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--Qwen2.5-3B-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 0.18.1

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": [
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "down_proj",
+    "gate_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:e2f51df241da488414daa3e56f59f8c968ea91e605fdf1908eaa9786fe9e28ca
+size 239536272

added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|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 +1,54 @@
+{%- 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 %}

loss.txt

@@ -0,0 +1 @@
+50,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 sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "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
+    }
+  },
+  "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": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,1234 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.0005,
+  "eval_steps": 500,
+  "global_step": 50,
+  "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": 9365.0,
+      "completions/max_terminated_length": 9365.0,
+      "completions/mean_length": 8020.65625,
+      "completions/mean_terminated_length": 8020.65625,
+      "completions/min_length": 4249.0,
+      "completions/min_terminated_length": 4249.0,
+      "entropy": 0.09263912006281316,
+      "epoch": 1e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.3079737424850464,
+      "kl": 0.0,
+      "learning_rate": 0.0,
+      "loss": -0.0363,
+      "num_tokens": 283668.0,
+      "reward": -0.4597676992416382,
+      "reward_std": 0.3370218873023987,
+      "rewards/rollout_reward_func/mean": -0.4597676992416382,
+      "rewards/rollout_reward_func/std": 0.37389153242111206,
+      "sampling/importance_sampling_ratio/max": 1.752305507659912,
+      "sampling/importance_sampling_ratio/mean": 0.9996651411056519,
+      "sampling/importance_sampling_ratio/min": 0.4149494767189026,
+      "sampling/sampling_logp_difference/max": 0.8795984983444214,
+      "sampling/sampling_logp_difference/mean": 0.009596828371286392,
+      "step": 1,
+      "step_time": 96.616219112002
+    },
+    {
+      "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": 0.09263912006281316,
+      "epoch": 2e-05,
+      "grad_norm": 1.297884225845337,
+      "kl": 0.0,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": -0.0363,
+      "step": 2,
+      "step_time": 28.28962996199698
+    },
+    {
+      "clip_ratio/high_max": 0.005740102380514145,
+      "clip_ratio/high_mean": 0.0028700511902570724,
+      "clip_ratio/low_mean": 0.0018344554118812084,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.004704506660345942,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9371.0,
+      "completions/max_terminated_length": 9371.0,
+      "completions/mean_length": 7904.6875,
+      "completions/mean_terminated_length": 7904.6875,
+      "completions/min_length": 2043.0,
+      "completions/min_terminated_length": 2043.0,
+      "entropy": 0.11998538533225656,
+      "epoch": 3e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.5289855003356934,
+      "kl": 0.0012301416445552604,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": -0.0457,
+      "num_tokens": 564087.0,
+      "reward": -0.5441880226135254,
+      "reward_std": 0.3924558162689209,
+      "rewards/rollout_reward_func/mean": -0.5441880226135254,
+      "rewards/rollout_reward_func/std": 0.42554038763046265,
+      "sampling/importance_sampling_ratio/max": 1.5868604183197021,
+      "sampling/importance_sampling_ratio/mean": 0.9994746446609497,
+      "sampling/importance_sampling_ratio/min": 0.5226752758026123,
+      "sampling/sampling_logp_difference/max": 0.6487948894500732,
+      "sampling/sampling_logp_difference/mean": 0.011393455788493156,
+      "step": 3,
+      "step_time": 96.45885010200436
+    },
+    {
+      "clip_ratio/high_max": 0.005765739129856229,
+      "clip_ratio/high_mean": 0.0028828695649281144,
+      "clip_ratio/low_mean": 0.0021901361469645053,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.00507300574099645,
+      "entropy": 0.12003705860115588,
+      "epoch": 4e-05,
+      "grad_norm": 1.2704079151153564,
+      "kl": 0.001491897305186285,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": -0.047,
+      "step": 4,
+      "step_time": 29.195899179991102
+    },
+    {
+      "clip_ratio/high_max": 0.005963329633232206,
+      "clip_ratio/high_mean": 0.002981664816616103,
+      "clip_ratio/low_mean": 0.0014880952658131719,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.004469760082429275,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 8882.0,
+      "completions/max_terminated_length": 8882.0,
+      "completions/mean_length": 7734.1875,
+      "completions/mean_terminated_length": 7734.1875,
+      "completions/min_length": 1783.0,
+      "completions/min_terminated_length": 1783.0,
+      "entropy": 0.09044346120208502,
+      "epoch": 5e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.5453312397003174,
+      "kl": 0.0007876164445406175,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": -0.0298,
+      "num_tokens": 838350.0,
+      "reward": -0.5028736591339111,
+      "reward_std": 0.371159166097641,
+      "rewards/rollout_reward_func/mean": -0.5028736591339111,
+      "rewards/rollout_reward_func/std": 0.38670846819877625,
+      "sampling/importance_sampling_ratio/max": 1.758213996887207,
+      "sampling/importance_sampling_ratio/mean": 0.9994957447052002,
+      "sampling/importance_sampling_ratio/min": 0.3028993308544159,
+      "sampling/sampling_logp_difference/max": 1.194354772567749,
+      "sampling/sampling_logp_difference/mean": 0.009732999838888645,
+      "step": 5,
+      "step_time": 93.71373151599619
+    },
+    {
+      "clip_ratio/high_max": 0.006641630490776151,
+      "clip_ratio/high_mean": 0.004027052520541474,
+      "clip_ratio/low_mean": 0.002363648993195966,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.00639070151373744,
+      "entropy": 0.08915826864540577,
+      "epoch": 6e-05,
+      "grad_norm": 1.28467857837677,
+      "kl": 0.0010278278514306294,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": -0.0305,
