checkpoint step 250

.gitattributes

@@ -37,3 +37,4 @@ checkpoint-50/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoint-150/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-250/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoint-250/README.md

@@ -0,0 +1,209 @@
+---
+base_model: unsloth/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/Qwen2.5-7B-Instruct
+- grpo
+- lora
+- transformers
+- trl
+---
+
+# 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 0.19.1

checkpoint-250/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "unsloth/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "up_proj",
+    "k_proj",
+    "v_proj",
+    "gate_proj",
+    "q_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoint-250/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f1bc45a0aeab4a94479459bdc28937b43f4835fbc5db962bc392d333fb9c80b
+size 80792880

checkpoint-250/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoint-250/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8082a4b83a1ad4d8b26a7c5b25c6fbd82e03cd1e8eeb7fc97080a17e49a01c5f
+size 161816251

checkpoint-250/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9959c6e3e1b617a6724ce1d70ffea05ccf5ce9f07a4e9daed34597feda7daa5
+size 14645

checkpoint-250/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:539e1a1b51bdc68591f96c992e4b9aed6ce307481cdb426a0e8ffb9a3a8acb4b
+size 1465

checkpoint-250/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
+size 11421892

checkpoint-250/tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 32768,
+  "pad_token": "<|vision_pad|>",
+  "padding_side": "left",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-250/trainer_state.json

@@ -0,0 +1,1384 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.010775862068965518,
+  "eval_steps": 500,
+  "global_step": 250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 130.0,
+      "completions/max_terminated_length": 130.0,
+      "completions/mean_length": 68.5,
+      "completions/mean_terminated_length": 68.5,
+      "completions/min_length": 26.8,
+      "completions/min_terminated_length": 26.8,
+      "entropy": 0.25826080311089755,
+      "epoch": 0.00021551724137931034,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.10238287448883057,
+      "num_tokens": 12064.0,
+      "reward": 0.3290319949388504,
+      "reward_std": 0.40028320252895355,
+      "rewards/reward_fn/mean": 0.3290319949388504,
+      "rewards/reward_fn/std": 0.400283208489418,
+      "step": 5,
+      "step_time": 22.194597821800016
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 87.6,
+      "completions/max_terminated_length": 87.6,
+      "completions/mean_length": 48.525,
+      "completions/mean_terminated_length": 48.525,
+      "completions/min_length": 17.0,
+      "completions/min_terminated_length": 17.0,
+      "entropy": 0.2673827801831067,
+      "epoch": 0.0004310344827586207,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.25e-06,
+      "loss": 0.05849265456199646,
+      "num_tokens": 23089.0,
+      "reward": 0.45969198942184447,
+      "reward_std": 0.2795014828443527,
+      "rewards/reward_fn/mean": 0.45969198942184447,
+      "rewards/reward_fn/std": 0.27950150668621065,
+      "step": 10,
+      "step_time": 16.331819552399928
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 85.6,
+      "completions/max_terminated_length": 85.6,
+      "completions/mean_length": 47.35,
+      "completions/mean_terminated_length": 47.35,
+      "completions/min_length": 25.0,
+      "completions/min_terminated_length": 25.0,
+      "entropy": 0.20174795808270574,
+      "epoch": 0.000646551724137931,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.5e-06,
+      "loss": -0.03555725216865539,
+      "num_tokens": 34239.0,
+      "reward": 0.5545999944210053,
+      "reward_std": 0.32832055240869523,
+      "rewards/reward_fn/mean": 0.5545999944210053,
+      "rewards/reward_fn/std": 0.3283205583691597,
+      "step": 15,
+      "step_time": 16.44816417159991
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 98.2,
+      "completions/max_terminated_length": 98.2,
+      "completions/mean_length": 53.6,
+      "completions/mean_terminated_length": 53.6,
+      "completions/min_length": 25.0,
+      "completions/min_terminated_length": 25.0,
+      "entropy": 0.2620095924474299,
+      "epoch": 0.0008620689655172414,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.75e-06,
+      "loss": 0.017401468753814698,
+      "num_tokens": 45399.0,
+      "reward": 0.4466240078210831,
+      "reward_std": 0.27573536019772293,
+      "rewards/reward_fn/mean": 0.4466240078210831,
+      "rewards/reward_fn/std": 0.27573537137359383,
+      "step": 20,
+      "step_time": 17.988385831799953
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 123.8,
+      "completions/max_terminated_length": 123.8,
+      "completions/mean_length": 59.15,
+      "completions/mean_terminated_length": 59.15,
+      "completions/min_length": 21.8,
+      "completions/min_terminated_length": 21.8,
+      "entropy": 0.33701689867302775,
