49103742837905884,
+      "logits/rejected": 0.16804730892181396,
+      "logps/chosen": -917.3203735351562,
+      "logps/rejected": -1026.76171875,
+      "loss": 5.4038,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.007693195715546608,
+      "rewards/margins": 0.04031064733862877,
+      "rewards/rejected": -0.03261745348572731,
+      "step": 30
+    },
+    {
+      "epoch": 0.007756544584493166,
+      "grad_norm": 168.09117126464844,
+      "learning_rate": 7.750000000000001e-07,
+      "logits/chosen": 0.6098487377166748,
+      "logits/rejected": 0.7861579656600952,
+      "logps/chosen": -730.9053344726562,
+      "logps/rejected": -774.8155517578125,
+      "loss": 5.3484,
+      "rewards/accuracies": 0.8125,
+      "rewards/chosen": 0.021628666669130325,
+      "rewards/margins": 0.05140829086303711,
+      "rewards/rejected": -0.029779626056551933,
+      "step": 31
+    },
+    {
+      "epoch": 0.008006755700121979,
+      "grad_norm": 205.30642700195312,
+      "learning_rate": 8.000000000000001e-07,
+      "logits/chosen": 0.7579723596572876,
+      "logits/rejected": 0.9500037431716919,
+      "logps/chosen": -919.2657470703125,
+      "logps/rejected": -679.0538330078125,
+      "loss": 5.6396,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.006780529394745827,
+      "rewards/margins": -0.017520543187856674,
+      "rewards/rejected": 0.0243010763078928,
+      "step": 32
+    },
+    {
+      "epoch": 0.008256966815750789,
+      "grad_norm": 162.000732421875,
+      "learning_rate": 8.250000000000001e-07,
+      "logits/chosen": 0.7313171029090881,
+      "logits/rejected": 1.0325978994369507,
+      "logps/chosen": -763.1948852539062,
+      "logps/rejected": -727.3697509765625,
+      "loss": 5.5084,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.010264062322676182,
+      "rewards/margins": 0.013461539521813393,
+      "rewards/rejected": -0.00319747906178236,
+      "step": 33
+    },
+    {
+      "epoch": 0.008507177931379601,
+      "grad_norm": 175.06857299804688,
+      "learning_rate": 8.500000000000001e-07,
+      "logits/chosen": 0.6649048328399658,
+      "logits/rejected": 0.6554955244064331,
+      "logps/chosen": -856.1851806640625,
+      "logps/rejected": -801.0748291015625,
+      "loss": 5.6566,
+      "rewards/accuracies": 0.375,
+      "rewards/chosen": 0.006980515085160732,
+      "rewards/margins": -0.02326059527695179,
+      "rewards/rejected": 0.030241109430789948,
+      "step": 34
+    },
+    {
+      "epoch": 0.008757389047008414,
+      "grad_norm": 177.32518005371094,
+      "learning_rate": 8.75e-07,
+      "logits/chosen": 0.34936243295669556,
+      "logits/rejected": 0.42776912450790405,
+      "logps/chosen": -846.9901733398438,
+      "logps/rejected": -750.77392578125,
+      "loss": 5.5082,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.021079493686556816,
+      "rewards/margins": 0.013807770796120167,
+      "rewards/rejected": 0.007271719165146351,
+      "step": 35
+    },
+    {
+      "epoch": 0.009007600162637226,
+      "grad_norm": 199.04859924316406,
+      "learning_rate": 9.000000000000001e-07,
+      "logits/chosen": 0.6959156394004822,
+      "logits/rejected": 0.6144693493843079,
+      "logps/chosen": -941.1321411132812,
+      "logps/rejected": -962.2286987304688,
+      "loss": 5.5607,
+      "rewards/accuracies": 0.3125,
+      "rewards/chosen": 0.009562971070408821,
+      "rewards/margins": -0.0015952119138091803,
+      "rewards/rejected": 0.011158181354403496,
+      "step": 36
+    },
+    {
+      "epoch": 0.009257811278266036,
+      "grad_norm": 179.16551208496094,
+      "learning_rate": 9.25e-07,
+      "logits/chosen": 0.7614805698394775,
+      "logits/rejected": 0.7655118107795715,
+      "logps/chosen": -778.8309326171875,
+      "logps/rejected": -778.1507568359375,
+      "loss": 5.7471,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.0017023803666234016,
+      "rewards/margins": -0.04353148117661476,
+      "rewards/rejected": 0.04523386061191559,
+      "step": 37
+    },
+    {
+      "epoch": 0.009508022393894849,
+      "grad_norm": 172.88616943359375,
+      "learning_rate": 9.500000000000001e-07,
+      "logits/chosen": 0.43299418687820435,
+      "logits/rejected": 0.3225078284740448,
+      "logps/chosen": -901.8963012695312,
+      "logps/rejected": -913.7239990234375,
+      "loss": 5.4793,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.04441376030445099,
+      "rewards/margins": 0.020983794704079628,
+      "rewards/rejected": 0.02342996746301651,
+      "step": 38
+    },
+    {
+      "epoch": 0.009758233509523661,