+      "step": 6,
+      "step_time": 27.6442580659932
+    },
+    {
+      "clip_ratio/high_max": 0.006255197920836508,
+      "clip_ratio/high_mean": 0.003127598960418254,
+      "clip_ratio/low_mean": 0.0010739664430730045,
+      "clip_ratio/low_min": 0.0006944444612599909,
+      "clip_ratio/region_mean": 0.004201565403491259,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9327.0,
+      "completions/max_terminated_length": 9327.0,
+      "completions/mean_length": 7944.875,
+      "completions/mean_terminated_length": 7944.875,
+      "completions/min_length": 1618.0,
+      "completions/min_terminated_length": 1618.0,
+      "entropy": 0.0936605577589944,
+      "epoch": 7e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.6153287887573242,
+      "kl": 0.001322501270351495,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": -0.04,
+      "num_tokens": 1120382.0,
+      "reward": -0.40193918347358704,
+      "reward_std": 0.42798829078674316,
+      "rewards/rollout_reward_func/mean": -0.40193918347358704,
+      "rewards/rollout_reward_func/std": 0.45951688289642334,
+      "sampling/importance_sampling_ratio/max": 2.626508951187134,
+      "sampling/importance_sampling_ratio/mean": 0.9997774362564087,
+      "sampling/importance_sampling_ratio/min": 0.4664866328239441,
+      "sampling/sampling_logp_difference/max": 0.9656555652618408,
+      "sampling/sampling_logp_difference/mean": 0.011292756535112858,
+      "step": 7,
+      "step_time": 93.59508053499303
+    },
+    {
+      "clip_ratio/high_max": 0.004923702392261475,
+      "clip_ratio/high_mean": 0.0028252233169041574,
+      "clip_ratio/low_mean": 0.006011958757881075,
+      "clip_ratio/low_min": 0.0007102272938936949,
+      "clip_ratio/region_mean": 0.008837182191200554,
+      "entropy": 0.08954936638474464,
+      "epoch": 8e-05,
+      "grad_norm": 1.4120911359786987,
+      "kl": 0.0029542466588736715,
+      "learning_rate": 2e-05,
+      "loss": -0.0404,
+      "step": 8,
+      "step_time": 28.796026483003516
+    },
+    {
+      "clip_ratio/high_max": 0.004531709826551378,
+      "clip_ratio/high_mean": 0.00261697854148224,
+      "clip_ratio/low_mean": 0.0036588753282558173,
+      "clip_ratio/low_min": 0.0014044943964108825,
+      "clip_ratio/region_mean": 0.006275853869738057,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9217.0,
+      "completions/max_terminated_length": 9217.0,
+      "completions/mean_length": 7774.4375,
+      "completions/mean_terminated_length": 7774.4375,
+      "completions/min_length": 860.0,
+      "completions/min_terminated_length": 860.0,
+      "entropy": 0.09250054159201682,
+      "epoch": 9e-05,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.8319895267486572,
+      "kl": 0.003183710745361168,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": -0.0399,
+      "num_tokens": 1396422.0,
+      "reward": -0.35600194334983826,
+      "reward_std": 0.46349450945854187,
+      "rewards/rollout_reward_func/mean": -0.35600194334983826,
+      "rewards/rollout_reward_func/std": 0.4859367907047272,
+      "sampling/importance_sampling_ratio/max": 2.4923911094665527,
+      "sampling/importance_sampling_ratio/mean": 0.9995814561843872,
+      "sampling/importance_sampling_ratio/min": 0.34728506207466125,
+      "sampling/sampling_logp_difference/max": 1.0576093196868896,
+      "sampling/sampling_logp_difference/mean": 0.011536622419953346,
+      "step": 9,
+      "step_time": 91.41392734301189
+    },
+    {
+      "clip_ratio/high_max": 0.005798806145321578,
+      "clip_ratio/high_mean": 0.0032545167196076363,
+      "clip_ratio/low_mean": 0.005793917312985286,
+      "clip_ratio/low_min": 0.0007022471982054412,
+      "clip_ratio/region_mean": 0.009048434003489092,
+      "entropy": 0.08986990223638713,
+      "epoch": 0.0001,
+      "grad_norm": 1.1754401922225952,
+      "kl": 0.006458986033976544,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": -0.0398,
+      "step": 10,
+      "step_time": 27.723220527001104
+    },
+    {
+      "clip_ratio/high_max": 0.002881916589103639,
+      "clip_ratio/high_mean": 0.0014409582945518196,
+      "clip_ratio/low_mean": 0.0003511235991027206,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0017920818936545402,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9155.0,
+      "completions/max_terminated_length": 9155.0,
+      "completions/mean_length": 7173.3125,
+      "completions/mean_terminated_length": 7173.3125,
+      "completions/min_length": 887.0,
+      "completions/min_terminated_length": 887.0,
+      "entropy": 0.06778168666642159,
+      "epoch": 0.00011,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.0174713134765625,
+      "kl": 0.004350664900812262,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": -0.2438,
+      "num_tokens": 1653204.0,
+      "reward": -0.4879741668701172,
+      "reward_std": 0.49708276987075806,
+      "rewards/rollout_reward_func/mean": -0.4879741668701172,
+      "rewards/rollout_reward_func/std": 0.5053675770759583,
+      "sampling/importance_sampling_ratio/max": 1.3829487562179565,
+      "sampling/importance_sampling_ratio/mean": 1.0001708269119263,
+      "sampling/importance_sampling_ratio/min": 0.6129343509674072,
+      "sampling/sampling_logp_difference/max": 0.48949742317199707,
+      "sampling/sampling_logp_difference/mean": 0.006062386557459831,
+      "step": 11,
+      "step_time": 88.94538035600272
+    },
+    {
+      "clip_ratio/high_max": 0.0021720757358707488,
+      "clip_ratio/high_mean": 0.0010860378679353744,
+      "clip_ratio/low_mean": 0.005371870560338721,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.006457908428274095,
+      "entropy": 0.0648728254600428,
+      "epoch": 0.00012,
+      "grad_norm": 0.9108887910842896,
+      "kl": 0.010094417615619022,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": -0.2448,
+      "step": 12,
+      "step_time": 27.213586286001373
+    },
+    {
+      "clip_ratio/high_max": 0.009980108763556927,
+      "clip_ratio/high_mean": 0.004990054381778464,
+      "clip_ratio/low_mean": 0.00396484971861355,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.008954904100392014,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9369.0,
