Upload task output 1

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,209 @@
+---
+base_model: None
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:/cache/models/Qwen--Qwen3-4B-Instruct-2507
+- 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 0.18.1

adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": null,
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "k_proj",
+    "q_proj",
+    "gate_proj",
+    "o_proj",
+    "v_proj",
+    "up_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f10923c51131df28cbbc3b2ebac46c8ea940027309c1f8f36b14a93c130c18
+size 264308896

added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

chat_template.jinja

@@ -0,0 +1,61 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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 %}

loss.txt

@@ -0,0 +1 @@
+53,no_eval

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

tokenizer_config.json

@@ -0,0 +1,239 @@
+{
+  "add_bos_token": false,
+  "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
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 1010000,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,1315 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.00053,
+  "eval_steps": 500,
+  "global_step": 53,
+  "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.48936397582292557,
+      "epoch": 1e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0005664514610543847,
+      "kl": 0.0,
+      "learning_rate": 0.0,
+      "loss": 0.0,
+      "num_tokens": 146997.0,
+      "reward": -3.158337116241455,
+      "reward_std": 11.52173900604248,
+      "rewards/rollout_reward_func/mean": -3.158337116241455,
+      "rewards/rollout_reward_func/std": 11.52173900604248,
+      "sampling/importance_sampling_ratio/max": 0.9331093430519104,
+      "sampling/importance_sampling_ratio/mean": 0.8982968330383301,
+      "sampling/importance_sampling_ratio/min": 0.8507044315338135,
+      "sampling/sampling_logp_difference/max": 0.15258322656154633,
+      "sampling/sampling_logp_difference/mean": 0.03600664064288139,
+      "step": 1,
+      "step_time": 78.72238844999993
+    },
+    {
+      "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,
+      "entropy": 0.48936397582292557,
+      "epoch": 2e-05,
+      "grad_norm": 0.0005677085719071329,
+      "kl": 0.0,
+      "learning_rate": 2.2857142857142855e-07,
+      "loss": 0.0,
+      "step": 2,
+      "step_time": 15.87692344900006
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.010416666977107525,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.48732541501522064,
+      "epoch": 3e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.002515916945412755,
+      "kl": 0.003698288830491947,
+      "learning_rate": 4.571428571428571e-07,
+      "loss": -0.0,
+      "num_tokens": 300422.0,
+      "reward": -3.6841847896575928,
+      "reward_std": 11.474555015563965,
+      "rewards/rollout_reward_func/mean": -3.6841847896575928,
+      "rewards/rollout_reward_func/std": 11.474555015563965,
+      "sampling/importance_sampling_ratio/max": 0.9328096508979797,
+      "sampling/importance_sampling_ratio/mean": 0.8980345726013184,
+      "sampling/importance_sampling_ratio/min": 0.8055793642997742,
+      "sampling/sampling_logp_difference/max": 0.14251333475112915,
+      "sampling/sampling_logp_difference/mean": 0.03639774024486542,
+      "step": 3,
+      "step_time": 80.83202714800018
+    },
+    {
+      "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,
+      "entropy": 0.47548339515924454,
+      "epoch": 4e-05,
+      "grad_norm": 0.0018481042934581637,
+      "kl": 0.0001756823347314196,
+      "learning_rate": 6.857142857142857e-07,
+      "loss": -0.0,
+      "step": 4,
+      "step_time": 16.098328456000672
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.875,
+      "completions/mean_terminated_length": 2.875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.5311543047428131,
+      "epoch": 5e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0028014357667416334,
+      "kl": 0.0004833272454334292,
+      "learning_rate": 9.142857142857142e-07,
+      "loss": 0.0001,
+      "num_tokens": 456990.0,
+      "reward": -1.6541011333465576,
+      "reward_std": 11.05872631072998,
+      "rewards/rollout_reward_func/mean": -1.6541011333465576,
+      "rewards/rollout_reward_func/std": 11.058727264404297,
+      "sampling/importance_sampling_ratio/max": 0.9697506427764893,
+      "sampling/importance_sampling_ratio/mean": 0.8795021772384644,
+      "sampling/importance_sampling_ratio/min": 0.5391229391098022,
+      "sampling/sampling_logp_difference/max": 0.4758959412574768,