+      "grad_norm": 156.6342315673828,
+      "learning_rate": 9.750000000000002e-07,
+      "logits/chosen": 0.7074395418167114,
+      "logits/rejected": 0.7963012456893921,
+      "logps/chosen": -675.8837890625,
+      "logps/rejected": -675.045654296875,
+      "loss": 5.6631,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.037299610674381256,
+      "rewards/margins": -0.025847911834716797,
+      "rewards/rejected": 0.06314751505851746,
+      "step": 39
+    },
+    {
+      "epoch": 0.010008444625152473,
+      "grad_norm": 184.23532104492188,
+      "learning_rate": 1.0000000000000002e-06,
+      "logits/chosen": 0.9331844449043274,
+      "logits/rejected": 0.8512438535690308,
+      "logps/chosen": -769.69482421875,
+      "logps/rejected": -836.7721557617188,
+      "loss": 5.7472,
+      "rewards/accuracies": 0.3125,
+      "rewards/chosen": -0.015815066173672676,
+      "rewards/margins": -0.04652342572808266,
+      "rewards/rejected": 0.03070836141705513,
+      "step": 40
+    },
+    {
+      "epoch": 0.010258655740781284,
+      "grad_norm": 178.39248657226562,
+      "learning_rate": 1.025e-06,
+      "logits/chosen": 1.0110334157943726,
+      "logits/rejected": 0.8221983909606934,
+      "logps/chosen": -727.4151611328125,
+      "logps/rejected": -740.2017822265625,
+      "loss": 5.7696,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": -0.01262521930038929,
+      "rewards/margins": -0.04993200674653053,
+      "rewards/rejected": 0.037306781858205795,
+      "step": 41
+    },
+    {
+      "epoch": 0.010508866856410096,
+      "grad_norm": 162.9761199951172,
+      "learning_rate": 1.0500000000000001e-06,
+      "logits/chosen": 0.5978808403015137,
+      "logits/rejected": 0.6464583873748779,
+      "logps/chosen": -785.3408203125,
+      "logps/rejected": -706.8924560546875,
+      "loss": 5.6026,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.028894497081637383,
+      "rewards/margins": -0.010278724133968353,
+      "rewards/rejected": 0.039173223078250885,
+      "step": 42
+    },
+    {
+      "epoch": 0.010759077972038908,
+      "grad_norm": 191.45413208007812,
+      "learning_rate": 1.075e-06,
+      "logits/chosen": 0.3623913526535034,
+      "logits/rejected": 0.4811972677707672,
+      "logps/chosen": -922.0300903320312,
+      "logps/rejected": -847.462646484375,
+      "loss": 5.5087,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.040454890578985214,
+      "rewards/margins": 0.016093183308839798,
+      "rewards/rejected": 0.024361703544855118,
+      "step": 43
+    },
+    {
+      "epoch": 0.011009289087667719,
+      "grad_norm": 156.37103271484375,
+      "learning_rate": 1.1e-06,
+      "logits/chosen": 0.8890228271484375,
+      "logits/rejected": 1.0089755058288574,
+      "logps/chosen": -613.6019287109375,
+      "logps/rejected": -764.8602294921875,
+      "loss": 5.485,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.06841964274644852,
+      "rewards/margins": 0.01816377602517605,
+      "rewards/rejected": 0.05025586858391762,
+      "step": 44
+    },
+    {
+      "epoch": 0.011259500203296531,
+      "grad_norm": 168.50357055664062,
+      "learning_rate": 1.125e-06,
+      "logits/chosen": 0.6311800479888916,
+      "logits/rejected": 0.51740962266922,
+      "logps/chosen": -670.5458374023438,
+      "logps/rejected": -749.3095703125,
+      "loss": 5.4536,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.05805638059973717,
+      "rewards/margins": 0.030191292986273766,
+      "rewards/rejected": 0.027865087613463402,
+      "step": 45
+    },
+    {
+      "epoch": 0.011509711318925343,
+      "grad_norm": 163.52545166015625,
+      "learning_rate": 1.1500000000000002e-06,
+      "logits/chosen": 0.5339171290397644,
+      "logits/rejected": 0.6589304208755493,
+      "logps/chosen": -849.0599365234375,
+      "logps/rejected": -787.2460327148438,
+      "loss": 5.395,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": 0.09280490130186081,
+      "rewards/margins": 0.04091768339276314,
+      "rewards/rejected": 0.05188722535967827,
+      "step": 46
+    },
+    {
+      "epoch": 0.011759922434554156,
+      "grad_norm": 175.00405883789062,
+      "learning_rate": 1.175e-06,
+      "logits/chosen": 0.6448819637298584,
+      "logits/rejected": 0.5809612274169922,
+      "logps/chosen": -872.1870727539062,
+      "logps/rejected": -852.7489013671875,
+      "loss": 5.4143,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.06790357083082199,
+      "rewards/margins": 0.03818032890558243,
+      "rewards/rejected": 0.029723238199949265,