+      "completions/max_terminated_length": 9369.0,
+      "completions/mean_length": 8193.15625,
+      "completions/mean_terminated_length": 8193.15625,
+      "completions/min_length": 960.0,
+      "completions/min_terminated_length": 960.0,
+      "entropy": 0.10193604091182351,
+      "epoch": 0.00013,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.485856294631958,
+      "kl": 0.030039603414479643,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": -0.1495,
+      "num_tokens": 1943132.0,
+      "reward": -0.5804459452629089,
+      "reward_std": 0.27082452178001404,
+      "rewards/rollout_reward_func/mean": -0.5804459452629089,
+      "rewards/rollout_reward_func/std": 0.2952437400817871,
+      "sampling/importance_sampling_ratio/max": 1.951629877090454,
+      "sampling/importance_sampling_ratio/mean": 0.9995810985565186,
+      "sampling/importance_sampling_ratio/min": 0.08306944370269775,
+      "sampling/sampling_logp_difference/max": 2.4880783557891846,
+      "sampling/sampling_logp_difference/mean": 0.016808483749628067,
+      "step": 13,
+      "step_time": 94.51923282000644
+    },
+    {
+      "clip_ratio/high_max": 0.015696270624175668,
+      "clip_ratio/high_mean": 0.007848135312087834,
+      "clip_ratio/low_mean": 0.00827725778799504,
+      "clip_ratio/low_min": 0.0021828330936841667,
+      "clip_ratio/region_mean": 0.016125393158290535,
+      "entropy": 0.10211672727018595,
+      "epoch": 0.00014,
+      "grad_norm": 1.529998779296875,
+      "kl": 0.045463139831554145,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": -0.1549,
+      "step": 14,
+      "step_time": 27.938481579996733
+    },
+    {
+      "clip_ratio/high_max": 0.002873942954465747,
+      "clip_ratio/high_mean": 0.0017880951054394245,
+      "clip_ratio/low_mean": 0.0017879052029456943,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.003576000308385119,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9252.0,
+      "completions/max_terminated_length": 9252.0,
+      "completions/mean_length": 7854.34375,
+      "completions/mean_terminated_length": 7854.34375,
+      "completions/min_length": 1084.0,
+      "completions/min_terminated_length": 1084.0,
+      "entropy": 0.064073825487867,
+      "epoch": 0.00015,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 2.1420814990997314,
+      "kl": 0.013782255366095342,
+      "learning_rate": 4e-05,
+      "loss": 0.0498,
+      "num_tokens": 2221867.0,
+      "reward": -0.3640046715736389,
+      "reward_std": 0.5706602334976196,
+      "rewards/rollout_reward_func/mean": -0.3640046715736389,
+      "rewards/rollout_reward_func/std": 0.5570085644721985,
+      "sampling/importance_sampling_ratio/max": 2.0640242099761963,
+      "sampling/importance_sampling_ratio/mean": 1.001070499420166,
+      "sampling/importance_sampling_ratio/min": 0.4299520254135132,
+      "sampling/sampling_logp_difference/max": 0.8440816402435303,
+      "sampling/sampling_logp_difference/mean": 0.008871052414178848,
+      "step": 15,
+      "step_time": 95.68331051399582
+    },
+    {
+      "clip_ratio/high_max": 0.009986817487515509,
+      "clip_ratio/high_mean": 0.005344532371964306,
+      "clip_ratio/low_mean": 0.005893140041735023,
+      "clip_ratio/low_min": 0.0014044943964108825,
+      "clip_ratio/region_mean": 0.01123767253011465,
+      "entropy": 0.0662008763756603,
+      "epoch": 0.00016,
+      "grad_norm": 1.10405695438385,
+      "kl": 0.022447960553108715,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.045,
+      "step": 16,
+      "step_time": 27.621052484992106
+    },
+    {
+      "clip_ratio/high_max": 0.006080338731408119,
+      "clip_ratio/high_mean": 0.0030401693657040596,
+      "clip_ratio/low_mean": 0.003163026151014492,
+      "clip_ratio/low_min": 0.0007183908019214869,
+      "clip_ratio/region_mean": 0.0062031955167185515,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9357.0,
+      "completions/max_terminated_length": 9357.0,
+      "completions/mean_length": 7752.09375,
+      "completions/mean_terminated_length": 7752.09375,
+      "completions/min_length": 2165.0,
+      "completions/min_terminated_length": 2165.0,
+      "entropy": 0.07411375921219587,
+      "epoch": 0.00017,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.4593960046768188,
+      "kl": 0.05461276904679835,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.0396,
+      "num_tokens": 2497538.0,
+      "reward": -0.2799840569496155,
+      "reward_std": 0.5249596834182739,
+      "rewards/rollout_reward_func/mean": -0.2799840569496155,
+      "rewards/rollout_reward_func/std": 0.5806505680084229,
+      "sampling/importance_sampling_ratio/max": 1.9463616609573364,
+      "sampling/importance_sampling_ratio/mean": 0.9976865649223328,
+      "sampling/importance_sampling_ratio/min": 0.06184462085366249,
+      "sampling/sampling_logp_difference/max": 2.783130168914795,
+      "sampling/sampling_logp_difference/mean": 0.014597892761230469,
+      "step": 17,
+      "step_time": 93.46702274899872
+    },
+    {
+      "clip_ratio/high_max": 0.01004306401591748,
+      "clip_ratio/high_mean": 0.00502153200795874,
+      "clip_ratio/low_mean": 0.009563877363689244,
+      "clip_ratio/low_min": 0.001420638000126928,
+      "clip_ratio/region_mean": 0.014585409400751814,
+      "entropy": 0.07483723293989897,
+      "epoch": 0.00018,
+      "grad_norm": 1.2538055181503296,
+      "kl": 0.17985437362222,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.036,
+      "step": 18,
+      "step_time": 27.61892309099494
+    },
+    {
+      "clip_ratio/high_max": 0.007797640864737332,
+      "clip_ratio/high_mean": 0.004960351681802422,
+      "clip_ratio/low_mean": 0.003771967807551846,
+      "clip_ratio/low_min": 0.0013297871919348836,
+      "clip_ratio/region_mean": 0.008732319489354268,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9323.0,
+      "completions/max_terminated_length": 9323.0,
+      "completions/mean_length": 8057.5625,
+      "completions/mean_terminated_length": 8057.5625,
+      "completions/min_length": 4131.0,
+      "completions/min_terminated_length": 4131.0,
