Instructions to use kishan51/binary_grpo_checkpoint500 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kishan51/binary_grpo_checkpoint500 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/Qwen2.5-0.5B")
model = PeftModel.from_pretrained(base_model, "kishan51/binary_grpo_checkpoint500")

Transformers

How to use kishan51/binary_grpo_checkpoint500 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="kishan51/binary_grpo_checkpoint500")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("kishan51/binary_grpo_checkpoint500", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use kishan51/binary_grpo_checkpoint500 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kishan51/binary_grpo_checkpoint500"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kishan51/binary_grpo_checkpoint500",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kishan51/binary_grpo_checkpoint500

SGLang

How to use kishan51/binary_grpo_checkpoint500 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "kishan51/binary_grpo_checkpoint500" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kishan51/binary_grpo_checkpoint500",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "kishan51/binary_grpo_checkpoint500" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kishan51/binary_grpo_checkpoint500",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use kishan51/binary_grpo_checkpoint500 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for kishan51/binary_grpo_checkpoint500 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for kishan51/binary_grpo_checkpoint500 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for kishan51/binary_grpo_checkpoint500 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="kishan51/binary_grpo_checkpoint500",
    max_seq_length=2048,
)

Docker Model Runner
How to use kishan51/binary_grpo_checkpoint500 with Docker Model Runner:
```
docker model run hf.co/kishan51/binary_grpo_checkpoint500
```

kishan51 commited on Apr 12

Commit

f218ff9

verified ·

1 Parent(s): 2426b4f

Upload folder using huggingface_hub

Browse files

Files changed (16) hide show

.gitattributes +1 -0
README.md +210 -0
adapter_config.json +50 -0
adapter_model.safetensors +3 -0
added_tokens.json +25 -0
chat_template.jinja +4 -0
merges.txt +0 -0
optimizer.pt +3 -0
rng_state.pth +3 -0
scheduler.pt +3 -0
special_tokens_map.json +25 -0
tokenizer.json +3 -0
tokenizer_config.json +216 -0
trainer_state.json +1384 -0
training_args.bin +3 -0
vocab.json +0 -0

.gitattributes CHANGED Viewed

@@ -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

 *.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 ADDED Viewed

	@@ -0,0 +1,210 @@

+---
+base_model: unsloth/Qwen2.5-0.5B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:unsloth/Qwen2.5-0.5B
+- grpo
+- lora
+- transformers
+- trl
+- unsloth
+---
+# 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 ADDED Viewed

	@@ -0,0 +1,50 @@

+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": {
+    "base_model_class": "Qwen2ForCausalLM",
+    "parent_library": "transformers.models.qwen2.modeling_qwen2",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/Qwen2.5-0.5B",
+  "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": 0.0,
+  "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": [
+    "up_proj",
+    "q_proj",
+    "down_proj",
+    "k_proj",
+    "o_proj",
+    "gate_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:4f05aa03bc225c19cbf71c5dfc90ae3428521ceb8e760458dd8911dadf7ec9a6
+size 70430032

added_tokens.json ADDED Viewed

	@@ -0,0 +1,25 @@

+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|PAD_TOKEN|>": 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 ADDED Viewed

	@@ -0,0 +1,4 @@

+{% if messages[0]['role'] == 'system' %}{{ messages[0]['content'] + eos_token }}{% set loop_messages = messages[1:] %}{% else %}{{ 'You are given a problem.
+Think about the problem and provide your working out.
+Place it between <start_working_out> and <end_working_out>.
+Then, provide your solution between <SOLUTION></SOLUTION>' + eos_token }}{% set loop_messages = messages %}{% endif %}{% for message in loop_messages %}{% if message['role'] == 'user' %}{{ message['content'] }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<start_working_out>' }}{% endif %}

merges.txt ADDED Viewed

The diff for this file is too large to render. See raw diff

optimizer.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:52b85f4efbb568bf2cd1669daa8037e45f85e476f8c25a8930d2ecd4ef88316b
+size 36140069

rng_state.pth ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:1ca619867566630d37bb5d4ade20d201cd6de1a3031ae267cbe60aa7e9195a6c
+size 14645

scheduler.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:f369d8647b02587ea9d43b517f5ba41d7c98e522781f1b62206828d755e77757
+size 1465

special_tokens_map.json ADDED Viewed

	@@ -0,0 +1,25 @@

+{
+  "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": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|PAD_TOKEN|>"
+}

tokenizer.json ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:a11add1ccce473bcd467355ba4e231ba727eee54501528658b0d6fa18112957e
+size 11422253

tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,216 @@

+{
+  "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": "<|PAD_TOKEN|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "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": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 32768,
+  "pad_token": "<|PAD_TOKEN|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json ADDED Viewed