+      "sampling/sampling_logp_difference/mean": 0.04838065803050995,
+      "step": 5,
+      "step_time": 86.28426266100041
+    },
+    {
+      "clip_ratio/high_max": 0.02500000037252903,
+      "clip_ratio/high_mean": 0.012500000186264515,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.012500000186264515,
+      "entropy": 0.5359498672187328,
+      "epoch": 6e-05,
+      "grad_norm": 0.0022959199268370867,
+      "kl": 0.002978163921227406,
+      "learning_rate": 1.1428571428571428e-06,
+      "loss": 0.0001,
+      "step": 6,
+      "step_time": 16.73348437599998
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.5112962238490582,
+      "epoch": 7e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.00931193120777607,
+      "kl": 5.794192431451961e-05,
+      "learning_rate": 1.3714285714285715e-06,
+      "loss": 0.0,
+      "num_tokens": 606759.0,
+      "reward": -1.4654988050460815,
+      "reward_std": 12.649130821228027,
+      "rewards/rollout_reward_func/mean": -1.4654988050460815,
+      "rewards/rollout_reward_func/std": 12.649128913879395,
+      "sampling/importance_sampling_ratio/max": 1.0540623664855957,
+      "sampling/importance_sampling_ratio/mean": 0.9033976197242737,
+      "sampling/importance_sampling_ratio/min": 0.8677928447723389,
+      "sampling/sampling_logp_difference/max": 0.17156457901000977,
+      "sampling/sampling_logp_difference/mean": 0.037677086889743805,
+      "step": 7,
+      "step_time": 80.9153498870005
+    },
+    {
+      "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,
+      "entropy": 0.5051367729902267,
+      "epoch": 8e-05,
+      "grad_norm": 0.004157723858952522,
+      "kl": 4.364301761228262e-05,
+      "learning_rate": 1.6e-06,
+      "loss": 0.0,
+      "step": 8,
+      "step_time": 16.163697053000305
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.875,
+      "completions/mean_terminated_length": 2.875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.5464204885065556,
+      "epoch": 9e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.003582538105547428,
+      "kl": 0.0011944683974434156,
+      "learning_rate": 1.8285714285714284e-06,
+      "loss": 0.0,
+      "num_tokens": 757249.0,
+      "reward": -7.697338104248047,
+      "reward_std": 8.61632251739502,
+      "rewards/rollout_reward_func/mean": -7.697338104248047,
+      "rewards/rollout_reward_func/std": 8.61632251739502,
+      "sampling/importance_sampling_ratio/max": 1.1465768814086914,
+      "sampling/importance_sampling_ratio/mean": 0.9066257476806641,
+      "sampling/importance_sampling_ratio/min": 0.8192122578620911,
+      "sampling/sampling_logp_difference/max": 0.23310279846191406,
+      "sampling/sampling_logp_difference/mean": 0.042134545743465424,
+      "step": 9,
+      "step_time": 87.67375510900001
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.012500000186264515,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.012500000186264515,
+      "entropy": 0.5470796823501587,
+      "epoch": 0.0001,
+      "grad_norm": 0.004420289304107428,
+      "kl": 0.0015031041582567184,
+      "learning_rate": 2.057142857142857e-06,
+      "loss": -0.0,
+      "step": 10,
+      "step_time": 16.138566152999374
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.010416666977107525,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.4877597391605377,
+      "epoch": 0.00011,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0015756486682221293,
+      "kl": 0.0009349783420873337,
+      "learning_rate": 2.2857142857142856e-06,
+      "loss": 0.0,
+      "num_tokens": 910271.0,
+      "reward": -1.2807047367095947,
+      "reward_std": 11.581576347351074,
+      "rewards/rollout_reward_func/mean": -1.2807047367095947,
+      "rewards/rollout_reward_func/std": 11.581576347351074,
+      "sampling/importance_sampling_ratio/max": 1.075101375579834,
+      "sampling/importance_sampling_ratio/mean": 0.9127196669578552,
+      "sampling/importance_sampling_ratio/min": 0.7523821592330933,
+      "sampling/sampling_logp_difference/max": 0.17058521509170532,
+      "sampling/sampling_logp_difference/mean": 0.04037006199359894,
+      "step": 11,
+      "step_time": 84.74441413100021
+    },
+    {
+      "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,
+      "entropy": 0.48288414254784584,
+      "epoch": 0.00012,
+      "grad_norm": 0.0031028895173221827,
+      "kl": 0.00019705418833382282,
+      "learning_rate": 2.5142857142857142e-06,
+      "loss": 0.0,
+      "step": 12,
+      "step_time": 16.036360359999435
+    },
+    {
+      "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": 4.0,
+      "completions/max_terminated_length": 4.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.49491142481565475,