+      "step": 47
+    },
+    {
+      "epoch": 0.012010133550182966,
+      "grad_norm": 213.84231567382812,
+      "learning_rate": 1.2000000000000002e-06,
+      "logits/chosen": 0.6562647223472595,
+      "logits/rejected": 0.5508157014846802,
+      "logps/chosen": -949.4231567382812,
+      "logps/rejected": -897.4124755859375,
+      "loss": 5.7273,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.05144808441400528,
+      "rewards/margins": -0.04000968486070633,
+      "rewards/rejected": 0.09145776927471161,
+      "step": 48
+    },
+    {
+      "epoch": 0.012260344665811778,
+      "grad_norm": 141.68014526367188,
+      "learning_rate": 1.2250000000000001e-06,
+      "logits/chosen": 1.1136417388916016,
+      "logits/rejected": 1.1333364248275757,
+      "logps/chosen": -519.7319946289062,
+      "logps/rejected": -594.5286865234375,
+      "loss": 5.5898,
+      "rewards/accuracies": 0.375,
+      "rewards/chosen": 0.08642894774675369,
+      "rewards/margins": -0.005818536505103111,
+      "rewards/rejected": 0.09224748611450195,
+      "step": 49
+    },
+    {
+      "epoch": 0.01251055578144059,
+      "grad_norm": 205.3349151611328,
+      "learning_rate": 1.25e-06,
+      "logits/chosen": 0.7967091798782349,
+      "logits/rejected": 0.7253572344779968,
+      "logps/chosen": -896.8436279296875,
+      "logps/rejected": -919.3460083007812,
+      "loss": 5.8913,
+      "rewards/accuracies": 0.375,
+      "rewards/chosen": 0.013068106025457382,
+      "rewards/margins": -0.07888536155223846,
+      "rewards/rejected": 0.09195347130298615,
+      "step": 50
+    },
+    {
+      "epoch": 0.012760766897069403,
+      "grad_norm": 181.29290771484375,
+      "learning_rate": 1.275e-06,
+      "logits/chosen": 0.514256477355957,
+      "logits/rejected": 0.28395047783851624,
+      "logps/chosen": -790.8712768554688,
+      "logps/rejected": -954.5174560546875,
+      "loss": 5.5535,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.06918986141681671,
+      "rewards/margins": 0.0045003220438957214,
+      "rewards/rejected": 0.06468954682350159,
+      "step": 51
+    },
+    {
+      "epoch": 0.013010978012698213,
+      "grad_norm": 186.6272735595703,
+      "learning_rate": 1.3e-06,
+      "logits/chosen": 0.6851955056190491,
+      "logits/rejected": 0.7106304168701172,
+      "logps/chosen": -874.3208618164062,
+      "logps/rejected": -860.9207763671875,
+      "loss": 5.3796,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.06201258301734924,
+      "rewards/margins": 0.04732148349285126,
+      "rewards/rejected": 0.014691092073917389,
+      "step": 52
+    },
+    {
+      "epoch": 0.013261189128327026,
+      "grad_norm": 157.80934143066406,
+      "learning_rate": 1.3250000000000002e-06,
+      "logits/chosen": 0.8752857446670532,
+      "logits/rejected": 1.064914584159851,
+      "logps/chosen": -615.7212524414062,
+      "logps/rejected": -567.2990112304688,
+      "loss": 5.4081,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.09691645205020905,
+      "rewards/margins": 0.03854731097817421,
+      "rewards/rejected": 0.05836913734674454,
+      "step": 53
+    },
+    {
+      "epoch": 0.013511400243955838,
+      "grad_norm": 178.67034912109375,
+      "learning_rate": 1.3500000000000002e-06,
+      "logits/chosen": 0.711281418800354,
+      "logits/rejected": 0.9150132536888123,
+      "logps/chosen": -859.6094360351562,
+      "logps/rejected": -774.975830078125,
+      "loss": 5.4574,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.10661716759204865,
+      "rewards/margins": 0.029118256643414497,
+      "rewards/rejected": 0.0774989128112793,
+      "step": 54
+    },
+    {
+      "epoch": 0.01376161135958465,
+      "grad_norm": 171.77268981933594,
+      "learning_rate": 1.3750000000000002e-06,
+      "logits/chosen": 0.4057692587375641,
+      "logits/rejected": 0.34246376156806946,
+      "logps/chosen": -969.0261840820312,
+      "logps/rejected": -971.955810546875,
+      "loss": 5.1942,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": 0.08421164751052856,
+      "rewards/margins": 0.09731664508581161,
+      "rewards/rejected": -0.01310500968247652,
+      "step": 55
+    },
+    {
+      "epoch": 0.01401182247521346,
+      "grad_norm": 181.45034790039062,
+      "learning_rate": 1.4000000000000001e-06,
+      "logits/chosen": 0.7378779649734497,
+      "logits/rejected": 0.6882550716400146,
+      "logps/chosen": -918.377685546875,
+      "logps/rejected": -950.7130737304688,
+      "loss": 5.4139,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.1188017874956131,