+      "epoch": 0.0010775862068965517,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.965517241379311e-06,
+      "loss": -0.02674364447593689,
+      "num_tokens": 56993.0,
+      "reward": 0.5006439983844757,
+      "reward_std": 0.2351181447505951,
+      "rewards/reward_fn/mean": 0.5006439983844757,
+      "rewards/reward_fn/std": 0.23511814773082734,
+      "step": 25,
+      "step_time": 21.11461545459997
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 110.2,
+      "completions/max_terminated_length": 110.2,
+      "completions/mean_length": 47.95,
+      "completions/mean_terminated_length": 47.95,
+      "completions/min_length": 15.8,
+      "completions/min_terminated_length": 15.8,
+      "entropy": 0.24504410615190864,
+      "epoch": 0.001293103448275862,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.922413793103449e-06,
+      "loss": -0.05023183822631836,
+      "num_tokens": 67659.0,
+      "reward": 0.6457239985466003,
+      "reward_std": 0.11179900387069211,
+      "rewards/reward_fn/mean": 0.6457239985466003,
+      "rewards/reward_fn/std": 0.11179901438299567,
+      "step": 30,
+      "step_time": 19.134268762199827
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.1,
+      "completions/max_length": 150.8,
+      "completions/max_terminated_length": 120.2,
+      "completions/mean_length": 73.375,
+      "completions/mean_terminated_length": 54.36666717529297,
+      "completions/min_length": 15.2,
+      "completions/min_terminated_length": 15.2,
+      "entropy": 0.23285924410447478,
+      "epoch": 0.0015086206896551724,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.879310344827586e-06,
+      "loss": -0.05149807929992676,
+      "num_tokens": 79734.0,
+      "reward": 0.46414998471736907,
+      "reward_std": 0.4071369742392562,
+      "rewards/reward_fn/mean": 0.46414998471736907,
+      "rewards/reward_fn/std": 0.40713699695188554,
+      "step": 35,
+      "step_time": 24.679699858600042
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 125.4,
+      "completions/max_terminated_length": 125.4,
+      "completions/mean_length": 68.925,
+      "completions/mean_terminated_length": 68.925,
+      "completions/min_length": 26.0,
+      "completions/min_terminated_length": 26.0,
+      "entropy": 0.3110098702833056,
+      "epoch": 0.0017241379310344827,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.836206896551724e-06,
+      "loss": 0.030162644386291505,
+      "num_tokens": 91695.0,
+      "reward": 0.5143539935350419,
+      "reward_std": 0.2790891878306866,
+      "rewards/reward_fn/mean": 0.5143539935350419,
+      "rewards/reward_fn/std": 0.279089218378067,
+      "step": 40,
+      "step_time": 21.565513409599998
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 100.0,
+      "completions/max_terminated_length": 100.0,
+      "completions/mean_length": 53.875,
+      "completions/mean_terminated_length": 53.875,
+      "completions/min_length": 16.6,
+      "completions/min_terminated_length": 16.6,
+      "entropy": 0.23047098610550165,
+      "epoch": 0.001939655172413793,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.793103448275862e-06,
+      "loss": -0.04350074529647827,
+      "num_tokens": 102994.0,
+      "reward": 0.6398920059204102,
+      "reward_std": 0.2279826147481799,
+      "rewards/reward_fn/mean": 0.6398920059204102,
+      "rewards/reward_fn/std": 0.22798261437565087,
+      "step": 45,
+      "step_time": 18.23975218899968
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 133.4,
+      "completions/max_terminated_length": 133.4,
+      "completions/mean_length": 75.975,
+      "completions/mean_terminated_length": 75.975,
+      "completions/min_length": 23.8,
+      "completions/min_terminated_length": 23.8,
+      "entropy": 0.34613882582634686,
+      "epoch": 0.0021551724137931034,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.75e-06,
+      "loss": -0.03668657541275024,
+      "num_tokens": 115289.0,
+      "reward": 0.33042599707841874,
+      "reward_std": 0.5918701648712158,
+      "rewards/reward_fn/mean": 0.33042599707841874,
+      "rewards/reward_fn/std": 0.5918702006340026,
+      "step": 50,
+      "step_time": 22.43372436519985
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 92.2,
+      "completions/max_terminated_length": 92.2,
+      "completions/mean_length": 50.65,
+      "completions/mean_terminated_length": 50.65,
+      "completions/min_length": 18.0,
+      "completions/min_terminated_length": 18.0,
+      "entropy": 0.3391244841972366,
+      "epoch": 0.0023706896551724138,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.706896551724138e-06,
+      "loss": -0.0521969735622406,
+      "num_tokens": 126111.0,
+      "reward": 0.6139639973640442,
+      "reward_std": 0.18208687230944634,
+      "rewards/reward_fn/mean": 0.6139639973640442,
+      "rewards/reward_fn/std": 0.18208688013255597,
+      "step": 55,
+      "step_time": 16.889007100199933
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 164.0,