	@@ -0,0 +1,1384 @@

+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6666666666666666,
+  "eval_steps": 500,
+  "global_step": 500,
+  "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,
+      "completion_length": 1142.7,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1142.7,
+      "completions/max_terminated_length": 902.5,
+      "completions/mean_length": 314.125,
+      "completions/mean_terminated_length": 290.0325042724609,
+      "completions/min_length": 22.9,
+      "completions/min_terminated_length": 22.9,
+      "epoch": 0.013333333333333334,
+      "frac_reward_zero_std": 0.575,
+      "grad_norm": 0.03041459619998932,
+      "kl": 0.04187124692685681,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.0,
+      "num_tokens": 67628.0,
+      "reward": -2.325,
+      "reward_std": 0.23799361288547516,
+      "rewards/reward_func/mean": -2.325,
+      "rewards/reward_func/std": 0.37357009053230283,
+      "step": 10
+    },
+    {
+      "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,
+      "completion_length": 1404.2,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 1404.2,
+      "completions/max_terminated_length": 983.1,
+      "completions/mean_length": 350.225,
+      "completions/mean_terminated_length": 286.28453063964844,
+      "completions/min_length": 21.8,
+      "completions/min_terminated_length": 21.8,
+      "epoch": 0.02666666666666667,
+      "frac_reward_zero_std": 0.525,
+      "grad_norm": 77.34902954101562,
+      "kl": 0.17394654396921397,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.0002,
+      "num_tokens": 140212.0,
+      "reward": -2.31875,
+      "reward_std": 0.2846687763929367,
+      "rewards/reward_func/mean": -2.31875,
+      "rewards/reward_func/std": 0.4204031050205231,
+      "step": 20
+    },
+    {
+      "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,
+      "completion_length": 1156.8,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1156.8,
+      "completions/max_terminated_length": 965.6,
+      "completions/mean_length": 319.85625,
+      "completions/mean_terminated_length": 281.527326965332,
+      "completions/min_length": 21.8,
+      "completions/min_terminated_length": 21.8,
+      "epoch": 0.04,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 54.36892318725586,
+      "kl": 0.198792327567935,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0002,
+      "num_tokens": 207321.0,
+      "reward": -2.26875,
+      "reward_std": 0.38884909451007843,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.5729059964418411,
+      "step": 30
+    },
+    {
+      "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,
+      "completion_length": 1250.6,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1250.6,
+      "completions/max_terminated_length": 877.9,
+      "completions/mean_length": 273.88125,
+      "completions/mean_terminated_length": 233.0241729736328,
+      "completions/min_length": 19.5,
+      "completions/min_terminated_length": 19.5,
+      "epoch": 0.05333333333333334,
+      "frac_reward_zero_std": 0.525,
+      "grad_norm": 62.96562194824219,
+      "kl": 23.047201949357987,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.023,
+      "num_tokens": 268846.0,
+      "reward": -2.3375,
+      "reward_std": 0.2683012694120407,
+      "rewards/reward_func/mean": -2.3375,
+      "rewards/reward_func/std": 0.3865860253572464,
+      "step": 40
+    },
+    {
+      "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,
+      "completion_length": 1285.8,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 1285.8,
+      "completions/max_terminated_length": 984.4,
+      "completions/mean_length": 335.86875,
+      "completions/mean_terminated_length": 274.3717956542969,
+      "completions/min_length": 36.2,
+      "completions/min_terminated_length": 36.2,
+      "epoch": 0.06666666666666667,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 50.94906997680664,
+      "kl": 43.621040439605714,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0436,
+      "num_tokens": 339865.0,
+      "reward": -2.325,
+      "reward_std": 0.29716877937316893,
+      "rewards/reward_func/mean": -2.325,
+      "rewards/reward_func/std": 0.40061245262622835,
+      "step": 50
+    },
+    {
+      "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,
+      "completion_length": 1238.2,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1238.2,