+      "epoch": 0.00013,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0027884214650839567,
+      "kl": 0.0024777476012687316,
+      "learning_rate": 2.742857142857143e-06,
+      "loss": 0.0,
+      "num_tokens": 1049539.0,
+      "reward": -7.930685997009277,
+      "reward_std": 10.647638320922852,
+      "rewards/rollout_reward_func/mean": -7.930685997009277,
+      "rewards/rollout_reward_func/std": 10.647637367248535,
+      "sampling/importance_sampling_ratio/max": 1.1072638034820557,
+      "sampling/importance_sampling_ratio/mean": 0.9023596048355103,
+      "sampling/importance_sampling_ratio/min": 0.5414906740188599,
+      "sampling/sampling_logp_difference/max": 0.4979434013366699,
+      "sampling/sampling_logp_difference/mean": 0.0404733270406723,
+      "step": 13,
+      "step_time": 80.93550676899986
+    },
+    {
+      "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,
+      "entropy": 0.503012377768755,
+      "epoch": 0.00014,
+      "grad_norm": 0.002723120851442218,
+      "kl": 0.00012674138329771267,
+      "learning_rate": 2.9714285714285716e-06,
+      "loss": 0.0,
+      "step": 14,
+      "step_time": 16.415520546000607
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.5103708207607269,
+      "epoch": 0.00015,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0012367322342470288,
+      "kl": 0.0009814446365226104,
+      "learning_rate": 3.2e-06,
+      "loss": -0.0,
+      "num_tokens": 1198467.0,
+      "reward": -1.7910524606704712,
+      "reward_std": 14.345645904541016,
+      "rewards/rollout_reward_func/mean": -1.7910524606704712,
+      "rewards/rollout_reward_func/std": 14.345645904541016,
+      "sampling/importance_sampling_ratio/max": 0.9826624393463135,
+      "sampling/importance_sampling_ratio/mean": 0.9046387672424316,
+      "sampling/importance_sampling_ratio/min": 0.8640578985214233,
+      "sampling/sampling_logp_difference/max": 0.1460454910993576,
+      "sampling/sampling_logp_difference/mean": 0.03705396130681038,
+      "step": 15,
+      "step_time": 84.01811012399958
+    },
+    {
+      "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,
+      "entropy": 0.5053780749440193,
+      "epoch": 0.00016,
+      "grad_norm": 0.0012754682684317231,
+      "kl": 0.0003874969786465954,
+      "learning_rate": 3.428571428571428e-06,
+      "loss": -0.0,
+      "step": 16,
+      "step_time": 16.14125642999943
+    },
+    {
+      "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": 4.0,
+      "completions/max_terminated_length": 4.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.4758913367986679,
+      "epoch": 0.00017,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.000835731509141624,
+      "kl": 2.2824605100169038e-05,
+      "learning_rate": 3.657142857142857e-06,
+      "loss": -0.0,
+      "num_tokens": 1352552.0,
+      "reward": -2.025902271270752,
+      "reward_std": 9.88846492767334,
+      "rewards/rollout_reward_func/mean": -2.025902271270752,
+      "rewards/rollout_reward_func/std": 9.88846492767334,
+      "sampling/importance_sampling_ratio/max": 0.9374787211418152,
+      "sampling/importance_sampling_ratio/mean": 0.9018987417221069,
+      "sampling/importance_sampling_ratio/min": 0.8059737086296082,
+      "sampling/sampling_logp_difference/max": 0.1600051373243332,
+      "sampling/sampling_logp_difference/mean": 0.034846384078264236,
+      "step": 17,
+      "step_time": 87.90823437400013
+    },
+    {
+      "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,
+      "entropy": 0.47047749534249306,
+      "epoch": 0.00018,
+      "grad_norm": 0.000915601325687021,
+      "kl": 3.488661536721338e-05,
+      "learning_rate": 3.885714285714286e-06,
+      "loss": -0.0,
+      "step": 18,
+      "step_time": 16.648166115000095
+    },
+    {
+      "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": 4.0,
+      "completions/max_terminated_length": 4.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.4496545232832432,
+      "epoch": 0.00019,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0004858207830693573,
+      "kl": 3.536299266215792e-05,
+      "learning_rate": 4.114285714285714e-06,
+      "loss": 0.0,
+      "num_tokens": 1501094.0,
+      "reward": -3.7144558429718018,
+      "reward_std": 11.072718620300293,
+      "rewards/rollout_reward_func/mean": -3.7144558429718018,
+      "rewards/rollout_reward_func/std": 11.072717666625977,
+      "sampling/importance_sampling_ratio/max": 0.9308063983917236,
+      "sampling/importance_sampling_ratio/mean": 0.899206817150116,
+      "sampling/importance_sampling_ratio/min": 0.66890549659729,