+      "completions/max_terminated_length": 162.4,
+      "completions/mean_length": 82.1,
+      "completions/mean_terminated_length": 78.87857208251953,
+      "completions/min_length": 29.4,
+      "completions/min_terminated_length": 29.4,
+      "entropy": 0.28240502553526314,
+      "epoch": 0.002586206896551724,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.663793103448276e-06,
+      "loss": -0.0108717679977417,
+      "num_tokens": 138667.0,
+      "reward": 0.6616780042648316,
+      "reward_std": 0.18088504523038865,
+      "rewards/reward_fn/mean": 0.6616780042648316,
+      "rewards/reward_fn/std": 0.18088504672050476,
+      "step": 60,
+      "step_time": 26.46498533039994
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 134.8,
+      "completions/max_terminated_length": 133.0,
+      "completions/mean_length": 69.55,
+      "completions/mean_terminated_length": 63.233334350585935,
+      "completions/min_length": 15.8,
+      "completions/min_terminated_length": 15.8,
+      "entropy": 0.3688268234953284,
+      "epoch": 0.0028017241379310344,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.620689655172414e-06,
+      "loss": 0.09875075221061706,
+      "num_tokens": 150585.0,
+      "reward": 0.5008320093154908,
+      "reward_std": 0.2221744753420353,
+      "rewards/reward_fn/mean": 0.5008320093154908,
+      "rewards/reward_fn/std": 0.22217447832226753,
+      "step": 65,
+      "step_time": 22.69940989719953
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 149.4,
+      "completions/max_terminated_length": 141.4,
+      "completions/mean_length": 82.125,
+      "completions/mean_terminated_length": 76.2107162475586,
+      "completions/min_length": 34.8,
+      "completions/min_terminated_length": 34.8,
+      "entropy": 0.2419425747357309,
+      "epoch": 0.003017241379310345,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.577586206896552e-06,
+      "loss": -0.06608393192291259,
+      "num_tokens": 162726.0,
+      "reward": 0.35277000069618225,
+      "reward_std": 0.37774557205848397,
+      "rewards/reward_fn/mean": 0.35277000069618225,
+      "rewards/reward_fn/std": 0.37774557354860006,
+      "step": 70,
+      "step_time": 24.30778650340026
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 137.6,
+      "completions/max_terminated_length": 135.6,
+      "completions/mean_length": 77.55,
+      "completions/mean_terminated_length": 74.46785888671874,
+      "completions/min_length": 41.8,
+      "completions/min_terminated_length": 41.8,
+      "entropy": 0.35793030727654696,
+      "epoch": 0.003232758620689655,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.53448275862069e-06,
+      "loss": 0.10921311378479004,
+      "num_tokens": 175364.0,
+      "reward": 0.15934600830078124,
+      "reward_std": 0.5374872148036957,
+      "rewards/reward_fn/mean": 0.15934600830078124,
+      "rewards/reward_fn/std": 0.5374872386455536,
+      "step": 75,
+      "step_time": 23.099894465200123
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 95.6,
+      "completions/max_terminated_length": 95.6,
+      "completions/mean_length": 53.175,
+      "completions/mean_terminated_length": 53.175,
+      "completions/min_length": 15.8,
+      "completions/min_terminated_length": 15.8,
+      "entropy": 0.2758270605234429,
+      "epoch": 0.0034482758620689655,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.4913793103448275e-06,
+      "loss": 0.11924625635147094,
+      "num_tokens": 186407.0,
+      "reward": 0.6774819850921631,
+      "reward_std": 0.11054287778679281,
+      "rewards/reward_fn/mean": 0.6774819850921631,
+      "rewards/reward_fn/std": 0.11054288879968226,
+      "step": 80,
+      "step_time": 17.348075130399685
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 137.6,
+      "completions/max_terminated_length": 137.6,
+      "completions/mean_length": 68.25,
+      "completions/mean_terminated_length": 68.25,
+      "completions/min_length": 30.0,
+      "completions/min_terminated_length": 30.0,
+      "entropy": 0.2696234828326851,
+      "epoch": 0.003663793103448276,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.4482758620689656e-06,
+      "loss": -0.03822631537914276,
+      "num_tokens": 198241.0,
+      "reward": 0.5305620029568672,
+      "reward_std": 0.4029154841089621,
+      "rewards/reward_fn/mean": 0.5305620029568672,
+      "rewards/reward_fn/std": 0.4029154877178371,
+      "step": 85,
+      "step_time": 22.834117503599828
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 131.2,
+      "completions/max_terminated_length": 131.2,
+      "completions/mean_length": 77.45,
+      "completions/mean_terminated_length": 77.45,
+      "completions/min_length": 19.8,
+      "completions/min_terminated_length": 19.8,
+      "entropy": 0.3707127865403891,
+      "epoch": 0.003879310344827586,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.405172413793104e-06,
+      "loss": 0.035861659049987796,
+      "num_tokens": 210935.0,
+      "reward": 0.3616240084171295,