+      "entropy": 0.09318487590644509,
+      "epoch": 0.00019,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.8037179708480835,
+      "kl": 0.08203080226667225,
+      "learning_rate": 5.142857142857143e-05,
+      "loss": -0.1161,
+      "num_tokens": 2782910.0,
+      "reward": -0.5081856846809387,
+      "reward_std": 0.34735041856765747,
+      "rewards/rollout_reward_func/mean": -0.5081856846809387,
+      "rewards/rollout_reward_func/std": 0.34861209988594055,
+      "sampling/importance_sampling_ratio/max": 2.6022117137908936,
+      "sampling/importance_sampling_ratio/mean": 1.0005253553390503,
+      "sampling/importance_sampling_ratio/min": 0.39166536927223206,
+      "sampling/sampling_logp_difference/max": 0.9563617706298828,
+      "sampling/sampling_logp_difference/mean": 0.016585635021328926,
+      "step": 19,
+      "step_time": 96.19064523599809
+    },
+    {
+      "clip_ratio/high_max": 0.009210843476466835,
+      "clip_ratio/high_mean": 0.0056830079411156476,
+      "clip_ratio/low_mean": 0.01183583348756656,
+      "clip_ratio/low_min": 0.005575570685323328,
+      "clip_ratio/region_mean": 0.01751884145778604,
+      "entropy": 0.08642621210310608,
+      "epoch": 0.0002,
+      "grad_norm": 1.6954180002212524,
+      "kl": 0.1374246042687446,
+      "learning_rate": 5.428571428571428e-05,
+      "loss": -0.1201,
+      "step": 20,
+      "step_time": 27.602408401988214
+    },
+    {
+      "clip_ratio/high_max": 0.005014301335904747,
+      "clip_ratio/high_mean": 0.002883656619815156,
+      "clip_ratio/low_mean": 0.0028673816996160895,
+      "clip_ratio/low_min": 0.0006944444612599909,
+      "clip_ratio/region_mean": 0.005751038319431245,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9403.0,
+      "completions/max_terminated_length": 9403.0,
+      "completions/mean_length": 7930.5625,
+      "completions/mean_terminated_length": 7930.5625,
+      "completions/min_length": 634.0,
+      "completions/min_terminated_length": 634.0,
+      "entropy": 0.07444759807549417,
+      "epoch": 0.00021,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.1838347911834717,
+      "kl": 0.09432762704091147,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": -0.159,
+      "num_tokens": 3064148.0,
+      "reward": -0.3027535676956177,
+      "reward_std": 0.49178069829940796,
+      "rewards/rollout_reward_func/mean": -0.3027535676956177,
+      "rewards/rollout_reward_func/std": 0.5297390818595886,
+      "sampling/importance_sampling_ratio/max": 2.3851823806762695,
+      "sampling/importance_sampling_ratio/mean": 1.0000920295715332,
+      "sampling/importance_sampling_ratio/min": 0.3727108836174011,
+      "sampling/sampling_logp_difference/max": 0.9869523048400879,
+      "sampling/sampling_logp_difference/mean": 0.011310078203678131,
+      "step": 21,
+      "step_time": 91.98894712300535
+    },
+    {
+      "clip_ratio/high_max": 0.011702421237714589,
+      "clip_ratio/high_mean": 0.0065865046926774085,
+      "clip_ratio/low_mean": 0.011994066560873762,
+      "clip_ratio/low_min": 0.0024457787512801588,
+      "clip_ratio/region_mean": 0.01858057125355117,
+      "entropy": 0.07004613918252289,
+      "epoch": 0.00022,
+      "grad_norm": 1.0932140350341797,
+      "kl": 0.06876913830637932,
+      "learning_rate": 6e-05,
+      "loss": -0.168,
+      "step": 22,
+      "step_time": 27.827071762007108
+    },
+    {
+      "clip_ratio/high_max": 0.007137945271097124,
+      "clip_ratio/high_mean": 0.003936619672458619,
+      "clip_ratio/low_mean": 0.0028658679220825434,
+      "clip_ratio/low_min": 0.0007267441833391786,
+      "clip_ratio/region_mean": 0.006802487594541162,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9274.0,
+      "completions/max_terminated_length": 9274.0,
+      "completions/mean_length": 8095.0625,
+      "completions/mean_terminated_length": 8095.0625,
+      "completions/min_length": 1533.0,
+      "completions/min_terminated_length": 1533.0,
+      "entropy": 0.0666239457204938,
+      "epoch": 0.00023,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.2763183116912842,
+      "kl": 0.05845395056530833,
+      "learning_rate": 6.285714285714286e-05,
+      "loss": 0.3004,
+      "num_tokens": 3350379.0,
+      "reward": -0.3713611960411072,
+      "reward_std": 0.44723719358444214,
+      "rewards/rollout_reward_func/mean": -0.3713611960411072,
+      "rewards/rollout_reward_func/std": 0.5041891932487488,
+      "sampling/importance_sampling_ratio/max": 3.0,
+      "sampling/importance_sampling_ratio/mean": 1.0009047985076904,
+      "sampling/importance_sampling_ratio/min": 0.4158317744731903,
+      "sampling/sampling_logp_difference/max": 1.114884614944458,
+      "sampling/sampling_logp_difference/mean": 0.01109105534851551,
+      "step": 23,
+      "step_time": 93.21978440000748
+    },
+    {
+      "clip_ratio/high_max": 0.014727697416674346,
+      "clip_ratio/high_mean": 0.00882161219487898,
+      "clip_ratio/low_mean": 0.01083813124569133,
+      "clip_ratio/low_min": 0.003661759605165571,
+      "clip_ratio/region_mean": 0.01965974346967414,
+      "entropy": 0.06382836727425456,
+      "epoch": 0.00024,
+      "grad_norm": 1.133302092552185,
+      "kl": 0.07930020266212523,
+      "learning_rate": 6.571428571428571e-05,
+      "loss": 0.2957,
+      "step": 24,
+      "step_time": 28.035423179004283
+    },
+    {
+      "clip_ratio/high_max": 0.004302181303501129,
+      "clip_ratio/high_mean": 0.0021510906517505646,
+      "clip_ratio/low_mean": 0.0007800387684255838,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0029311294783838093,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 8580.0,
+      "completions/max_terminated_length": 8580.0,
+      "completions/mean_length": 6758.75,
+      "completions/mean_terminated_length": 6758.75,
+      "completions/min_length": 272.0,
+      "completions/min_terminated_length": 272.0,
+      "entropy": 0.04931171424686909,
+      "epoch": 0.00025,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.291664719581604,
+      "kl": 0.07758830557577312,
+      "learning_rate": 6.857142857142858e-05,
+      "loss": -0.2042,
+      "num_tokens": 3593214.0,