+      "sampling/sampling_logp_difference/max": 0.32890522480010986,
+      "sampling/sampling_logp_difference/mean": 0.03617248684167862,
+      "step": 19,
+      "step_time": 80.7782728230004
+    },
+    {
+      "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,
+      "entropy": 0.44183528050780296,
+      "epoch": 0.0002,
+      "grad_norm": 0.00042393436888232827,
+      "kl": 0.00015307544185816369,
+      "learning_rate": 4.342857142857142e-06,
+      "loss": 0.0,
+      "step": 20,
+      "step_time": 16.09888975399963
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.5056380964815617,
+      "epoch": 0.00021,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.00032198833650909364,
+      "kl": 4.869016663633374e-05,
+      "learning_rate": 4.571428571428571e-06,
+      "loss": -0.0,
+      "num_tokens": 1651955.0,
+      "reward": -1.5917177200317383,
+      "reward_std": 13.17538070678711,
+      "rewards/rollout_reward_func/mean": -1.5917177200317383,
+      "rewards/rollout_reward_func/std": 13.175381660461426,
+      "sampling/importance_sampling_ratio/max": 0.9360762238502502,
+      "sampling/importance_sampling_ratio/mean": 0.9005730152130127,
+      "sampling/importance_sampling_ratio/min": 0.5624815821647644,
+      "sampling/sampling_logp_difference/max": 0.6438815593719482,
+      "sampling/sampling_logp_difference/mean": 0.038462258875370026,
+      "step": 21,
+      "step_time": 86.13508222099972
+    },
+    {
+      "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,
+      "entropy": 0.4966563805937767,
+      "epoch": 0.00022,
+      "grad_norm": 0.0003532824048306793,
+      "kl": 8.798080261840369e-05,
+      "learning_rate": 4.8e-06,
+      "loss": -0.0,
+      "step": 22,
+      "step_time": 16.589822670999638
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.46145135909318924,
+      "epoch": 0.00023,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0017623284365981817,
+      "kl": 8.294408689835109e-05,
+      "learning_rate": 5.0285714285714285e-06,
+      "loss": 0.0,
+      "num_tokens": 1801431.0,
+      "reward": -3.1079812049865723,
+      "reward_std": 15.382396697998047,
+      "rewards/rollout_reward_func/mean": -3.1079812049865723,
+      "rewards/rollout_reward_func/std": 15.382396697998047,
+      "sampling/importance_sampling_ratio/max": 0.9533551335334778,
+      "sampling/importance_sampling_ratio/mean": 0.9110425710678101,
+      "sampling/importance_sampling_ratio/min": 0.8689562678337097,
+      "sampling/sampling_logp_difference/max": 0.12607532739639282,
+      "sampling/sampling_logp_difference/mean": 0.031991466879844666,
+      "step": 23,
+      "step_time": 84.70673029099953
+    },
+    {
+      "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,
+      "entropy": 0.4546354450285435,
+      "epoch": 0.00024,
+      "grad_norm": 0.002184377983212471,
+      "kl": 0.00013984292854729574,
+      "learning_rate": 5.257142857142857e-06,
+      "loss": 0.0,
+      "step": 24,
+      "step_time": 15.989325058999839
+    },
+    {
+      "clip_ratio/high_max": 0.02083333395421505,
+      "clip_ratio/high_mean": 0.010416666977107525,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.50193877145648,
+      "epoch": 0.00025,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0028643528930842876,
+      "kl": 0.0006277600778048509,
+      "learning_rate": 5.485714285714286e-06,
+      "loss": -0.0,
+      "num_tokens": 1954111.0,
+      "reward": -1.1967151165008545,
+      "reward_std": 9.904693603515625,
+      "rewards/rollout_reward_func/mean": -1.1967151165008545,
+      "rewards/rollout_reward_func/std": 9.904692649841309,
+      "sampling/importance_sampling_ratio/max": 1.0367523431777954,
+      "sampling/importance_sampling_ratio/mean": 0.9070800542831421,
+      "sampling/importance_sampling_ratio/min": 0.641558825969696,
+      "sampling/sampling_logp_difference/max": 0.2708436846733093,
+      "sampling/sampling_logp_difference/mean": 0.03570127859711647,
+      "step": 25,
+      "step_time": 88.23927604999926
+    },
+    {
+      "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,
+      "entropy": 0.493778470903635,
+      "epoch": 0.00026,
+      "grad_norm": 0.002861190587282181,
+      "kl": 0.006509271008326323,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": -0.0,
+      "step": 26,
+      "step_time": 17.19145845399953
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.90625,
+      "completions/mean_terminated_length": 2.90625,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.4369519129395485,
+      "epoch": 0.00027,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0023894549813121557,