+      "reward_std": 0.48762375079095366,
+      "rewards/reward_fn/mean": 0.3616240084171295,
+      "rewards/reward_fn/std": 0.48762374445796014,
+      "step": 90,
+      "step_time": 22.48314213299982
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.075,
+      "completions/max_length": 118.2,
+      "completions/max_terminated_length": 116.0,
+      "completions/mean_length": 66.9,
+      "completions/mean_terminated_length": 56.7,
+      "completions/min_length": 25.0,
+      "completions/min_terminated_length": 25.0,
+      "entropy": 0.342819757014513,
+      "epoch": 0.004094827586206897,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.362068965517242e-06,
+      "loss": -0.051665693521499634,
+      "num_tokens": 222647.0,
+      "reward": 0.46663198471069334,
+      "reward_std": 0.3400567059754394,
+      "rewards/reward_fn/mean": 0.46663198471069334,
+      "rewards/reward_fn/std": 0.34005671155173334,
+      "step": 95,
+      "step_time": 20.31996373819993
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 82.2,
+      "completions/max_terminated_length": 82.2,
+      "completions/mean_length": 39.975,
+      "completions/mean_terminated_length": 39.975,
+      "completions/min_length": 16.4,
+      "completions/min_terminated_length": 16.4,
+      "entropy": 0.25842549465596676,
+      "epoch": 0.004310344827586207,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.31896551724138e-06,
+      "loss": 0.029389530420303345,
+      "num_tokens": 233466.0,
+      "reward": 0.5307899951934815,
+      "reward_std": 0.28180397795513273,
+      "rewards/reward_fn/mean": 0.5307899951934815,
+      "rewards/reward_fn/std": 0.28180398060940204,
+      "step": 100,
+      "step_time": 15.753981888000453
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 77.8,
+      "completions/max_terminated_length": 77.8,
+      "completions/mean_length": 49.6,
+      "completions/mean_terminated_length": 49.6,
+      "completions/min_length": 16.4,
+      "completions/min_terminated_length": 16.4,
+      "entropy": 0.27716706816572695,
+      "epoch": 0.004525862068965518,
+      "frac_reward_zero_std": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.275862068965518e-06,
+      "loss": 0.011613726615905762,
+      "num_tokens": 244550.0,
+      "reward": 0.6918019950389862,
+      "reward_std": 0.10386521331965923,
+      "rewards/reward_fn/mean": 0.6918019950389862,
+      "rewards/reward_fn/std": 0.10386521825566888,
+      "step": 105,
+      "step_time": 15.171548829199946
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 94.0,
+      "completions/max_terminated_length": 94.0,
+      "completions/mean_length": 47.1,
+      "completions/mean_terminated_length": 47.1,
+      "completions/min_length": 16.0,
+      "completions/min_terminated_length": 16.0,
+      "entropy": 0.1942274721484864,
+      "epoch": 0.0047413793103448275,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.232758620689656e-06,
+      "loss": -0.000987192615866661,
+      "num_tokens": 255962.0,
+      "reward": 0.1787539839744568,
+      "reward_std": 0.5241983281448483,
+      "rewards/reward_fn/mean": 0.1787539839744568,
+      "rewards/reward_fn/std": 0.5241983408108354,
+      "step": 110,
+      "step_time": 17.40271624299976
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 108.4,
+      "completions/max_terminated_length": 108.4,
+      "completions/mean_length": 54.025,
+      "completions/mean_terminated_length": 54.025,
+      "completions/min_length": 26.8,
+      "completions/min_terminated_length": 26.8,
+      "entropy": 0.21391218816424953,
+      "epoch": 0.004956896551724138,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.189655172413794e-06,
+      "loss": -0.07643186450004577,
+      "num_tokens": 267563.0,
+      "reward": 0.5654179990291596,
+      "reward_std": 0.2869014423340559,
+      "rewards/reward_fn/mean": 0.5654179990291596,
+      "rewards/reward_fn/std": 0.2869014598429203,
+      "step": 115,
+      "step_time": 19.255251389799923
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 136.0,
+      "completions/max_terminated_length": 124.6,
+      "completions/mean_length": 80.575,
+      "completions/mean_terminated_length": 77.74285888671875,
+      "completions/min_length": 33.0,
+      "completions/min_terminated_length": 33.0,
+      "entropy": 0.31822405869606885,
+      "epoch": 0.005172413793103448,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.146551724137932e-06,
+      "loss": 0.01645863652229309,
+      "num_tokens": 280074.0,
+      "reward": 0.06659798473119735,
+      "reward_std": 0.5049000799655914,
+      "rewards/reward_fn/mean": 0.06659798473119735,
+      "rewards/reward_fn/std": 0.5049000844359398,
+      "step": 120,
+      "step_time": 22.649546912000005
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 138.4,
+      "completions/max_terminated_length": 136.2,
+      "completions/mean_length": 68.85,
+      "completions/mean_terminated_length": 61.975,
+      "completions/min_length": 25.4,