+      "rewards/margins": 0.03934526816010475,
+      "rewards/rejected": 0.07945652306079865,
+      "step": 56
+    },
+    {
+      "epoch": 0.014262033590842273,
+      "grad_norm": 184.7382049560547,
+      "learning_rate": 1.425e-06,
+      "logits/chosen": 0.7330397963523865,
+      "logits/rejected": 0.7624054551124573,
+      "logps/chosen": -737.2774658203125,
+      "logps/rejected": -834.9508666992188,
+      "loss": 5.3769,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.1038268432021141,
+      "rewards/margins": 0.05010106414556503,
+      "rewards/rejected": 0.05372576788067818,
+      "step": 57
+    },
+    {
+      "epoch": 0.014512244706471085,
+      "grad_norm": 177.06936645507812,
+      "learning_rate": 1.45e-06,
+      "logits/chosen": 0.9588703513145447,
+      "logits/rejected": 0.8182455897331238,
+      "logps/chosen": -761.6469116210938,
+      "logps/rejected": -787.0045776367188,
+      "loss": 5.4959,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.09994152933359146,
+      "rewards/margins": 0.01606077142059803,
+      "rewards/rejected": 0.08388075977563858,
+      "step": 58
+    },
+    {
+      "epoch": 0.014762455822099898,
+      "grad_norm": 148.8179473876953,
+      "learning_rate": 1.475e-06,
+      "logits/chosen": 0.7903648614883423,
+      "logits/rejected": 0.9569006562232971,
+      "logps/chosen": -744.4844360351562,
+      "logps/rejected": -699.2374877929688,
+      "loss": 5.1259,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.14937205612659454,
+      "rewards/margins": 0.12116777896881104,
+      "rewards/rejected": 0.028204288333654404,
+      "step": 59
+    },
+    {
+      "epoch": 0.015012666937728708,
+      "grad_norm": 184.97242736816406,
+      "learning_rate": 1.5e-06,
+      "logits/chosen": 0.8606523275375366,
+      "logits/rejected": 0.8736893534660339,
+      "logps/chosen": -881.1505126953125,
+      "logps/rejected": -857.4742431640625,
+      "loss": 5.538,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.10718020796775818,
+      "rewards/margins": 0.009757664054632187,
+      "rewards/rejected": 0.09742254763841629,
+      "step": 60
+    },
+    {
+      "epoch": 0.01526287805335752,
+      "grad_norm": 189.88743591308594,
+      "learning_rate": 1.525e-06,
+      "logits/chosen": 0.45436331629753113,
+      "logits/rejected": 0.3766404390335083,
+      "logps/chosen": -769.0473022460938,
+      "logps/rejected": -877.458740234375,
+      "loss": 5.6783,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.13664618134498596,
+      "rewards/margins": -0.02574186399579048,
+      "rewards/rejected": 0.16238804161548615,
+      "step": 61
+    },
+    {
+      "epoch": 0.015513089168986333,
+      "grad_norm": 179.95240783691406,
+      "learning_rate": 1.5500000000000002e-06,
+      "logits/chosen": 0.28293538093566895,
+      "logits/rejected": 0.23024167120456696,
+      "logps/chosen": -830.7689208984375,
+      "logps/rejected": -917.7525634765625,
+      "loss": 5.3777,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.15008211135864258,
+      "rewards/margins": 0.047852493822574615,
+      "rewards/rejected": 0.10222961753606796,
+      "step": 62
+    },
+    {
+      "epoch": 0.015763300284615145,
+      "grad_norm": 194.25332641601562,
+      "learning_rate": 1.5750000000000002e-06,
+      "logits/chosen": 0.7159794569015503,
+      "logits/rejected": 0.6552922129631042,
+      "logps/chosen": -979.01611328125,
+      "logps/rejected": -962.4176025390625,
+      "loss": 5.3516,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.12703093886375427,
+      "rewards/margins": 0.05819063261151314,
+      "rewards/rejected": 0.06884030997753143,
+      "step": 63
+    },
+    {
+      "epoch": 0.016013511400243957,
+      "grad_norm": 194.19081115722656,
+      "learning_rate": 1.6000000000000001e-06,
+      "logits/chosen": 0.4464403986930847,
+      "logits/rejected": 0.5157870054244995,
+      "logps/chosen": -936.5851440429688,
+      "logps/rejected": -935.0670776367188,
+      "loss": 5.531,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.1870882213115692,
+      "rewards/margins": 0.014117714017629623,
+      "rewards/rejected": 0.1729704886674881,
+      "step": 64
+    },
+    {
+      "epoch": 0.016263722515872766,
+      "grad_norm": 205.0813751220703,
+      "learning_rate": 1.6250000000000001e-06,
+      "logits/chosen": 0.3094771206378937,
+      "logits/rejected": 0.3199611306190491,
+      "logps/chosen": -1150.50146484375,
+      "logps/rejected": -1083.0101318359375,
+      "loss": 5.4969,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.16603738069534302,