+      "completions/max_terminated_length": 1031.5,
+      "completions/mean_length": 331.9375,
+      "completions/mean_terminated_length": 301.5768478393555,
+      "completions/min_length": 28.5,
+      "completions/min_terminated_length": 28.5,
+      "epoch": 0.08,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 54.65301513671875,
+      "kl": 18.56655087471008,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0186,
+      "num_tokens": 410479.0,
+      "reward": -2.24375,
+      "reward_std": 0.43427384495735166,
+      "rewards/reward_func/mean": -2.24375,
+      "rewards/reward_func/std": 0.6787217557430267,
+      "step": 60
+    },
+    {
+      "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,
+      "completion_length": 1068.7,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1068.7,
+      "completions/max_terminated_length": 728.7,
+      "completions/mean_length": 276.9625,
+      "completions/mean_terminated_length": 235.63208923339843,
+      "completions/min_length": 17.2,
+      "completions/min_terminated_length": 17.2,
+      "epoch": 0.09333333333333334,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 76.15401458740234,
+      "kl": 44.26115598678589,
+      "learning_rate": 4.7888888888888894e-06,
+      "loss": 0.0443,
+      "num_tokens": 472129.0,
+      "reward": -2.25625,
+      "reward_std": 0.3837158679962158,
+      "rewards/reward_func/mean": -2.25625,
+      "rewards/reward_func/std": 0.5701160997152328,
+      "step": 70
+    },
+    {
+      "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,
+      "completion_length": 949.2,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 949.2,
+      "completions/max_terminated_length": 865.0,
+      "completions/mean_length": 280.79375,
+      "completions/mean_terminated_length": 257.9735137939453,
+      "completions/min_length": 29.4,
+      "completions/min_terminated_length": 29.4,
+      "epoch": 0.10666666666666667,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 82.72232818603516,
+      "kl": 26.10418074131012,
+      "learning_rate": 4.677777777777778e-06,
+      "loss": 0.0261,
+      "num_tokens": 534296.0,
+      "reward": -2.225,
+      "reward_std": 0.5031255036592484,
+      "rewards/reward_func/mean": -2.225,
+      "rewards/reward_func/std": 0.6819553226232529,
+      "step": 80
+    },
+    {
+      "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,
+      "completion_length": 947.2,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 947.2,
+      "completions/max_terminated_length": 814.9,
+      "completions/mean_length": 277.56875,
+      "completions/mean_terminated_length": 261.88625183105466,
+      "completions/min_length": 27.9,
+      "completions/min_terminated_length": 27.9,
+      "epoch": 0.12,
+      "frac_reward_zero_std": 0.325,
+      "grad_norm": 69.23222351074219,
+      "kl": 22.169475841522218,
+      "learning_rate": 4.566666666666667e-06,
+      "loss": 0.0222,
+      "num_tokens": 595867.0,
+      "reward": -2.23125,
+      "reward_std": 0.43445787727832796,
+      "rewards/reward_func/mean": -2.23125,
+      "rewards/reward_func/std": 0.5476110845804214,
+      "step": 90
+    },
+    {
+      "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,
+      "completion_length": 1021.2,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1021.2,
+      "completions/max_terminated_length": 858.1,
+      "completions/mean_length": 291.4375,
+      "completions/mean_terminated_length": 265.9560745239258,
+      "completions/min_length": 25.5,
+      "completions/min_terminated_length": 25.5,
+      "epoch": 0.13333333333333333,
+      "frac_reward_zero_std": 0.6,
+      "grad_norm": 55.19097137451172,
+      "kl": 38.881617593765256,
+      "learning_rate": 4.455555555555555e-06,
+      "loss": 0.0389,
+      "num_tokens": 659557.0,
+      "reward": -2.25,
+      "reward_std": 0.3368710219860077,
+      "rewards/reward_func/mean": -2.25,
+      "rewards/reward_func/std": 0.5741409182548523,
+      "step": 100
+    },
+    {
+      "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,
+      "completion_length": 1103.4,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 1103.4,
+      "completions/max_terminated_length": 926.0,
+      "completions/mean_length": 286.44375,
+      "completions/mean_terminated_length": 269.99083709716797,