+      "reward": -0.5959107875823975,
+      "reward_std": 0.35876262187957764,
+      "rewards/rollout_reward_func/mean": -0.5959107875823975,
+      "rewards/rollout_reward_func/std": 0.39302268624305725,
+      "sampling/importance_sampling_ratio/max": 2.010491371154785,
+      "sampling/importance_sampling_ratio/mean": 1.0006413459777832,
+      "sampling/importance_sampling_ratio/min": 0.42770469188690186,
+      "sampling/sampling_logp_difference/max": 0.8493223190307617,
+      "sampling/sampling_logp_difference/mean": 0.008722992613911629,
+      "step": 25,
+      "step_time": 86.72631602899492
+    },
+    {
+      "clip_ratio/high_max": 0.010786514729261398,
+      "clip_ratio/high_mean": 0.005393257364630699,
+      "clip_ratio/low_mean": 0.011662953183986247,
+      "clip_ratio/low_min": 0.0019696029485203326,
+      "clip_ratio/region_mean": 0.017056210490409285,
+      "entropy": 0.05728732608258724,
+      "epoch": 0.00026,
+      "grad_norm": 2.188659906387329,
+      "kl": 0.08259193820413202,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": -0.2036,
+      "step": 26,
+      "step_time": 26.119606291009404
+    },
+    {
+      "clip_ratio/high_max": 0.0061968869413249195,
+      "clip_ratio/high_mean": 0.0030984434706624597,
+      "clip_ratio/low_mean": 0.0010736910917330533,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0041721345914993435,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 8940.0,
+      "completions/max_terminated_length": 8940.0,
+      "completions/mean_length": 7771.3125,
+      "completions/mean_terminated_length": 7771.3125,
+      "completions/min_length": 1515.0,
+      "completions/min_terminated_length": 1515.0,
+      "entropy": 0.0601597810164094,
+      "epoch": 0.00027,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.141928791999817,
+      "kl": 0.05808806075947359,
+      "learning_rate": 7.428571428571429e-05,
+      "loss": 0.1398,
+      "num_tokens": 3868467.0,
+      "reward": -0.36368632316589355,
+      "reward_std": 0.5357357263565063,
+      "rewards/rollout_reward_func/mean": -0.36368632316589355,
+      "rewards/rollout_reward_func/std": 0.5756723880767822,
+      "sampling/importance_sampling_ratio/max": 1.8575561046600342,
+      "sampling/importance_sampling_ratio/mean": 1.0000970363616943,
+      "sampling/importance_sampling_ratio/min": 0.3134206235408783,
+      "sampling/sampling_logp_difference/max": 1.1602091789245605,
+      "sampling/sampling_logp_difference/mean": 0.009103494696319103,
+      "step": 27,
+      "step_time": 92.96613138498651
+    },
+    {
+      "clip_ratio/high_max": 0.0164202400483191,
+      "clip_ratio/high_mean": 0.00821012002415955,
+      "clip_ratio/low_mean": 0.013974322064314038,
+      "clip_ratio/low_min": 0.0007183908019214869,
+      "clip_ratio/region_mean": 0.022184442059369758,
+      "entropy": 0.0649251660797745,
+      "epoch": 0.00028,
+      "grad_norm": 0.8147655725479126,
+      "kl": 0.10712331312242895,
+      "learning_rate": 7.714285714285715e-05,
+      "loss": 0.1369,
+      "step": 28,
+      "step_time": 27.32312696599911
+    },
+    {
+      "clip_ratio/high_max": 0.002155932132154703,
+      "clip_ratio/high_mean": 0.0010779660660773516,
+      "clip_ratio/low_mean": 0.0007102272938936949,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0017881933599710464,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9165.0,
+      "completions/max_terminated_length": 9165.0,
+      "completions/mean_length": 7336.8125,
+      "completions/mean_terminated_length": 7336.8125,
+      "completions/min_length": 1667.0,
+      "completions/min_terminated_length": 1667.0,
+      "entropy": 0.06351292878389359,
+      "epoch": 0.00029,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.7936819195747375,
+      "kl": 0.04702634894056246,
+      "learning_rate": 8e-05,
+      "loss": -0.2535,
+      "num_tokens": 4129721.0,
+      "reward": -0.4376564919948578,
+      "reward_std": 0.39146924018859863,
+      "rewards/rollout_reward_func/mean": -0.4376564919948578,
+      "rewards/rollout_reward_func/std": 0.4162859320640564,
+      "sampling/importance_sampling_ratio/max": 1.9884843826293945,
+      "sampling/importance_sampling_ratio/mean": 1.0006194114685059,
+      "sampling/importance_sampling_ratio/min": 0.5153306126594543,
+      "sampling/sampling_logp_difference/max": 0.6873726844787598,
+      "sampling/sampling_logp_difference/mean": 0.00740803312510252,
+      "step": 29,
+      "step_time": 91.82200075300352
+    },
+    {
+      "clip_ratio/high_max": 0.005039906303863972,
+      "clip_ratio/high_mean": 0.00323018038761802,
+      "clip_ratio/low_mean": 0.010294274339685217,
+      "clip_ratio/low_min": 0.004024621332064271,
+      "clip_ratio/region_mean": 0.013524454698199406,
+      "entropy": 0.06765187811106443,
+      "epoch": 0.0003,
+      "grad_norm": 0.7302473783493042,
+      "kl": 0.050249568943399936,
+      "learning_rate": 8.285714285714287e-05,
+      "loss": -0.2607,
+      "step": 30,
+      "step_time": 28.11271668801055
+    },
+    {
+      "clip_ratio/high_max": 0.00493967067450285,
+      "clip_ratio/high_mean": 0.002469835337251425,
+      "clip_ratio/low_mean": 0.0007143090479075909,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0031841443851590157,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9146.0,
+      "completions/max_terminated_length": 9146.0,
+      "completions/mean_length": 7793.5625,
+      "completions/mean_terminated_length": 7793.5625,
+      "completions/min_length": 263.0,
+      "completions/min_terminated_length": 263.0,
+      "entropy": 0.08614894514903426,
+      "epoch": 0.00031,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.0549495220184326,
+      "kl": 0.06513052270747721,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": -0.0626,
+      "num_tokens": 4405861.0,
+      "reward": -0.37940478324890137,
+      "reward_std": 0.45117825269699097,
+      "rewards/rollout_reward_func/mean": -0.37940478324890137,
+      "rewards/rollout_reward_func/std": 0.48283621668815613,
+      "sampling/importance_sampling_ratio/max": 1.4535597562789917,
+      "sampling/importance_sampling_ratio/mean": 0.9976645112037659,