+      "rewards/margins": 0.017452813684940338,
+      "rewards/rejected": 0.14858455955982208,
+      "step": 65
+    },
+    {
+      "epoch": 0.016513933631501578,
+      "grad_norm": 196.21629333496094,
+      "learning_rate": 1.6500000000000003e-06,
+      "logits/chosen": 0.7236325740814209,
+      "logits/rejected": 0.35501039028167725,
+      "logps/chosen": -707.7012329101562,
+      "logps/rejected": -963.8251342773438,
+      "loss": 5.4652,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.1357288509607315,
+      "rewards/margins": 0.02503032796084881,
+      "rewards/rejected": 0.11069851368665695,
+      "step": 66
+    },
+    {
+      "epoch": 0.01676414474713039,
+      "grad_norm": 196.8893585205078,
+      "learning_rate": 1.6750000000000003e-06,
+      "logits/chosen": 0.7199596166610718,
+      "logits/rejected": 0.754000723361969,
+      "logps/chosen": -926.88037109375,
+      "logps/rejected": -847.78369140625,
+      "loss": 5.6266,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.09751587361097336,
+      "rewards/margins": -0.006441976875066757,
+      "rewards/rejected": 0.10395783185958862,
+      "step": 67
+    },
+    {
+      "epoch": 0.017014355862759203,
+      "grad_norm": 180.5210723876953,
+      "learning_rate": 1.7000000000000002e-06,
+      "logits/chosen": 0.6137507557868958,
+      "logits/rejected": 0.5803739428520203,
+      "logps/chosen": -890.4776611328125,
+      "logps/rejected": -861.68701171875,
+      "loss": 5.1542,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.16620349884033203,
+      "rewards/margins": 0.11186676472425461,
+      "rewards/rejected": 0.05433673784136772,
+      "step": 68
+    },
+    {
+      "epoch": 0.017264566978388015,
+      "grad_norm": 166.2334442138672,
+      "learning_rate": 1.725e-06,
+      "logits/chosen": 0.7205318212509155,
+      "logits/rejected": 0.8995532989501953,
+      "logps/chosen": -789.7824096679688,
+      "logps/rejected": -698.5999755859375,
+      "loss": 5.2653,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.13012146949768066,
+      "rewards/margins": 0.07733415067195892,
+      "rewards/rejected": 0.05278730392456055,
+      "step": 69
+    },
+    {
+      "epoch": 0.017514778094016827,
+      "grad_norm": 184.89102172851562,
+      "learning_rate": 1.75e-06,
+      "logits/chosen": 0.5266979932785034,
+      "logits/rejected": 0.41255348920822144,
+      "logps/chosen": -937.231201171875,
+      "logps/rejected": -911.324462890625,
+      "loss": 5.4461,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.16596612334251404,
+      "rewards/margins": 0.03995685651898384,
+      "rewards/rejected": 0.1260092705488205,
+      "step": 70
+    },
+    {
+      "epoch": 0.01776498920964564,
+      "grad_norm": 175.320068359375,
+      "learning_rate": 1.7750000000000002e-06,
+      "logits/chosen": 0.5139020085334778,
+      "logits/rejected": 0.3155631721019745,
+      "logps/chosen": -860.723876953125,
+      "logps/rejected": -881.3247680664062,
+      "loss": 5.1899,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.22478055953979492,
+      "rewards/margins": 0.10203619301319122,
+      "rewards/rejected": 0.1227443665266037,
+      "step": 71
+    },
+    {
+      "epoch": 0.018015200325274452,
+      "grad_norm": 187.54153442382812,
+      "learning_rate": 1.8000000000000001e-06,
+      "logits/chosen": 0.4344066381454468,
+      "logits/rejected": 0.4636542797088623,
+      "logps/chosen": -944.456787109375,
+      "logps/rejected": -1080.532958984375,
+      "loss": 5.3375,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.18390604853630066,
+      "rewards/margins": 0.06977145373821259,
+      "rewards/rejected": 0.11413460224866867,
+      "step": 72
+    },
+    {
+      "epoch": 0.01826541144090326,
+      "grad_norm": 163.3697509765625,
+      "learning_rate": 1.825e-06,
+      "logits/chosen": 0.6777754426002502,
+      "logits/rejected": 0.7649651169776917,
+      "logps/chosen": -683.1030883789062,
+      "logps/rejected": -761.2905883789062,
+      "loss": 5.4525,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.1391131728887558,
+      "rewards/margins": 0.04012186825275421,
+      "rewards/rejected": 0.09899129718542099,
+      "step": 73
+    },
+    {
+      "epoch": 0.018515622556532073,
+      "grad_norm": 182.06051635742188,
+      "learning_rate": 1.85e-06,
+      "logits/chosen": 0.40922242403030396,
+      "logits/rejected": 0.5510671734809875,
+      "logps/chosen": -805.5526123046875,
+      "logps/rejected": -794.8642578125,