+      "completions/min_length": 23.5,
+      "completions/min_terminated_length": 23.5,
+      "epoch": 0.14666666666666667,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 44.06207275390625,
+      "kl": 10.720393800735474,
+      "learning_rate": 4.344444444444445e-06,
+      "loss": 0.0107,
+      "num_tokens": 722664.0,
+      "reward": -2.3,
+      "reward_std": 0.3366025388240814,
+      "rewards/reward_func/mean": -2.3,
+      "rewards/reward_func/std": 0.4434426426887512,
+      "step": 110
+    },
+    {
+      "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,
+      "completion_length": 762.9,
+      "completions/clipped_ratio": 0.0,
+      "completions/max_length": 762.9,
+      "completions/max_terminated_length": 762.9,
+      "completions/mean_length": 244.29375,
+      "completions/mean_terminated_length": 244.29375,
+      "completions/min_length": 28.0,
+      "completions/min_terminated_length": 28.0,
+      "epoch": 0.16,
+      "frac_reward_zero_std": 0.525,
+      "grad_norm": 98.8595199584961,
+      "kl": 3.3859562277793884,
+      "learning_rate": 4.233333333333334e-06,
+      "loss": 0.0034,
+      "num_tokens": 778203.0,
+      "reward": -2.28125,
+      "reward_std": 0.3689758062362671,
+      "rewards/reward_func/mean": -2.28125,
+      "rewards/reward_func/std": 0.5127301692962647,
+      "step": 120
+    },
+    {
+      "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,
+      "completion_length": 1152.1,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1152.1,
+      "completions/max_terminated_length": 764.6,
+      "completions/mean_length": 291.2875,
+      "completions/mean_terminated_length": 257.580419921875,
+      "completions/min_length": 21.6,
+      "completions/min_terminated_length": 21.6,
+      "epoch": 0.17333333333333334,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 113.74595642089844,
+      "kl": 5.273260289430619,
+      "learning_rate": 4.122222222222222e-06,
+      "loss": 0.0053,
+      "num_tokens": 842517.0,
+      "reward": -2.3,
+      "reward_std": 0.31371368169784547,
+      "rewards/reward_func/mean": -2.3,
+      "rewards/reward_func/std": 0.42710726857185366,
+      "step": 130
+    },
+    {
+      "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,
+      "completion_length": 1270.1,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1270.1,
+      "completions/max_terminated_length": 928.7,
+      "completions/mean_length": 317.6875,
+      "completions/mean_terminated_length": 270.543994140625,
+      "completions/min_length": 26.9,
+      "completions/min_terminated_length": 26.9,
+      "epoch": 0.18666666666666668,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 66.77887725830078,
+      "kl": 10.910183477401734,
+      "learning_rate": 4.011111111111111e-06,
+      "loss": 0.0109,
+      "num_tokens": 911419.0,
+      "reward": -2.20625,
+      "reward_std": 0.4734624236822128,
+      "rewards/reward_func/mean": -2.20625,
+      "rewards/reward_func/std": 0.7416160374879837,
+      "step": 140
+    },
+    {
+      "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,
+      "completion_length": 927.5,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 927.5,
+      "completions/max_terminated_length": 674.2,
+      "completions/mean_length": 267.95625,
+      "completions/mean_terminated_length": 243.36625213623046,
+      "completions/min_length": 22.6,
+      "completions/min_terminated_length": 22.6,
+      "epoch": 0.2,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 3.4511423110961914,
+      "kl": 7.64983651638031,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0076,
+      "num_tokens": 970928.0,
+      "reward": -2.325,
+      "reward_std": 0.2786566108465195,
+      "rewards/reward_func/mean": -2.325,
+      "rewards/reward_func/std": 0.3554343104362488,
+      "step": 150
+    },
+    {
+      "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,
+      "completion_length": 1125.2,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1125.2,
+      "completions/max_terminated_length": 787.6,
+      "completions/mean_length": 282.80625,
+      "completions/mean_terminated_length": 234.86697082519532,
+      "completions/min_length": 17.5,
+      "completions/min_terminated_length": 17.5,
+      "epoch": 0.21333333333333335,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 78.64956665039062,