+      "kl": 0.00021573404501395999,
+      "learning_rate": 5.942857142857143e-06,
+      "loss": 0.0,
+      "num_tokens": 2103766.0,
+      "reward": -4.149303436279297,
+      "reward_std": 11.891597747802734,
+      "rewards/rollout_reward_func/mean": -4.149303436279297,
+      "rewards/rollout_reward_func/std": 11.891597747802734,
+      "sampling/importance_sampling_ratio/max": 0.9366415739059448,
+      "sampling/importance_sampling_ratio/mean": 0.9109276533126831,
+      "sampling/importance_sampling_ratio/min": 0.7386661171913147,
+      "sampling/sampling_logp_difference/max": 0.17405521869659424,
+      "sampling/sampling_logp_difference/mean": 0.03263023495674133,
+      "step": 27,
+      "step_time": 84.90730395599985
+    },
+    {
+      "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,
+      "entropy": 0.4287884756922722,
+      "epoch": 0.00028,
+      "grad_norm": 0.003158987034112215,
+      "kl": 0.0010838782172868378,
+      "learning_rate": 6.171428571428571e-06,
+      "loss": 0.0,
+      "step": 28,
+      "step_time": 16.073362338999686
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.90625,
+      "completions/mean_terminated_length": 2.90625,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.47590041905641556,
+      "epoch": 0.00029,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0019009847892448306,
+      "kl": 0.002789211330764374,
+      "learning_rate": 6.4e-06,
+      "loss": -0.0,
+      "num_tokens": 2255958.0,
+      "reward": -1.561665415763855,
+      "reward_std": 11.948278427124023,
+      "rewards/rollout_reward_func/mean": -1.561665415763855,
+      "rewards/rollout_reward_func/std": 11.948278427124023,
+      "sampling/importance_sampling_ratio/max": 0.951849639415741,
+      "sampling/importance_sampling_ratio/mean": 0.8972554802894592,
+      "sampling/importance_sampling_ratio/min": 0.328483909368515,
+      "sampling/sampling_logp_difference/max": 1.0063753128051758,
+      "sampling/sampling_logp_difference/mean": 0.04340764880180359,
+      "step": 29,
+      "step_time": 91.08275033600057
+    },
+    {
+      "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,
+      "entropy": 0.47105172276496887,
+      "epoch": 0.0003,
+      "grad_norm": 0.002664634957909584,
+      "kl": 0.0012541578576019674,
+      "learning_rate": 6.628571428571428e-06,
+      "loss": -0.0,
+      "step": 30,
+      "step_time": 16.112164762000702
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.3540297709405422,
+      "epoch": 0.00031,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.00012217683251947165,
+      "kl": 0.00030874026197125204,
+      "learning_rate": 6.857142857142856e-06,
+      "loss": -0.0,
+      "num_tokens": 2407027.0,
+      "reward": 0.8819369077682495,
+      "reward_std": 9.408540725708008,
+      "rewards/rollout_reward_func/mean": 0.8819369077682495,
+      "rewards/rollout_reward_func/std": 9.408540725708008,
+      "sampling/importance_sampling_ratio/max": 0.9496375322341919,
+      "sampling/importance_sampling_ratio/mean": 0.926586389541626,
+      "sampling/importance_sampling_ratio/min": 0.8927135467529297,
+      "sampling/sampling_logp_difference/max": 0.11497166007757187,
+      "sampling/sampling_logp_difference/mean": 0.025536321103572845,
+      "step": 31,
+      "step_time": 87.01734963399849
+    },
+    {
+      "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,
+      "entropy": 0.34124791249632835,
+      "epoch": 0.00032,
+      "grad_norm": 0.00012294144835323095,
+      "kl": 0.0003481047842797125,
+      "learning_rate": 7.085714285714285e-06,
+      "loss": -0.0,
+      "step": 32,
+      "step_time": 16.119427530000394
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.37663908302783966,
+      "epoch": 0.00033,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.00028033516719006,
+      "kl": 0.00011349991063980269,
+      "learning_rate": 7.314285714285714e-06,
+      "loss": 0.0,
+      "num_tokens": 2556831.0,
+      "reward": -5.249267101287842,
+      "reward_std": 9.296294212341309,
+      "rewards/rollout_reward_func/mean": -5.249267101287842,
+      "rewards/rollout_reward_func/std": 9.296294212341309,
+      "sampling/importance_sampling_ratio/max": 0.9602704644203186,
+      "sampling/importance_sampling_ratio/mean": 0.9252135753631592,
+      "sampling/importance_sampling_ratio/min": 0.8558757305145264,
+      "sampling/sampling_logp_difference/max": 0.1554594784975052,
+      "sampling/sampling_logp_difference/mean": 0.027311213314533234,
+      "step": 33,
+      "step_time": 91.05606818099977
+    },
+    {
+      "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,