+      "completions/min_terminated_length": 25.4,
+      "entropy": 0.33708888398250564,
+      "epoch": 0.005387931034482759,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 1.25,
+      "learning_rate": 4.103448275862069e-06,
+      "loss": 0.16269092559814452,
+      "num_tokens": 291760.0,
+      "reward": 0.09610399603843689,
+      "reward_std": 0.6181229472160339,
+      "rewards/reward_fn/mean": 0.09610399603843689,
+      "rewards/reward_fn/std": 0.6181229710578918,
+      "step": 125,
+      "step_time": 22.892032967400247
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 121.6,
+      "completions/max_terminated_length": 119.6,
+      "completions/mean_length": 54.625,
+      "completions/mean_terminated_length": 45.516667175292966,
+      "completions/min_length": 15.4,
+      "completions/min_terminated_length": 15.4,
+      "entropy": 0.2316889252353576,
+      "epoch": 0.005603448275862069,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.060344827586207e-06,
+      "loss": 0.05887578129768371,
+      "num_tokens": 302777.0,
+      "reward": 0.47827199250459673,
+      "reward_std": 0.31257240688428284,
+      "rewards/reward_fn/mean": 0.47827199250459673,
+      "rewards/reward_fn/std": 0.31257240623235705,
+      "step": 130,
+      "step_time": 20.756939186200178
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 83.4,
+      "completions/max_terminated_length": 83.4,
+      "completions/mean_length": 43.375,
+      "completions/mean_terminated_length": 43.375,
+      "completions/min_length": 17.0,
+      "completions/min_terminated_length": 17.0,
+      "entropy": 0.14220066292909905,
+      "epoch": 0.00581896551724138,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.017241379310346e-06,
+      "loss": 0.0067549549043178555,
+      "num_tokens": 313896.0,
+      "reward": 0.5540139883756637,
+      "reward_std": 0.3137821853160858,
+      "rewards/reward_fn/mean": 0.5540139883756637,
+      "rewards/reward_fn/std": 0.31378220971673726,
+      "step": 135,
+      "step_time": 16.04840490619972
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 65.2,
+      "completions/max_terminated_length": 65.2,
+      "completions/mean_length": 33.825,
+      "completions/mean_terminated_length": 33.825,
+      "completions/min_length": 15.8,
+      "completions/min_terminated_length": 15.8,
+      "entropy": 0.16708708838559688,
+      "epoch": 0.00603448275862069,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.974137931034483e-06,
+      "loss": 0.000922510027885437,
+      "num_tokens": 324093.0,
+      "reward": 0.5665820002555847,
+      "reward_std": 0.3455619063111953,
+      "rewards/reward_fn/mean": 0.5665820002555847,
+      "rewards/reward_fn/std": 0.34556191868614405,
+      "step": 140,
+      "step_time": 13.251863584200146
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.1,
+      "completions/max_length": 175.8,
+      "completions/max_terminated_length": 149.4,
+      "completions/mean_length": 76.9,
+      "completions/mean_terminated_length": 58.73142929077149,
+      "completions/min_length": 25.6,
+      "completions/min_terminated_length": 25.6,
+      "entropy": 0.2752605409361422,
+      "epoch": 0.00625,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.931034482758621e-06,
+      "loss": -0.012939244508743286,
+      "num_tokens": 336493.0,
+      "reward": 0.03320800065994263,
+      "reward_std": 0.6782135218381882,
+      "rewards/reward_fn/mean": 0.03320800065994263,
+      "rewards/reward_fn/std": 0.6782135486602783,
+      "step": 145,
+      "step_time": 27.717435833599847
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 159.6,
+      "completions/max_terminated_length": 159.4,
+      "completions/mean_length": 86.4,
+      "completions/mean_terminated_length": 78.2,
+      "completions/min_length": 27.0,
+      "completions/min_terminated_length": 27.0,
+      "entropy": 0.3023837681859732,
+      "epoch": 0.00646551724137931,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.25,
+      "learning_rate": 3.887931034482759e-06,
+      "loss": -0.004035864770412445,
+      "num_tokens": 349213.0,
+      "reward": 0.2794659972190857,
+      "reward_std": 0.33909483794122935,
+      "rewards/reward_fn/mean": 0.2794659972190857,
+      "rewards/reward_fn/std": 0.3390948371961713,
+      "step": 150,
+      "step_time": 25.6254667358
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 120.8,
+      "completions/max_terminated_length": 117.2,
+      "completions/mean_length": 76.35,
+      "completions/mean_terminated_length": 74.68571472167969,
+      "completions/min_length": 42.2,
+      "completions/min_terminated_length": 42.2,
+      "entropy": 0.2444805505729164,
+      "epoch": 0.006681034482758621,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.844827586206897e-06,
+      "loss": 0.06990929841995239,
+      "num_tokens": 361243.0,
+      "reward": 0.31372600197792055,
+      "reward_std": 0.16792407557368277,
+      "rewards/reward_fn/mean": 0.31372600197792055,