+      "sampling/importance_sampling_ratio/min": 0.5075654983520508,
+      "sampling/sampling_logp_difference/max": 0.6781294345855713,
+      "sampling/sampling_logp_difference/mean": 0.009683198295533657,
+      "step": 31,
+      "step_time": 92.44875282800785
+    },
+    {
+      "clip_ratio/high_max": 0.011355375638231635,
+      "clip_ratio/high_mean": 0.006735137314535677,
+      "clip_ratio/low_mean": 0.012324601178988814,
+      "clip_ratio/low_min": 0.0007022471982054412,
+      "clip_ratio/region_mean": 0.019059738871874288,
+      "entropy": 0.08400337002240121,
+      "epoch": 0.00032,
+      "grad_norm": 0.6007691621780396,
+      "kl": 0.08906090061645955,
+      "learning_rate": 8.857142857142857e-05,
+      "loss": -0.0698,
+      "step": 32,
+      "step_time": 28.252189476006606
+    },
+    {
+      "clip_ratio/high_max": 0.007888395455665886,
+      "clip_ratio/high_mean": 0.003944197727832943,
+      "clip_ratio/low_mean": 0.004384168569231406,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.00832836632616818,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9394.0,
+      "completions/max_terminated_length": 9394.0,
+      "completions/mean_length": 7661.125,
+      "completions/mean_terminated_length": 7661.125,
+      "completions/min_length": 1012.0,
+      "completions/min_terminated_length": 1012.0,
+      "entropy": 0.08313656039535999,
+      "epoch": 0.00033,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.035186767578125,
+      "kl": 0.09069790283683687,
+      "learning_rate": 9.142857142857143e-05,
+      "loss": 0.0166,
+      "num_tokens": 4677397.0,
+      "reward": -0.3240296244621277,
+      "reward_std": 0.4347441494464874,
+      "rewards/rollout_reward_func/mean": -0.3240296244621277,
+      "rewards/rollout_reward_func/std": 0.4854819178581238,
+      "sampling/importance_sampling_ratio/max": 1.7231584787368774,
+      "sampling/importance_sampling_ratio/mean": 0.9987027645111084,
+      "sampling/importance_sampling_ratio/min": 0.06808681786060333,
+      "sampling/sampling_logp_difference/max": 2.686971664428711,
+      "sampling/sampling_logp_difference/mean": 0.010714043863117695,
+      "step": 33,
+      "step_time": 93.2897496730111
+    },
+    {
+      "clip_ratio/high_max": 0.020736676699016243,
+      "clip_ratio/high_mean": 0.01142766050179489,
+      "clip_ratio/low_mean": 0.010279498674208298,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.02170715932152234,
+      "entropy": 0.07817789679393172,
+      "epoch": 0.00034,
+      "grad_norm": 0.6795229911804199,
+      "kl": 0.110447074577678,
+      "learning_rate": 9.428571428571429e-05,
+      "loss": 0.0053,
+      "step": 34,
+      "step_time": 28.798946735994832
+    },
+    {
+      "clip_ratio/high_max": 0.004271204466931522,
+      "clip_ratio/high_mean": 0.002135602233465761,
+      "clip_ratio/low_mean": 0.0010695144010242075,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0032051166344899684,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9245.0,
+      "completions/max_terminated_length": 9245.0,
+      "completions/mean_length": 8244.75,
+      "completions/mean_terminated_length": 8244.75,
+      "completions/min_length": 291.0,
+      "completions/min_terminated_length": 291.0,
+      "entropy": 0.06729014380834997,
+      "epoch": 0.00035,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.4786404371261597,
+      "kl": 0.10362116701435298,
+      "learning_rate": 9.714285714285715e-05,
+      "loss": -0.2709,
+      "num_tokens": 4968197.0,
+      "reward": -0.3916197419166565,
+      "reward_std": 0.35996532440185547,
+      "rewards/rollout_reward_func/mean": -0.3916197419166565,
+      "rewards/rollout_reward_func/std": 0.3666594326496124,
+      "sampling/importance_sampling_ratio/max": 1.5868641138076782,
+      "sampling/importance_sampling_ratio/mean": 1.0002307891845703,
+      "sampling/importance_sampling_ratio/min": 0.6188454031944275,
+      "sampling/sampling_logp_difference/max": 0.4798997640609741,
+      "sampling/sampling_logp_difference/mean": 0.007114737294614315,
+      "step": 35,
+      "step_time": 92.1033384639959
+    },
+    {
+      "clip_ratio/high_max": 0.017084260645788163,
+      "clip_ratio/high_mean": 0.009252357733203098,
+      "clip_ratio/low_mean": 0.012271191226318479,
+      "clip_ratio/low_min": 0.0014044943964108825,
+      "clip_ratio/region_mean": 0.021523548872210085,
+      "entropy": 0.06766591034829617,
+      "epoch": 0.00036,
+      "grad_norm": 0.7554082870483398,
+      "kl": 0.11198568157851696,
+      "learning_rate": 0.0001,
+      "loss": -0.2783,
+      "step": 36,
+      "step_time": 28.797730328002217
+    },
+    {
+      "clip_ratio/high_max": 0.0014046717551536858,
+      "clip_ratio/high_mean": 0.0007023358775768429,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0007023358775768429,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9282.0,
+      "completions/max_terminated_length": 9282.0,
+      "completions/mean_length": 7701.8125,
+      "completions/mean_terminated_length": 7701.8125,
+      "completions/min_length": 1350.0,
+      "completions/min_terminated_length": 1350.0,
+      "entropy": 0.06278971070423722,
+      "epoch": 0.00037,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.0583816766738892,
+      "kl": 0.12011543358676136,
+      "learning_rate": 9.918053502751771e-05,
+      "loss": -0.0297,
+      "num_tokens": 5241387.0,
+      "reward": -0.304426908493042,
+      "reward_std": 0.4464525580406189,
+      "rewards/rollout_reward_func/mean": -0.304426908493042,
+      "rewards/rollout_reward_func/std": 0.4538337290287018,
+      "sampling/importance_sampling_ratio/max": 1.6482372283935547,
+      "sampling/importance_sampling_ratio/mean": 0.9995368719100952,
+      "sampling/importance_sampling_ratio/min": 0.5574944615364075,
+      "sampling/sampling_logp_difference/max": 0.5843026638031006,
+      "sampling/sampling_logp_difference/mean": 0.007077127229422331,
+      "step": 37,
+      "step_time": 91.12368497201533
+    },
+    {
+      "clip_ratio/high_max": 0.010138559562619776,
+      "clip_ratio/high_mean": 0.006660482700681314,
+      "clip_ratio/low_mean": 0.01362931149196811,