+      "entropy": 0.37676023691892624,
+      "epoch": 0.00034,
+      "grad_norm": 0.0002865525020752102,
+      "kl": 0.00011139449998154305,
+      "learning_rate": 7.542857142857142e-06,
+      "loss": 0.0,
+      "step": 34,
+      "step_time": 16.66090228600069
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.010416666977107525,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.38403602689504623,
+      "epoch": 0.00035,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0007655026274733245,
+      "kl": 0.0008573867726227036,
+      "learning_rate": 7.771428571428572e-06,
+      "loss": 0.0,
+      "num_tokens": 2711341.0,
+      "reward": -3.2598910331726074,
+      "reward_std": 11.653153419494629,
+      "rewards/rollout_reward_func/mean": -3.2598910331726074,
+      "rewards/rollout_reward_func/std": 11.653153419494629,
+      "sampling/importance_sampling_ratio/max": 1.3437409400939941,
+      "sampling/importance_sampling_ratio/mean": 0.9472247362136841,
+      "sampling/importance_sampling_ratio/min": 0.872779130935669,
+      "sampling/sampling_logp_difference/max": 0.351446270942688,
+      "sampling/sampling_logp_difference/mean": 0.03257140517234802,
+      "step": 35,
+      "step_time": 97.56342971100003
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.010416666977107525,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "entropy": 0.37373727560043335,
+      "epoch": 0.00036,
+      "grad_norm": 0.0005624265759252012,
+      "kl": 0.0011794225774792721,
+      "learning_rate": 8e-06,
+      "loss": 0.0,
+      "step": 36,
+      "step_time": 16.08771912999964
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 3.0,
+      "completions/mean_terminated_length": 3.0,
+      "completions/min_length": 3.0,
+      "completions/min_terminated_length": 3.0,
+      "entropy": 0.3612252026796341,
+      "epoch": 0.00037,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.003778834594413638,
+      "kl": 0.005491923870067694,
+      "learning_rate": 7.999999999958871e-06,
+      "loss": 0.0,
+      "num_tokens": 2860979.0,
+      "reward": -3.204911947250366,
+      "reward_std": 13.321104049682617,
+      "rewards/rollout_reward_func/mean": -3.204911947250366,
+      "rewards/rollout_reward_func/std": 13.3211030960083,
+      "sampling/importance_sampling_ratio/max": 0.9534703493118286,
+      "sampling/importance_sampling_ratio/mean": 0.9272962808609009,
+      "sampling/importance_sampling_ratio/min": 0.812048077583313,
+      "sampling/sampling_logp_difference/max": 0.125631183385849,
+      "sampling/sampling_logp_difference/mean": 0.027381815016269684,
+      "step": 37,
+      "step_time": 93.92803074699941
+    },
+    {
+      "clip_ratio/high_max": 0.0416666679084301,
+      "clip_ratio/high_mean": 0.02083333395421505,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.02083333395421505,
+      "entropy": 0.3561956323683262,
+      "epoch": 0.00038,
+      "grad_norm": 0.00041312171379104257,
+      "kl": 0.009066745038580848,
+      "learning_rate": 7.999999999835487e-06,
+      "loss": -0.0,
+      "step": 38,
+      "step_time": 16.57034048200012
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.90625,
+      "completions/mean_terminated_length": 2.90625,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.37458185851573944,
+      "epoch": 0.00039,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.003013370558619499,
+      "kl": 0.0015790614006618853,
+      "learning_rate": 7.999999999629846e-06,
+      "loss": -0.0,
+      "num_tokens": 3014870.0,
+      "reward": -4.495431900024414,
+      "reward_std": 9.237488746643066,
+      "rewards/rollout_reward_func/mean": -4.495431900024414,
+      "rewards/rollout_reward_func/std": 9.237488746643066,
+      "sampling/importance_sampling_ratio/max": 0.955286979675293,
+      "sampling/importance_sampling_ratio/mean": 0.9233821630477905,
+      "sampling/importance_sampling_ratio/min": 0.7778754234313965,
+      "sampling/sampling_logp_difference/max": 0.19711723923683167,
+      "sampling/sampling_logp_difference/mean": 0.028256624937057495,
+      "step": 39,
+      "step_time": 95.70763315500017
+    },
+    {
+      "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,
+      "entropy": 0.3661472350358963,
+      "epoch": 0.0004,
+      "grad_norm": 0.0022348016500473022,
+      "kl": 0.0027103804022772238,
+      "learning_rate": 7.99999999934195e-06,
+      "loss": -0.0,
+      "step": 40,
+      "step_time": 16.061642451000353
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.28179030306637287,
+      "epoch": 0.00041,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 6.525115895783529e-05,