+      "rewards/reward_fn/std": 0.16792407706379892,
+      "step": 155,
+      "step_time": 20.657237647200326
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 86.2,
+      "completions/max_terminated_length": 78.2,
+      "completions/mean_length": 57.375,
+      "completions/mean_terminated_length": 54.985714721679685,
+      "completions/min_length": 36.2,
+      "completions/min_terminated_length": 36.2,
+      "entropy": 0.2041449310956523,
+      "epoch": 0.006896551724137931,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.801724137931035e-06,
+      "loss": -0.003799402713775635,
+      "num_tokens": 372606.0,
+      "reward": 0.6000540047883988,
+      "reward_std": 0.14438311448320745,
+      "rewards/reward_fn/mean": 0.6000540047883988,
+      "rewards/reward_fn/std": 0.1443831158336252,
+      "step": 160,
+      "step_time": 15.94448771479947
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.1,
+      "completions/max_length": 196.6,
+      "completions/max_terminated_length": 173.8,
+      "completions/mean_length": 114.125,
+      "completions/mean_terminated_length": 102.28333435058593,
+      "completions/min_length": 40.8,
+      "completions/min_terminated_length": 40.8,
+      "entropy": 0.3578500231727958,
+      "epoch": 0.007112068965517242,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.7586206896551727e-06,
+      "loss": 0.05463705062866211,
+      "num_tokens": 386899.0,
+      "reward": 0.29335000514984133,
+      "reward_std": 0.6011195421218872,
+      "rewards/reward_fn/mean": 0.29335000514984133,
+      "rewards/reward_fn/std": 0.6011195570230484,
+      "step": 165,
+      "step_time": 30.583657701599986
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 120.6,
+      "completions/max_terminated_length": 105.4,
+      "completions/mean_length": 64.775,
+      "completions/mean_terminated_length": 60.79285736083985,
+      "completions/min_length": 27.2,
+      "completions/min_terminated_length": 27.2,
+      "entropy": 0.3172524501627777,
+      "epoch": 0.007327586206896552,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.7155172413793107e-06,
+      "loss": -0.04064536094665527,
+      "num_tokens": 398878.0,
+      "reward": 0.2578879952430725,
+      "reward_std": 0.4871767422184348,
+      "rewards/reward_fn/mean": 0.2578879952430725,
+      "rewards/reward_fn/std": 0.4871767588891089,
+      "step": 170,
+      "step_time": 20.961449378399994
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 86.0,
+      "completions/max_terminated_length": 86.0,
+      "completions/mean_length": 49.1,
+      "completions/mean_terminated_length": 49.1,
+      "completions/min_length": 25.4,
+      "completions/min_terminated_length": 25.4,
+      "entropy": 0.28907319878344423,
+      "epoch": 0.007543103448275862,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.6724137931034487e-06,
+      "loss": 0.07596501111984252,
+      "num_tokens": 409894.0,
+      "reward": 0.2182680130004883,
+      "reward_std": 0.2655696720117703,
+      "rewards/reward_fn/mean": 0.2182680130004883,
+      "rewards/reward_fn/std": 0.26556967524811625,
+      "step": 175,
+      "step_time": 16.00376275539984
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.075,
+      "completions/max_length": 119.0,
+      "completions/max_terminated_length": 118.0,
+      "completions/mean_length": 75.025,
+      "completions/mean_terminated_length": 60.685000610351565,
+      "completions/min_length": 30.4,
+      "completions/min_terminated_length": 30.4,
+      "entropy": 0.2300539159856271,
+      "epoch": 0.007758620689655172,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.6293103448275863e-06,
+      "loss": 0.0008989959955215454,
+      "num_tokens": 422115.0,
+      "reward": 0.2748300015926361,
+      "reward_std": 0.4872116835438646,
+      "rewards/reward_fn/mean": 0.2748300015926361,
+      "rewards/reward_fn/std": 0.487211688212119,
+      "step": 180,
+      "step_time": 20.472563396999977
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 140.8,
+      "completions/max_terminated_length": 130.4,
+      "completions/mean_length": 77.75,
+      "completions/mean_terminated_length": 73.95714416503907,
+      "completions/min_length": 28.4,
+      "completions/min_terminated_length": 28.4,
+      "entropy": 0.2450535054318607,
+      "epoch": 0.007974137931034483,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5862068965517243e-06,
+      "loss": 0.041662493348121644,
+      "num_tokens": 434661.0,
+      "reward": 0.052784007787704465,
+      "reward_std": 0.2650376170873642,
+      "rewards/reward_fn/mean": 0.052784007787704465,
+      "rewards/reward_fn/std": 0.26503762900829314,
+      "step": 185,
+      "step_time": 23.318358728199563
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 195.6,
+      "completions/max_terminated_length": 192.8,
+      "completions/mean_length": 89.9,
+      "completions/mean_terminated_length": 84.275,
+      "completions/min_length": 26.2,
+      "completions/min_terminated_length": 26.2,