+      "clip_ratio/low_min": 0.003907083009835333,
+      "clip_ratio/region_mean": 0.020289794192649424,
+      "entropy": 0.05679044884163886,
+      "epoch": 0.00038,
+      "grad_norm": 0.618403971195221,
+      "kl": 0.118131990195252,
+      "learning_rate": 9.675795466159755e-05,
+      "loss": -0.0362,
+      "step": 38,
+      "step_time": 28.32057495099798
+    },
+    {
+      "clip_ratio/high_max": 0.005428613570984453,
+      "clip_ratio/high_mean": 0.0027143067854922265,
+      "clip_ratio/low_mean": 0.0010573608451522887,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.003771667630644515,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9203.0,
+      "completions/max_terminated_length": 9203.0,
+      "completions/mean_length": 7459.09375,
+      "completions/mean_terminated_length": 7459.09375,
+      "completions/min_length": 647.0,
+      "completions/min_terminated_length": 647.0,
+      "entropy": 0.05742924939841032,
+      "epoch": 0.00039,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.1807667016983032,
+      "kl": 0.11824507999699563,
+      "learning_rate": 9.283813728906054e-05,
+      "loss": -0.0685,
+      "num_tokens": 5507267.0,
+      "reward": -0.1265014261007309,
+      "reward_std": 0.6104403734207153,
+      "rewards/rollout_reward_func/mean": -0.1265014261007309,
+      "rewards/rollout_reward_func/std": 0.6330578327178955,
+      "sampling/importance_sampling_ratio/max": 1.6747703552246094,
+      "sampling/importance_sampling_ratio/mean": 0.9992620944976807,
+      "sampling/importance_sampling_ratio/min": 0.39614981412887573,
+      "sampling/sampling_logp_difference/max": 0.9259628653526306,
+      "sampling/sampling_logp_difference/mean": 0.006985975429415703,
+      "step": 39,
+      "step_time": 91.56473480099157
+    },
+    {
+      "clip_ratio/high_max": 0.017474214837420732,
+      "clip_ratio/high_mean": 0.008737107418710366,
+      "clip_ratio/low_mean": 0.01038203525240533,
+      "clip_ratio/low_min": 0.0021973324473947287,
+      "clip_ratio/region_mean": 0.01911914284573868,
+      "entropy": 0.05913685623090714,
+      "epoch": 0.0004,
+      "grad_norm": 11257.0869140625,
+      "kl": 15902.953878552886,
+      "learning_rate": 8.759239773845718e-05,
+      "loss": 59.5637,
+      "step": 40,
+      "step_time": 28.143666776995815
+    },
+    {
+      "clip_ratio/high_max": 0.0014367816038429737,
+      "clip_ratio/high_mean": 0.00109489681199193,
+      "clip_ratio/low_mean": 0.002375774463871494,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.003470671275863424,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9299.0,
+      "completions/max_terminated_length": 9299.0,
+      "completions/mean_length": 7399.59375,
+      "completions/mean_terminated_length": 7399.59375,
+      "completions/min_length": 311.0,
+      "completions/min_terminated_length": 311.0,
+      "entropy": 0.0707793568726629,
+      "epoch": 0.00041,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.779978334903717,
+      "kl": 0.1064646802842617,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": -0.006,
+      "num_tokens": 5770826.0,
+      "reward": -0.10253402590751648,
+      "reward_std": 0.587022602558136,
+      "rewards/rollout_reward_func/mean": -0.10253402590751648,
+      "rewards/rollout_reward_func/std": 0.58503657579422,
+      "sampling/importance_sampling_ratio/max": 1.7891288995742798,
+      "sampling/importance_sampling_ratio/mean": 1.000353455543518,
+      "sampling/importance_sampling_ratio/min": 0.33879733085632324,
+      "sampling/sampling_logp_difference/max": 1.082353115081787,
+      "sampling/sampling_logp_difference/mean": 0.008459154516458511,
+      "step": 41,
+      "step_time": 89.9564655849972
+    },
+    {
+      "clip_ratio/high_max": 0.013633344613481313,
+      "clip_ratio/high_mean": 0.0071758677077014,
+      "clip_ratio/low_mean": 0.013401742442511022,
+      "clip_ratio/low_min": 0.0028420357266440988,
+      "clip_ratio/region_mean": 0.020577610324835405,
+      "entropy": 0.07488202536478639,
+      "epoch": 0.00042,
+      "grad_norm": 0.6732333302497864,
+      "kl": 0.11995455622673035,
+      "learning_rate": 7.408813728906053e-05,
+      "loss": -0.0139,
+      "step": 42,
+      "step_time": 27.865521725998406
+    },
+    {
+      "clip_ratio/high_max": 0.001420638000126928,
+      "clip_ratio/high_mean": 0.000710319000063464,
+      "clip_ratio/low_mean": 0.002360046637477353,
+      "clip_ratio/low_min": 0.0007022471982054412,
+      "clip_ratio/region_mean": 0.0030703656375408173,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 8920.0,
+      "completions/max_terminated_length": 8920.0,
+      "completions/mean_length": 7293.5,
+      "completions/mean_terminated_length": 7293.5,
+      "completions/min_length": 876.0,
+      "completions/min_terminated_length": 876.0,
+      "entropy": 0.06767579750157893,
+      "epoch": 0.00043,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 0.9846507906913757,
+      "kl": 0.11756309028714895,
+      "learning_rate": 6.6419817372537e-05,
+      "loss": -0.105,
+      "num_tokens": 6030904.0,
+      "reward": -0.17516297101974487,
+      "reward_std": 0.6336148977279663,
+      "rewards/rollout_reward_func/mean": -0.17516297101974487,
+      "rewards/rollout_reward_func/std": 0.6060534715652466,
+      "sampling/importance_sampling_ratio/max": 1.5995110273361206,
+      "sampling/importance_sampling_ratio/mean": 1.0016493797302246,
+      "sampling/importance_sampling_ratio/min": 0.6211615800857544,
+      "sampling/sampling_logp_difference/max": 0.4761640429496765,
+      "sampling/sampling_logp_difference/mean": 0.00795302540063858,
+      "step": 43,
+      "step_time": 88.68364224100515
+    },
+    {
+      "clip_ratio/high_max": 0.011437574110459536,
+      "clip_ratio/high_mean": 0.006488003156846389,
+      "clip_ratio/low_mean": 0.011332564434269443,
+      "clip_ratio/low_min": 0.002825683041010052,
+      "clip_ratio/region_mean": 0.01782056744559668,
+      "entropy": 0.06705769780091941,
+      "epoch": 0.00044,
+      "grad_norm": 0.5778977274894714,
+      "kl": 0.13579032279085368,
+      "learning_rate": 5.858018262746301e-05,