+      "loss": 5.4129,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.17711277306079865,
+      "rewards/margins": 0.05670115351676941,
+      "rewards/rejected": 0.12041164189577103,
+      "step": 74
+    },
+    {
+      "epoch": 0.018765833672160885,
+      "grad_norm": 169.18785095214844,
+      "learning_rate": 1.8750000000000003e-06,
+      "logits/chosen": 0.2944774925708771,
+      "logits/rejected": 0.23562736809253693,
+      "logps/chosen": -874.0228271484375,
+      "logps/rejected": -862.84765625,
+      "loss": 5.2592,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.12976928055286407,
+      "rewards/margins": 0.08153662085533142,
+      "rewards/rejected": 0.048232655972242355,
+      "step": 75
+    },
+    {
+      "epoch": 0.019016044787789697,
+      "grad_norm": 178.00613403320312,
+      "learning_rate": 1.9000000000000002e-06,
+      "logits/chosen": 0.5909724235534668,
+      "logits/rejected": 0.3550012409687042,
+      "logps/chosen": -850.2745361328125,
+      "logps/rejected": -972.06005859375,
+      "loss": 4.9061,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.09781656414270401,
+      "rewards/margins": 0.1920575201511383,
+      "rewards/rejected": -0.0942409560084343,
+      "step": 76
+    },
+    {
+      "epoch": 0.01926625590341851,
+      "grad_norm": 137.01124572753906,
+      "learning_rate": 1.925e-06,
+      "logits/chosen": 1.2608228921890259,
+      "logits/rejected": 1.0536128282546997,
+      "logps/chosen": -552.6478271484375,
+      "logps/rejected": -569.1873168945312,
+      "loss": 5.2777,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.17236915230751038,
+      "rewards/margins": 0.07771128416061401,
+      "rewards/rejected": 0.09465786814689636,
+      "step": 77
+    },
+    {
+      "epoch": 0.019516467019047322,
+      "grad_norm": 186.46096801757812,
+      "learning_rate": 1.9500000000000004e-06,
+      "logits/chosen": 0.5851370096206665,
+      "logits/rejected": 0.578657329082489,
+      "logps/chosen": -837.231689453125,
+      "logps/rejected": -735.5185546875,
+      "loss": 5.1302,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.16862821578979492,
+      "rewards/margins": 0.12187151610851288,
+      "rewards/rejected": 0.04675669968128204,
+      "step": 78
+    },
+    {
+      "epoch": 0.019766678134676134,
+      "grad_norm": 185.38450622558594,
+      "learning_rate": 1.975e-06,
+      "logits/chosen": 0.5800808072090149,
+      "logits/rejected": 0.6927265524864197,
+      "logps/chosen": -804.7447509765625,
+      "logps/rejected": -930.7854614257812,
+      "loss": 5.1466,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.22438755631446838,
+      "rewards/margins": 0.11162024736404419,
+      "rewards/rejected": 0.11276731640100479,
+      "step": 79
+    },
+    {
+      "epoch": 0.020016889250304946,
+      "grad_norm": 167.37783813476562,
+      "learning_rate": 2.0000000000000003e-06,
+      "logits/chosen": 0.7262884974479675,
+      "logits/rejected": 0.8875728845596313,
+      "logps/chosen": -805.3521728515625,
+      "logps/rejected": -807.3448486328125,
+      "loss": 5.1817,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.253837913274765,
+      "rewards/margins": 0.10603541135787964,
+      "rewards/rejected": 0.14780253171920776,
+      "step": 80
+    },
+    {
+      "epoch": 0.020267100365933755,
+      "grad_norm": 158.7612762451172,
+      "learning_rate": 2.025e-06,
+      "logits/chosen": 1.1694684028625488,
+      "logits/rejected": 1.2965137958526611,
+      "logps/chosen": -718.0292358398438,
+      "logps/rejected": -655.6876220703125,
+      "loss": 5.387,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.23893362283706665,
+      "rewards/margins": 0.05224871635437012,
+      "rewards/rejected": 0.18668487668037415,
+      "step": 81
+    },
+    {
+      "epoch": 0.020517311481562567,
+      "grad_norm": 192.4062957763672,
+      "learning_rate": 2.05e-06,
+      "logits/chosen": 0.8099591135978699,
+      "logits/rejected": 0.6678645610809326,
+      "logps/chosen": -758.7017822265625,
+      "logps/rejected": -1070.0118408203125,
+      "loss": 5.2297,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.2544335424900055,
+      "rewards/margins": 0.1018560379743576,
+      "rewards/rejected": 0.1525774896144867,
+      "step": 82
+    },
+    {
+      "epoch": 0.02076752259719138,
+      "grad_norm": 160.0106964111328,
+      "learning_rate": 2.075e-06,
+      "logits/chosen": 0.7472668886184692,
+      "logits/rejected": 0.949428141117096,
+      "logps/chosen": -632.663818359375,
+      "logps/rejected": -727.7824096679688,