+      "kl": 31.174254989624025,
+      "learning_rate": 3.7888888888888893e-06,
+      "loss": 0.0312,
+      "num_tokens": 1032913.0,
+      "reward": -2.26875,
+      "reward_std": 0.4256745457649231,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.6542905002832413,
+      "step": 160
+    },
+    {
+      "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,
+      "completion_length": 1234.9,
+      "completions/clipped_ratio": 0.04375,
+      "completions/max_length": 1234.9,
+      "completions/max_terminated_length": 850.8,
+      "completions/mean_length": 293.04375,
+      "completions/mean_terminated_length": 236.12315979003907,
+      "completions/min_length": 17.5,
+      "completions/min_terminated_length": 17.5,
+      "epoch": 0.22666666666666666,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 74.24949645996094,
+      "kl": 17.39362871646881,
+      "learning_rate": 3.6777777777777778e-06,
+      "loss": 0.0174,
+      "num_tokens": 1096696.0,
+      "reward": -2.2375,
+      "reward_std": 0.3502942889928818,
+      "rewards/reward_func/mean": -2.2375,
+      "rewards/reward_func/std": 0.5108247697353363,
+      "step": 170
+    },
+    {
+      "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,
+      "completion_length": 1094.1,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1094.1,
+      "completions/max_terminated_length": 862.9,
+      "completions/mean_length": 306.3875,
+      "completions/mean_terminated_length": 275.8087554931641,
+      "completions/min_length": 19.2,
+      "completions/min_terminated_length": 19.2,
+      "epoch": 0.24,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 114.97906494140625,
+      "kl": 3.446433424949646,
+      "learning_rate": 3.566666666666667e-06,
+      "loss": 0.0034,
+      "num_tokens": 1163210.0,
+      "reward": -2.26875,
+      "reward_std": 0.36489740014076233,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.5350149899721146,
+      "step": 180
+    },
+    {
+      "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,
+      "completion_length": 1129.5,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1129.5,
+      "completions/max_terminated_length": 810.2,
+      "completions/mean_length": 286.28125,
+      "completions/mean_terminated_length": 238.94435272216796,
+      "completions/min_length": 23.6,
+      "completions/min_terminated_length": 23.6,
+      "epoch": 0.25333333333333335,
+      "frac_reward_zero_std": 0.55,
+      "grad_norm": 103.09403228759766,
+      "kl": 20.17943916320801,
+      "learning_rate": 3.455555555555556e-06,
+      "loss": 0.0202,
+      "num_tokens": 1226187.0,
+      "reward": -2.35625,
+      "reward_std": 0.26175799071788786,
+      "rewards/reward_func/mean": -2.35625,
+      "rewards/reward_func/std": 0.3903583735227585,
+      "step": 190
+    },
+    {
+      "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,
+      "completion_length": 1171.8,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1171.8,
+      "completions/max_terminated_length": 780.4,
+      "completions/mean_length": 289.95,
+      "completions/mean_terminated_length": 242.0620880126953,
+      "completions/min_length": 21.9,
+      "completions/min_terminated_length": 21.9,
+      "epoch": 0.26666666666666666,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 162.74258422851562,
+      "kl": 39.28117027282715,
+      "learning_rate": 3.3444444444444445e-06,
+      "loss": 0.0393,
+      "num_tokens": 1290015.0,
+      "reward": -2.225,
+      "reward_std": 0.42369674146175385,
+      "rewards/reward_func/mean": -2.225,
+      "rewards/reward_func/std": 0.6618488401174545,
+      "step": 200
+    },
+    {
+      "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,
+      "completion_length": 953.0,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 953.0,
+      "completions/max_terminated_length": 883.8,
+      "completions/mean_length": 306.9625,
+      "completions/mean_terminated_length": 285.81047973632815,
+      "completions/min_length": 20.9,
+      "completions/min_terminated_length": 20.9,
+      "epoch": 0.28,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 56.53829574584961,
+      "kl": 4.3735075235366825,
+      "learning_rate": 3.2333333333333334e-06,
+      "loss": 0.0044,
+      "num_tokens": 1355825.0,
+      "reward": -2.30625,
+      "reward_std": 0.2741025388240814,