+      "loss": -0.1148,
+      "step": 44,
+      "step_time": 26.98718312199344
+    },
+    {
+      "clip_ratio/high_max": 0.0022492663701996207,
+      "clip_ratio/high_mean": 0.0011246331850998104,
+      "clip_ratio/low_mean": 0.0028377448907122016,
+      "clip_ratio/low_min": 0.0006944444612599909,
+      "clip_ratio/region_mean": 0.003962378075812012,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9242.0,
+      "completions/max_terminated_length": 9242.0,
+      "completions/mean_length": 8146.5,
+      "completions/mean_terminated_length": 8146.5,
+      "completions/min_length": 1426.0,
+      "completions/min_terminated_length": 1426.0,
+      "entropy": 0.08103846106678247,
+      "epoch": 0.00045,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.409741759300232,
+      "kl": 0.14359550992958248,
+      "learning_rate": 5.0911862710939485e-05,
+      "loss": -0.034,
+      "num_tokens": 6318667.0,
+      "reward": -0.12613427639007568,
+      "reward_std": 0.5343547463417053,
+      "rewards/rollout_reward_func/mean": -0.12613427639007568,
+      "rewards/rollout_reward_func/std": 0.5416647791862488,
+      "sampling/importance_sampling_ratio/max": 3.0,
+      "sampling/importance_sampling_ratio/mean": 1.000476598739624,
+      "sampling/importance_sampling_ratio/min": 0.09892720729112625,
+      "sampling/sampling_logp_difference/max": 2.313370943069458,
+      "sampling/sampling_logp_difference/mean": 0.012712076306343079,
+      "step": 45,
+      "step_time": 93.79370664700036
+    },
+    {
+      "clip_ratio/high_max": 0.011134565633255988,
+      "clip_ratio/high_mean": 0.006269618694204837,
+      "clip_ratio/low_mean": 0.009181194531265646,
+      "clip_ratio/low_min": 0.0028904599603265524,
+      "clip_ratio/region_mean": 0.015450813429197297,
+      "entropy": 0.08389568043639883,
+      "epoch": 0.00046,
+      "grad_norm": 1.2059129476547241,
+      "kl": 0.24020475335419178,
+      "learning_rate": 4.375000000000001e-05,
+      "loss": -0.0432,
+      "step": 46,
+      "step_time": 28.168583785994997
+    },
+    {
+      "clip_ratio/high_max": 0.005004245089367032,
+      "clip_ratio/high_mean": 0.002502122544683516,
+      "clip_ratio/low_mean": 0.002046188630629331,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0045483111462090164,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9056.0,
+      "completions/max_terminated_length": 9056.0,
+      "completions/mean_length": 7947.84375,
+      "completions/mean_terminated_length": 7947.84375,
+      "completions/min_length": 4376.0,
+      "completions/min_terminated_length": 4376.0,
+      "entropy": 0.07940921117551625,
+      "epoch": 0.00047,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.2504568099975586,
+      "kl": 0.16516069741919637,
+      "learning_rate": 3.740760226154283e-05,
+      "loss": -0.0286,
+      "num_tokens": 6600118.0,
+      "reward": -0.12103764712810516,
+      "reward_std": 0.5211879014968872,
+      "rewards/rollout_reward_func/mean": -0.12103764712810516,
+      "rewards/rollout_reward_func/std": 0.5456745028495789,
+      "sampling/importance_sampling_ratio/max": 1.9942717552185059,
+      "sampling/importance_sampling_ratio/mean": 0.9991114139556885,
+      "sampling/importance_sampling_ratio/min": 0.11994559317827225,
+      "sampling/sampling_logp_difference/max": 2.1207170486450195,
+      "sampling/sampling_logp_difference/mean": 0.010312505066394806,
+      "step": 47,
+      "step_time": 94.76915054599158
+    },
+    {
+      "clip_ratio/high_max": 0.00876080454327166,
+      "clip_ratio/high_mean": 0.00438040227163583,
+      "clip_ratio/low_mean": 0.002355890115723014,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.006736292387358844,
+      "entropy": 0.08134529506787658,
+      "epoch": 0.00048,
+      "grad_norm": 0.805917501449585,
+      "kl": 0.16713016014546156,
+      "learning_rate": 3.2161862710939476e-05,
+      "loss": -0.0331,
+      "step": 48,
+      "step_time": 27.11353817199779
+    },
+    {
+      "clip_ratio/high_max": 0.004328028007876128,
+      "clip_ratio/high_mean": 0.002164014003938064,
+      "clip_ratio/low_mean": 0.0010534594766795635,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.003217473509721458,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 9558.0,
+      "completions/max_terminated_length": 9558.0,
+      "completions/mean_length": 8071.4375,
+      "completions/mean_terminated_length": 8071.4375,
+      "completions/min_length": 3757.0,
+      "completions/min_terminated_length": 3757.0,
+      "entropy": 0.09437589161098003,
+      "epoch": 0.00049,
+      "frac_reward_zero_std": 0.0,
+      "grad_norm": 1.223923921585083,
+      "kl": 0.17360988864675164,
+      "learning_rate": 2.8242045338402468e-05,
+      "loss": 0.0868,
+      "num_tokens": 6885078.0,
+      "reward": -0.17595598101615906,
+      "reward_std": 0.574884295463562,
+      "rewards/rollout_reward_func/mean": -0.17595598101615906,
+      "rewards/rollout_reward_func/std": 0.617475152015686,
+      "sampling/importance_sampling_ratio/max": 2.3295459747314453,
+      "sampling/importance_sampling_ratio/mean": 0.9994887709617615,
+      "sampling/importance_sampling_ratio/min": 0.21804022789001465,
+      "sampling/sampling_logp_difference/max": 1.5230756998062134,
+      "sampling/sampling_logp_difference/mean": 0.012081097811460495,
+      "step": 49,
+      "step_time": 92.44432610799413
+    },
+    {
+      "clip_ratio/high_max": 0.009452320169657469,
+      "clip_ratio/high_mean": 0.005073382286354899,
+      "clip_ratio/low_mean": 0.00776787509676069,
+      "clip_ratio/low_min": 0.002785580581985414,
+      "clip_ratio/region_mean": 0.012841257266700268,
+      "entropy": 0.09734066273085773,
+      "epoch": 0.0005,
+      "grad_norm": 1.335457682609558,
+      "kl": 0.17287601297721267,
+      "learning_rate": 2.5819464972482288e-05,
+      "loss": 0.0771,
+      "step": 50,
+      "step_time": 28.072686168001383
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 50,
+  "num_input_tokens_seen": 6885078,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e935e55bc95518f74954dabb1aeb2c31c84771846baba87cd5ac90c34aa6de35
+size 8145