+      "kl": 0.000415502121541067,
+      "learning_rate": 7.999999998971795e-06,
+      "loss": -0.0,
+      "num_tokens": 3164817.0,
+      "reward": 2.3517534732818604,
+      "reward_std": 12.381294250488281,
+      "rewards/rollout_reward_func/mean": 2.3517534732818604,
+      "rewards/rollout_reward_func/std": 12.381294250488281,
+      "sampling/importance_sampling_ratio/max": 0.9573846459388733,
+      "sampling/importance_sampling_ratio/mean": 0.9431747198104858,
+      "sampling/importance_sampling_ratio/min": 0.9145121574401855,
+      "sampling/sampling_logp_difference/max": 0.08903493732213974,
+      "sampling/sampling_logp_difference/mean": 0.01973060891032219,
+      "step": 41,
+      "step_time": 90.856857966
+    },
+    {
+      "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,
+      "entropy": 0.27641875483095646,
+      "epoch": 0.00042,
+      "grad_norm": 0.0001371208782074973,
+      "kl": 0.0004361617029644549,
+      "learning_rate": 7.999999998519386e-06,
+      "loss": -0.0,
+      "step": 42,
+      "step_time": 16.129788360000475
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.90625,
+      "completions/mean_terminated_length": 2.90625,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.32416083104908466,
+      "epoch": 0.00043,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0014907444128766656,
+      "kl": 0.0035448235757939983,
+      "learning_rate": 7.99999999798472e-06,
+      "loss": 0.0,
+      "num_tokens": 3317992.0,
+      "reward": 0.9361296892166138,
+      "reward_std": 10.966468811035156,
+      "rewards/rollout_reward_func/mean": 0.9361296892166138,
+      "rewards/rollout_reward_func/std": 10.966468811035156,
+      "sampling/importance_sampling_ratio/max": 1.039656639099121,
+      "sampling/importance_sampling_ratio/mean": 0.9434545636177063,
+      "sampling/importance_sampling_ratio/min": 0.8878768682479858,
+      "sampling/sampling_logp_difference/max": 0.11892110854387283,
+      "sampling/sampling_logp_difference/mean": 0.022497989237308502,
+      "step": 43,
+      "step_time": 92.78864051999926
+    },
+    {
+      "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,
+      "entropy": 0.3210602290928364,
+      "epoch": 0.00044,
+      "grad_norm": 0.0009183208458125591,
+      "kl": 0.0044001892892993055,
+      "learning_rate": 7.999999997367799e-06,
+      "loss": -0.0,
+      "step": 44,
+      "step_time": 16.18974080999942
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.2824931964278221,
+      "epoch": 0.00045,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.005496473051607609,
+      "kl": 0.09349646368718822,
+      "learning_rate": 7.999999996668619e-06,
+      "loss": 0.0,
+      "num_tokens": 3472410.0,
+      "reward": 1.872360348701477,
+      "reward_std": 12.642463684082031,
+      "rewards/rollout_reward_func/mean": 1.872360348701477,
+      "rewards/rollout_reward_func/std": 12.642463684082031,
+      "sampling/importance_sampling_ratio/max": 0.9575082063674927,
+      "sampling/importance_sampling_ratio/mean": 0.941848635673523,
+      "sampling/importance_sampling_ratio/min": 0.8221862316131592,
+      "sampling/sampling_logp_difference/max": 0.12410736083984375,
+      "sampling/sampling_logp_difference/mean": 0.0203759353607893,
+      "step": 45,
+      "step_time": 92.56621814599839
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.010416666977107525,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "entropy": 0.2739644553512335,
+      "epoch": 0.00046,
+      "grad_norm": 0.0008409882429987192,
+      "kl": 0.17168781322834548,
+      "learning_rate": 7.999999995887185e-06,
+      "loss": 0.0,
+      "step": 46,
+      "step_time": 16.97712109299937
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.90625,
+      "completions/mean_terminated_length": 2.90625,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.3127005323767662,
+      "epoch": 0.00047,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.005147633608430624,
+      "kl": 0.02404813828979968,
+      "learning_rate": 7.999999995023493e-06,
+      "loss": 0.0,
+      "num_tokens": 3623971.0,
+      "reward": -2.005354404449463,
+      "reward_std": 9.463374137878418,
+      "rewards/rollout_reward_func/mean": -2.005354404449463,
+      "rewards/rollout_reward_func/std": 9.463374137878418,
+      "sampling/importance_sampling_ratio/max": 1.8077476024627686,
+      "sampling/importance_sampling_ratio/mean": 0.9680489897727966,
+      "sampling/importance_sampling_ratio/min": 0.8782174587249756,
+      "sampling/sampling_logp_difference/max": 0.6487047672271729,