+      "rewards/reward_func/mean": -2.30625,
+      "rewards/reward_func/std": 0.3696858823299408,
+      "step": 210
+    },
+    {
+      "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,
+      "completion_length": 1269.1,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1269.1,
+      "completions/max_terminated_length": 1032.2,
+      "completions/mean_length": 292.15625,
+      "completions/mean_terminated_length": 260.2483383178711,
+      "completions/min_length": 17.8,
+      "completions/min_terminated_length": 17.8,
+      "epoch": 0.29333333333333333,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 74.96534729003906,
+      "kl": 16.07765164375305,
+      "learning_rate": 3.1222222222222228e-06,
+      "loss": 0.0161,
+      "num_tokens": 1418942.0,
+      "reward": -2.26875,
+      "reward_std": 0.35119174122810365,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.48719866275787355,
+      "step": 220
+    },
+    {
+      "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,
+      "completion_length": 1035.2,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1035.2,
+      "completions/max_terminated_length": 911.0,
+      "completions/mean_length": 256.95,
+      "completions/mean_terminated_length": 233.14875335693358,
+      "completions/min_length": 8.9,
+      "completions/min_terminated_length": 8.9,
+      "epoch": 0.30666666666666664,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 36.64857482910156,
+      "kl": 14.746024799346923,
+      "learning_rate": 3.0111111111111113e-06,
+      "loss": 0.0147,
+      "num_tokens": 1478462.0,
+      "reward": -2.225,
+      "reward_std": 0.4513490855693817,
+      "rewards/reward_func/mean": -2.225,
+      "rewards/reward_func/std": 0.6648978263139724,
+      "step": 230
+    },
+    {
+      "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,
+      "completion_length": 1116.4,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1116.4,
+      "completions/max_terminated_length": 758.3,
+      "completions/mean_length": 295.925,
+      "completions/mean_terminated_length": 255.58333892822264,
+      "completions/min_length": 21.1,
+      "completions/min_terminated_length": 21.1,
+      "epoch": 0.32,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 84.84314727783203,
+      "kl": 5.60067400932312,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0056,
+      "num_tokens": 1543490.0,
+      "reward": -2.26875,
+      "reward_std": 0.35537946224212646,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.4750110626220703,
+      "step": 240
+    },
+    {
+      "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,
+      "completion_length": 1023.8,
+      "completions/clipped_ratio": 0.00625,
+      "completions/max_length": 1023.8,
+      "completions/max_terminated_length": 945.7,
+      "completions/mean_length": 266.0375,
+      "completions/mean_terminated_length": 257.5279174804688,
+      "completions/min_length": 17.0,
+      "completions/min_terminated_length": 17.0,
+      "epoch": 0.3333333333333333,
+      "frac_reward_zero_std": 0.575,
+      "grad_norm": 81.47051239013672,
+      "kl": 18.14775743484497,
+      "learning_rate": 2.788888888888889e-06,
+      "loss": 0.0181,
+      "num_tokens": 1603676.0,
+      "reward": -2.3375,
+      "reward_std": 0.2645611822605133,
+      "rewards/reward_func/mean": -2.3375,
+      "rewards/reward_func/std": 0.4000595390796661,
+      "step": 250
+    },
+    {
+      "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,
+      "completion_length": 1079.9,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1079.9,
+      "completions/max_terminated_length": 947.9,
+      "completions/mean_length": 340.54375,
+      "completions/mean_terminated_length": 318.7557769775391,
+      "completions/min_length": 26.2,
+      "completions/min_terminated_length": 26.2,
+      "epoch": 0.3466666666666667,
+      "frac_reward_zero_std": 0.55,
+      "grad_norm": 54.74143600463867,
+      "kl": 20.487924671173097,
+      "learning_rate": 2.677777777777778e-06,
+      "loss": 0.0205,
+      "num_tokens": 1675795.0,
+      "reward": -2.3125,
+      "reward_std": 0.2760362923145294,
+      "rewards/reward_func/mean": -2.3125,
+      "rewards/reward_func/std": 0.3895877480506897,
+      "step": 260
+    },
+    {
+      "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,
+      "completion_length": 1192.1,