+      "entropy": 0.3329991169273853,
+      "epoch": 0.008189655172413794,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.5431034482758623e-06,
+      "loss": 0.014610832929611206,
+      "num_tokens": 447801.0,
+      "reward": 0.449290007352829,
+      "reward_std": 0.2617717768996954,
+      "rewards/reward_fn/mean": 0.449290007352829,
+      "rewards/reward_fn/std": 0.26177178248763083,
+      "step": 190,
+      "step_time": 30.514870558599977
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.075,
+      "completions/max_length": 174.6,
+      "completions/max_terminated_length": 167.2,
+      "completions/mean_length": 102.6,
+      "completions/mean_terminated_length": 93.59166717529297,
+      "completions/min_length": 33.8,
+      "completions/min_terminated_length": 33.8,
+      "entropy": 0.27610186783713286,
+      "epoch": 0.008405172413793103,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.5e-06,
+      "loss": 0.014060372114181518,
+      "num_tokens": 461297.0,
+      "reward": 0.18859200328588485,
+      "reward_std": 0.20913713360205294,
+      "rewards/reward_fn/mean": 0.18859200328588485,
+      "rewards/reward_fn/std": 0.20913713917834684,
+      "step": 195,
+      "step_time": 27.840963185400142
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.1,
+      "completions/max_length": 145.6,
+      "completions/max_terminated_length": 108.0,
+      "completions/mean_length": 63.075,
+      "completions/mean_terminated_length": 43.47500076293945,
+      "completions/min_length": 16.4,
+      "completions/min_terminated_length": 16.4,
+      "entropy": 0.20764350929530337,
+      "epoch": 0.008620689655172414,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.4568965517241384e-06,
+      "loss": 0.01216265857219696,
+      "num_tokens": 473228.0,
+      "reward": 0.4162260115146637,
+      "reward_std": 0.5489736057817935,
+      "rewards/reward_fn/mean": 0.4162260115146637,
+      "rewards/reward_fn/std": 0.5489736313116736,
+      "step": 200,
+      "step_time": 24.10546086580016
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 55.6,
+      "completions/max_terminated_length": 55.6,
+      "completions/mean_length": 23.825,
+      "completions/mean_terminated_length": 23.825,
+      "completions/min_length": 16.0,
+      "completions/min_terminated_length": 16.0,
+      "entropy": 0.13016226456093136,
+      "epoch": 0.008836206896551724,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.4137931034482764e-06,
+      "loss": 0.10798157453536987,
+      "num_tokens": 483053.0,
+      "reward": 0.7742319941520691,
+      "reward_std": 0.06249405808048323,
+      "rewards/reward_fn/mean": 0.7742319941520691,
+      "rewards/reward_fn/std": 0.06249405494891107,
+      "step": 205,
+      "step_time": 12.142383367000003
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 104.8,
+      "completions/max_terminated_length": 104.8,
+      "completions/mean_length": 50.125,
+      "completions/mean_terminated_length": 50.125,
+      "completions/min_length": 22.4,
+      "completions/min_terminated_length": 22.4,
+      "entropy": 0.3128367411205545,
+      "epoch": 0.009051724137931035,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.370689655172414e-06,
+      "loss": -0.012002941966056824,
+      "num_tokens": 494054.0,
+      "reward": 0.31811599135398866,
+      "reward_std": 0.6536980773322284,
+      "rewards/reward_fn/mean": 0.31811599135398866,
+      "rewards/reward_fn/std": 0.6536980641074479,
+      "step": 210,
+      "step_time": 18.774766382600138
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 79.8,
+      "completions/max_terminated_length": 79.8,
+      "completions/mean_length": 54.7,
+      "completions/mean_terminated_length": 54.7,
+      "completions/min_length": 38.4,
+      "completions/min_terminated_length": 38.4,
+      "entropy": 0.18855432691198076,
+      "epoch": 0.009267241379310344,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.327586206896552e-06,
+      "loss": -0.01920826584100723,
+      "num_tokens": 505226.0,
+      "reward": 0.5115379929542542,
+      "reward_std": 0.15735746768768877,
+      "rewards/reward_fn/mean": 0.5115379929542542,
+      "rewards/reward_fn/std": 0.1573574845213443,
+      "step": 215,
+      "step_time": 15.221036481800548
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 126.2,
+      "completions/max_terminated_length": 126.2,
+      "completions/mean_length": 59.2,
+      "completions/mean_terminated_length": 59.2,
+      "completions/min_length": 16.2,
+      "completions/min_terminated_length": 16.2,
+      "entropy": 0.31272616144269705,
+      "epoch": 0.009482758620689655,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.28448275862069e-06,
+      "loss": 0.13071467876434326,
+      "num_tokens": 516730.0,
+      "reward": 0.6471440196037292,
+      "reward_std": 0.15736471198033541,
+      "rewards/reward_fn/mean": 0.6471440196037292,
+      "rewards/reward_fn/std": 0.15736472072312607,
+      "step": 220,
+      "step_time": 21.683129164199453