+      "sampling/sampling_logp_difference/mean": 0.027968257665634155,
+      "step": 47,
+      "step_time": 90.81133413299995
+    },
+    {
+      "clip_ratio/high_max": 0.02083333395421505,
+      "clip_ratio/high_mean": 0.010416666977107525,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.010416666977107525,
+      "entropy": 0.31450361758470535,
+      "epoch": 0.00048,
+      "grad_norm": 0.010047705844044685,
+      "kl": 0.011364400590537116,
+      "learning_rate": 7.999999994077545e-06,
+      "loss": -0.0,
+      "step": 48,
+      "step_time": 16.131004170999404
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.31293279118835926,
+      "epoch": 0.00049,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0025514012668281794,
+      "kl": 0.004930681967380224,
+      "learning_rate": 7.999999993049342e-06,
+      "loss": -0.0,
+      "num_tokens": 3769693.0,
+      "reward": -1.6621102094650269,
+      "reward_std": 11.216950416564941,
+      "rewards/rollout_reward_func/mean": -1.6621102094650269,
+      "rewards/rollout_reward_func/std": 11.216950416564941,
+      "sampling/importance_sampling_ratio/max": 0.963409960269928,
+      "sampling/importance_sampling_ratio/mean": 0.9157481789588928,
+      "sampling/importance_sampling_ratio/min": 0.4775127172470093,
+      "sampling/sampling_logp_difference/max": 0.6763087511062622,
+      "sampling/sampling_logp_difference/mean": 0.03370527923107147,
+      "step": 49,
+      "step_time": 87.11181479400057
+    },
+    {
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.02083333395421505,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.02083333395421505,
+      "entropy": 0.305842787027359,
+      "epoch": 0.0005,
+      "grad_norm": 0.001049969345331192,
+      "kl": 0.01695527902484173,
+      "learning_rate": 7.999999991938882e-06,
+      "loss": -0.0,
+      "step": 50,
+      "step_time": 17.37328599800003
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.9375,
+      "completions/mean_terminated_length": 2.9375,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.26627207547426224,
+      "epoch": 0.00051,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.000988983316347003,
+      "kl": 0.0010694206866901368,
+      "learning_rate": 7.999999990746166e-06,
+      "loss": -0.0,
+      "num_tokens": 3918993.0,
+      "reward": 0.07662695646286011,
+      "reward_std": 12.704819679260254,
+      "rewards/rollout_reward_func/mean": 0.07662695646286011,
+      "rewards/rollout_reward_func/std": 12.704818725585938,
+      "sampling/importance_sampling_ratio/max": 0.9753377437591553,
+      "sampling/importance_sampling_ratio/mean": 0.950916051864624,
+      "sampling/importance_sampling_ratio/min": 0.8742276430130005,
+      "sampling/sampling_logp_difference/max": 0.08313722163438797,
+      "sampling/sampling_logp_difference/mean": 0.018695753067731857,
+      "step": 51,
+      "step_time": 84.65364402499836
+    },
+    {
+      "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,
+      "entropy": 0.26170745119452477,
+      "epoch": 0.00052,
+      "grad_norm": 0.0009646639809943736,
+      "kl": 0.0012894975807284936,
+      "learning_rate": 7.999999989471194e-06,
+      "loss": -0.0,
+      "step": 52,
+      "step_time": 15.9307113390023
+    },
+    {
+      "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": 3.0,
+      "completions/max_terminated_length": 3.0,
+      "completions/mean_length": 2.96875,
+      "completions/mean_terminated_length": 2.96875,
+      "completions/min_length": 2.0,
+      "completions/min_terminated_length": 2.0,
+      "entropy": 0.26491483487188816,
+      "epoch": 0.00053,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.0002617822028696537,
+      "kl": 0.0023044159715936985,
+      "learning_rate": 7.999999988113964e-06,
+      "loss": -0.0,
+      "num_tokens": 4071645.0,
+      "reward": -0.05424162745475769,
+      "reward_std": 9.241598129272461,
+      "rewards/rollout_reward_func/mean": -0.05424162745475769,
+      "rewards/rollout_reward_func/std": 9.241597175598145,
+      "sampling/importance_sampling_ratio/max": 0.9647132754325867,
+      "sampling/importance_sampling_ratio/mean": 0.9405180215835571,
+      "sampling/importance_sampling_ratio/min": 0.6948662400245667,
+      "sampling/sampling_logp_difference/max": 0.30833685398101807,
+      "sampling/sampling_logp_difference/mean": 0.021141711622476578,
+      "step": 53,
+      "step_time": 93.0931882169998
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 600000,
+  "num_input_tokens_seen": 4071645,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2010939fa76a3f4dbd78aaab9e103d580462957e8424de882b2bd2c6f761ef2c
+size 8081