+      "completions/clipped_ratio": 0.04375,
+      "completions/max_length": 1192.1,
+      "completions/max_terminated_length": 858.8,
+      "completions/mean_length": 302.7625,
+      "completions/mean_terminated_length": 246.45244750976562,
+      "completions/min_length": 12.2,
+      "completions/min_terminated_length": 12.2,
+      "epoch": 0.36,
+      "frac_reward_zero_std": 0.65,
+      "grad_norm": 0.1731022149324417,
+      "kl": 7.915937983989716,
+      "learning_rate": 2.566666666666667e-06,
+      "loss": 0.0079,
+      "num_tokens": 1740341.0,
+      "reward": -2.25,
+      "reward_std": 0.4122216373682022,
+      "rewards/reward_func/mean": -2.25,
+      "rewards/reward_func/std": 0.6856489062309266,
+      "step": 270
+    },
+    {
+      "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,
+      "completion_length": 1110.5,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1110.5,
+      "completions/max_terminated_length": 891.8,
+      "completions/mean_length": 309.71875,
+      "completions/mean_terminated_length": 286.5383361816406,
+      "completions/min_length": 24.0,
+      "completions/min_terminated_length": 24.0,
+      "epoch": 0.37333333333333335,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 122.04984283447266,
+      "kl": 3.3861916780471804,
+      "learning_rate": 2.455555555555556e-06,
+      "loss": 0.0034,
+      "num_tokens": 1806904.0,
+      "reward": -2.2375,
+      "reward_std": 0.4326493889093399,
+      "rewards/reward_func/mean": -2.2375,
+      "rewards/reward_func/std": 0.6847750276327134,
+      "step": 280
+    },
+    {
+      "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,
+      "completion_length": 1069.2,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1069.2,
+      "completions/max_terminated_length": 853.0,
+      "completions/mean_length": 301.1875,
+      "completions/mean_terminated_length": 269.63209075927733,
+      "completions/min_length": 18.7,
+      "completions/min_terminated_length": 18.7,
+      "epoch": 0.38666666666666666,
+      "frac_reward_zero_std": 0.575,
+      "grad_norm": 25.8874454498291,
+      "kl": 14.604678845405578,
+      "learning_rate": 2.3444444444444448e-06,
+      "loss": 0.0146,
+      "num_tokens": 1872494.0,
+      "reward": -2.36875,
+      "reward_std": 0.22780319154262543,
+      "rewards/reward_func/mean": -2.36875,
+      "rewards/reward_func/std": 0.3246472626924515,
+      "step": 290
+    },
+    {
+      "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,
+      "completion_length": 1228.1,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1228.1,
+      "completions/max_terminated_length": 817.4,
+      "completions/mean_length": 287.74375,
+      "completions/mean_terminated_length": 238.6010757446289,
+      "completions/min_length": 20.4,
+      "completions/min_terminated_length": 20.4,
+      "epoch": 0.4,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 50.89199447631836,
+      "kl": 12.896457767486572,
+      "learning_rate": 2.2333333333333333e-06,
+      "loss": 0.0129,
+      "num_tokens": 1936649.0,
+      "reward": -2.26875,
+      "reward_std": 0.3961044609546661,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.5396905303001404,
+      "step": 300
+    },
+    {
+      "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,
+      "completion_length": 1038.1,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 1038.1,
+      "completions/max_terminated_length": 889.8,
+      "completions/mean_length": 258.28125,
+      "completions/mean_terminated_length": 241.79458618164062,
+      "completions/min_length": 17.1,
+      "completions/min_terminated_length": 17.1,
+      "epoch": 0.41333333333333333,
+      "frac_reward_zero_std": 0.65,
+      "grad_norm": 48.88945770263672,
+      "kl": 4.541426551342011,
+      "learning_rate": 2.1222222222222226e-06,
+      "loss": 0.0045,
+      "num_tokens": 1994798.0,
+      "reward": -2.4,
+      "reward_std": 0.17886751294136047,
+      "rewards/reward_func/mean": -2.4,
+      "rewards/reward_func/std": 0.2989355862140656,
+      "step": 310
+    },
+    {
+      "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,
+      "completion_length": 1238.9,
+      "completions/clipped_ratio": 0.05625,
+      "completions/max_length": 1238.9,
+      "completions/max_terminated_length": 795.1,
+      "completions/mean_length": 315.44375,