+      "loss": 5.3828,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.2100038081407547,
+      "rewards/margins": 0.05636691302061081,
+      "rewards/rejected": 0.1536368876695633,
+      "step": 83
+    },
+    {
+      "epoch": 0.021017733712820192,
+      "grad_norm": 173.96224975585938,
+      "learning_rate": 2.1000000000000002e-06,
+      "logits/chosen": 0.8849693536758423,
+      "logits/rejected": 0.6512191295623779,
+      "logps/chosen": -855.689208984375,
+      "logps/rejected": -841.0733642578125,
+      "loss": 5.1656,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.3283468186855316,
+      "rewards/margins": 0.11740206927061081,
+      "rewards/rejected": 0.2109447419643402,
+      "step": 84
+    },
+    {
+      "epoch": 0.021267944828449004,
+      "grad_norm": 189.5381317138672,
+      "learning_rate": 2.125e-06,
+      "logits/chosen": 0.5075839757919312,
+      "logits/rejected": 0.2941034436225891,
+      "logps/chosen": -860.0894165039062,
+      "logps/rejected": -926.4581909179688,
+      "loss": 5.1784,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.25832295417785645,
+      "rewards/margins": 0.11188817024230957,
+      "rewards/rejected": 0.14643476903438568,
+      "step": 85
+    },
+    {
+      "epoch": 0.021518155944077817,
+      "grad_norm": 194.02357482910156,
+      "learning_rate": 2.15e-06,
+      "logits/chosen": 0.6765655279159546,
+      "logits/rejected": 0.49443772435188293,
+      "logps/chosen": -729.0982666015625,
+      "logps/rejected": -822.9295043945312,
+      "loss": 5.5596,
+      "rewards/accuracies": 0.4375,
+      "rewards/chosen": 0.21729154884815216,
+      "rewards/margins": 0.016531461849808693,
+      "rewards/rejected": 0.20076008141040802,
+      "step": 86
+    },
+    {
+      "epoch": 0.02176836705970663,
+      "grad_norm": 178.7899627685547,
+      "learning_rate": 2.1750000000000004e-06,
+      "logits/chosen": 0.7486958503723145,
+      "logits/rejected": 1.0325416326522827,
+      "logps/chosen": -824.9302978515625,
+      "logps/rejected": -687.9011840820312,
+      "loss": 5.2383,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.28955578804016113,
+      "rewards/margins": 0.09782307595014572,
+      "rewards/rejected": 0.19173270463943481,
+      "step": 87
+    },
+    {
+      "epoch": 0.022018578175335438,
+      "grad_norm": 164.25181579589844,
+      "learning_rate": 2.2e-06,
+      "logits/chosen": 0.9054800271987915,
+      "logits/rejected": 0.9131541848182678,
+      "logps/chosen": -796.5818481445312,
+      "logps/rejected": -905.0699462890625,
+      "loss": 4.957,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.4441540837287903,
+      "rewards/margins": 0.19320707023143768,
+      "rewards/rejected": 0.2509469985961914,
+      "step": 88
+    },
+    {
+      "epoch": 0.02226878929096425,
+      "grad_norm": 156.86618041992188,
+      "learning_rate": 2.2250000000000003e-06,
+      "logits/chosen": 0.9420207738876343,
+      "logits/rejected": 0.9043740034103394,
+      "logps/chosen": -697.5362548828125,
+      "logps/rejected": -696.8550415039062,
+      "loss": 5.0714,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.288621723651886,
+      "rewards/margins": 0.14416837692260742,
+      "rewards/rejected": 0.14445334672927856,
+      "step": 89
+    },
+    {
+      "epoch": 0.022519000406593062,
+      "grad_norm": 195.37326049804688,
+      "learning_rate": 2.25e-06,
+      "logits/chosen": 0.48666009306907654,
+      "logits/rejected": 0.5969275236129761,
+      "logps/chosen": -861.2529907226562,
+      "logps/rejected": -776.2673950195312,
+      "loss": 5.6532,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.26983776688575745,
+      "rewards/margins": -0.010628409683704376,
+      "rewards/rejected": 0.2804661691188812,
+      "step": 90
+    },
+    {
+      "epoch": 0.022769211522221874,
+      "grad_norm": 165.2438507080078,
+      "learning_rate": 2.2750000000000002e-06,
+      "logits/chosen": 0.8704694509506226,
+      "logits/rejected": 0.8926208019256592,
+      "logps/chosen": -675.2134399414062,
+      "logps/rejected": -655.6943969726562,
+      "loss": 5.3438,
+      "rewards/accuracies": 0.5625,
+      "rewards/chosen": 0.2891599237918854,
+      "rewards/margins": 0.06688322126865387,
+      "rewards/rejected": 0.2222766876220703,
+      "step": 91
+    },
+    {
+      "epoch": 0.023019422637850687,
+      "grad_norm": 183.89662170410156,
+      "learning_rate": 2.3000000000000004e-06,
+      "logits/chosen": 0.5281510353088379,
+      "logits/rejected": 0.5520732998847961,