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 75.6,
+      "completions/max_terminated_length": 75.6,
+      "completions/mean_length": 41.075,
+      "completions/mean_terminated_length": 41.075,
+      "completions/min_length": 21.2,
+      "completions/min_terminated_length": 21.2,
+      "entropy": 0.17901114720734768,
+      "epoch": 0.009698275862068966,
+      "frac_reward_zero_std": 0.2,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.2413793103448277e-06,
+      "loss": 0.0042393475770950316,
+      "num_tokens": 527637.0,
+      "reward": 0.5594579905271531,
+      "reward_std": 0.23395470455288886,
+      "rewards/reward_fn/mean": 0.5594579905271531,
+      "rewards/reward_fn/std": 0.23395471236435697,
+      "step": 225,
+      "step_time": 14.985401752000143
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 145.0,
+      "completions/max_terminated_length": 144.0,
+      "completions/mean_length": 77.85,
+      "completions/mean_terminated_length": 73.91785736083985,
+      "completions/min_length": 29.8,
+      "completions/min_terminated_length": 29.8,
+      "entropy": 0.28966000496875494,
+      "epoch": 0.009913793103448277,
+      "frac_reward_zero_std": 0.3,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.1982758620689657e-06,
+      "loss": -0.03884480595588684,
+      "num_tokens": 539999.0,
+      "reward": 0.2210400015115738,
+      "reward_std": 0.45898649455048146,
+      "rewards/reward_fn/mean": 0.2210400015115738,
+      "rewards/reward_fn/std": 0.458986507600639,
+      "step": 230,
+      "step_time": 23.7599374403997
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 117.2,
+      "completions/max_terminated_length": 117.2,
+      "completions/mean_length": 50.25,
+      "completions/mean_terminated_length": 50.25,
+      "completions/min_length": 15.4,
+      "completions/min_terminated_length": 15.4,
+      "entropy": 0.2198769545648247,
+      "epoch": 0.010129310344827586,
+      "frac_reward_zero_std": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1551724137931037e-06,
+      "loss": -0.013682875037193298,
+      "num_tokens": 551073.0,
+      "reward": 0.4487079918384552,
+      "reward_std": 0.5221681177150458,
+      "rewards/reward_fn/mean": 0.4487079918384552,
+      "rewards/reward_fn/std": 0.5221681545488537,
+      "step": 235,
+      "step_time": 20.431722840200383
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 118.6,
+      "completions/max_terminated_length": 118.6,
+      "completions/mean_length": 54.675,
+      "completions/mean_terminated_length": 54.675,
+      "completions/min_length": 16.2,
+      "completions/min_terminated_length": 16.2,
+      "entropy": 0.20829223558394006,
+      "epoch": 0.010344827586206896,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.1120689655172413e-06,
+      "loss": -0.030032938718795775,
+      "num_tokens": 562420.0,
+      "reward": 0.17456801533699035,
+      "reward_std": 0.46873694062232973,
+      "rewards/reward_fn/mean": 0.17456801533699035,
+      "rewards/reward_fn/std": 0.4687369393184781,
+      "step": 240,
+      "step_time": 20.53446474100001
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 73.8,
+      "completions/max_terminated_length": 73.8,
+      "completions/mean_length": 42.475,
+      "completions/mean_terminated_length": 42.475,
+      "completions/min_length": 15.8,
+      "completions/min_terminated_length": 15.8,
+      "entropy": 0.22942446915694745,
+      "epoch": 0.010560344827586207,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0689655172413797e-06,
+      "loss": 0.043540936708450315,
+      "num_tokens": 572919.0,
+      "reward": 0.5922660022974015,
+      "reward_std": 0.26113789578084834,
+      "rewards/reward_fn/mean": 0.5922660022974015,
+      "rewards/reward_fn/std": 0.2611379227571888,
+      "step": 245,
+      "step_time": 14.755489835400112
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 57.4,
+      "completions/max_terminated_length": 57.4,
+      "completions/mean_length": 31.925,
+      "completions/mean_terminated_length": 31.925,
+      "completions/min_length": 16.0,
+      "completions/min_terminated_length": 16.0,
+      "entropy": 0.16531706769019366,
+      "epoch": 0.010775862068965518,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.0258620689655178e-06,
+      "loss": 0.02320311963558197,
+      "num_tokens": 583104.0,
+      "reward": 0.7694700002670288,
+      "reward_std": 0.03285302005242556,
+      "rewards/reward_fn/mean": 0.7694700002670288,
+      "rewards/reward_fn/std": 0.03285302061121911,
+      "step": 250,
+      "step_time": 12.29388190760037
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 600,
+  "num_input_tokens_seen": 583104,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "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": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-250/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cf50531ca359d39fbfa05dc0896e03532a83d53b029182b6d7f757efab0c97a
+size 7185