+      "logps/chosen": -822.0519409179688,
+      "logps/rejected": -867.435791015625,
+      "loss": 5.0473,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": 0.33428192138671875,
+      "rewards/margins": 0.14445266127586365,
+      "rewards/rejected": 0.1898292601108551,
+      "step": 92
+    },
+    {
+      "epoch": 0.0232696337534795,
+      "grad_norm": 154.9501495361328,
+      "learning_rate": 2.325e-06,
+      "logits/chosen": 0.8877805471420288,
+      "logits/rejected": 1.1910347938537598,
+      "logps/chosen": -712.02587890625,
+      "logps/rejected": -649.3619995117188,
+      "loss": 4.8226,
+      "rewards/accuracies": 0.8125,
+      "rewards/chosen": 0.27764156460762024,
+      "rewards/margins": 0.21533805131912231,
+      "rewards/rejected": 0.06230354681611061,
+      "step": 93
+    },
+    {
+      "epoch": 0.02351984486910831,
+      "grad_norm": 144.99720764160156,
+      "learning_rate": 2.35e-06,
+      "logits/chosen": 0.6367812156677246,
+      "logits/rejected": 0.7492159008979797,
+      "logps/chosen": -862.7364501953125,
+      "logps/rejected": -663.7855224609375,
+      "loss": 4.9859,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.29428744316101074,
+      "rewards/margins": 0.17723265290260315,
+      "rewards/rejected": 0.11705479770898819,
+      "step": 94
+    },
+    {
+      "epoch": 0.023770055984737123,
+      "grad_norm": 198.8605194091797,
+      "learning_rate": 2.375e-06,
+      "logits/chosen": 0.24357648193836212,
+      "logits/rejected": 0.4175962209701538,
+      "logps/chosen": -919.9758911132812,
+      "logps/rejected": -958.36083984375,
+      "loss": 5.4862,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.22103404998779297,
+      "rewards/margins": 0.038531675934791565,
+      "rewards/rejected": 0.1825023740530014,
+      "step": 95
+    },
+    {
+      "epoch": 0.024020267100365932,
+      "grad_norm": 152.70587158203125,
+      "learning_rate": 2.4000000000000003e-06,
+      "logits/chosen": 0.2301986962556839,
+      "logits/rejected": 0.5025601983070374,
+      "logps/chosen": -877.77587890625,
+      "logps/rejected": -727.943359375,
+      "loss": 4.7149,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": 0.3718917667865753,
+      "rewards/margins": 0.24806731939315796,
+      "rewards/rejected": 0.12382444739341736,
+      "step": 96
+    },
+    {
+      "epoch": 0.024270478215994745,
+      "grad_norm": 152.5541534423828,
+      "learning_rate": 2.425e-06,
+      "logits/chosen": 0.5611572861671448,
+      "logits/rejected": 0.6057158708572388,
+      "logps/chosen": -712.0330200195312,
+      "logps/rejected": -764.8245849609375,
+      "loss": 4.8191,
+      "rewards/accuracies": 0.6875,
+      "rewards/chosen": 0.24860484898090363,
+      "rewards/margins": 0.20819023251533508,
+      "rewards/rejected": 0.04041462019085884,
+      "step": 97
+    },
+    {
+      "epoch": 0.024520689331623557,
+      "grad_norm": 188.4685516357422,
+      "learning_rate": 2.4500000000000003e-06,
+      "logits/chosen": 0.3046216070652008,
+      "logits/rejected": 0.4239407181739807,
+      "logps/chosen": -877.07373046875,
+      "logps/rejected": -900.893798828125,
+      "loss": 5.3107,
+      "rewards/accuracies": 0.5,
+      "rewards/chosen": 0.4058847427368164,
+      "rewards/margins": 0.0769483894109726,
+      "rewards/rejected": 0.328936368227005,
+      "step": 98
+    },
+    {
+      "epoch": 0.02477090044725237,
+      "grad_norm": 143.4902801513672,
+      "learning_rate": 2.475e-06,
+      "logits/chosen": 0.7912758588790894,
+      "logits/rejected": 0.6307110786437988,
+      "logps/chosen": -642.6576538085938,
+      "logps/rejected": -860.2074584960938,
+      "loss": 4.5087,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": 0.3818928897380829,
+      "rewards/margins": 0.3079994320869446,
+      "rewards/rejected": 0.0738934651017189,
+      "step": 99
+    },
+    {
+      "epoch": 0.02502111156288118,
+      "grad_norm": 142.76861572265625,
+      "learning_rate": 2.5e-06,
+      "logits/chosen": 1.039554238319397,
+      "logits/rejected": 1.1768317222595215,
+      "logps/chosen": -710.2076416015625,
+      "logps/rejected": -656.5585327148438,
+      "loss": 4.3711,
+      "rewards/accuracies": 0.9375,
+      "rewards/chosen": 0.33966264128685,
+      "rewards/margins": 0.3506268858909607,
+      "rewards/rejected": -0.010964250192046165,
+      "step": 100
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

last-checkpoint/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5ba1d346aa91501b4477ac53b83e0f97d1e49e34c6bc7cf85b91ff38f8939f2
+size 7096