+      "completions/mean_terminated_length": 243.35834197998048,
+      "completions/min_length": 15.4,
+      "completions/min_terminated_length": 15.4,
+      "epoch": 0.4266666666666667,
+      "frac_reward_zero_std": 0.525,
+      "grad_norm": 54.9058837890625,
+      "kl": 4.1825002312660216,
+      "learning_rate": 2.011111111111111e-06,
+      "loss": 0.0042,
+      "num_tokens": 2062905.0,
+      "reward": -2.29375,
+      "reward_std": 0.34910253882408143,
+      "rewards/reward_func/mean": -2.29375,
+      "rewards/reward_func/std": 0.538089805841446,
+      "step": 320
+    },
+    {
+      "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,
+      "completion_length": 1048.0,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 1048.0,
+      "completions/max_terminated_length": 917.3,
+      "completions/mean_length": 277.575,
+      "completions/mean_terminated_length": 261.64583587646484,
+      "completions/min_length": 19.7,
+      "completions/min_terminated_length": 19.7,
+      "epoch": 0.44,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 127.28370666503906,
+      "kl": 4.280037522315979,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.0043,
+      "num_tokens": 2124265.0,
+      "reward": -2.2375,
+      "reward_std": 0.44838539361953733,
+      "rewards/reward_func/mean": -2.2375,
+      "rewards/reward_func/std": 0.6401138246059418,
+      "step": 330
+    },
+    {
+      "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,
+      "completion_length": 927.5,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 927.5,
+      "completions/max_terminated_length": 798.2,
+      "completions/mean_length": 283.7125,
+      "completions/mean_terminated_length": 267.95792083740236,
+      "completions/min_length": 30.1,
+      "completions/min_terminated_length": 30.1,
+      "epoch": 0.4533333333333333,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 146.22303771972656,
+      "kl": 30.70813355445862,
+      "learning_rate": 1.788888888888889e-06,
+      "loss": 0.0307,
+      "num_tokens": 2187843.0,
+      "reward": -2.25625,
+      "reward_std": 0.39784331917762755,
+      "rewards/reward_func/mean": -2.25625,
+      "rewards/reward_func/std": 0.5563936829566956,
+      "step": 340
+    },
+    {
+      "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,
+      "completion_length": 1043.1,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1043.1,
+      "completions/max_terminated_length": 899.7,
+      "completions/mean_length": 307.55625,
+      "completions/mean_terminated_length": 284.3370895385742,
+      "completions/min_length": 24.6,
+      "completions/min_terminated_length": 24.6,
+      "epoch": 0.4666666666666667,
+      "frac_reward_zero_std": 0.55,
+      "grad_norm": 32.36188888549805,
+      "kl": 64.3428849220276,
+      "learning_rate": 1.6777777777777779e-06,
+      "loss": 0.0643,
+      "num_tokens": 2254616.0,
+      "reward": -2.26875,
+      "reward_std": 0.3944930166006088,
+      "rewards/reward_func/mean": -2.26875,
+      "rewards/reward_func/std": 0.6516756772994995,
+      "step": 350
+    },
+    {
+      "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,
+      "completion_length": 1245.1,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1245.1,
+      "completions/max_terminated_length": 964.9,
+      "completions/mean_length": 309.43125,
+      "completions/mean_terminated_length": 261.41810150146483,
+      "completions/min_length": 27.0,
+      "completions/min_terminated_length": 27.0,
+      "epoch": 0.48,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 38.63202667236328,
+      "kl": 21.073752880096436,
+      "learning_rate": 1.566666666666667e-06,
+      "loss": 0.0211,
+      "num_tokens": 2320821.0,
+      "reward": -2.275,
+      "reward_std": 0.39782515466213225,
+      "rewards/reward_func/mean": -2.275,
+      "rewards/reward_func/std": 0.5283936500549317,
+      "step": 360
+    },
+    {
+      "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,
+      "completion_length": 1342.0,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1342.0,
+      "completions/max_terminated_length": 1053.9,
+      "completions/mean_length": 318.3625,
+      "completions/mean_terminated_length": 278.52917327880857,
+      "completions/min_length": 21.3,
+      "completions/min_terminated_length": 21.3,
+      "epoch": 0.49333333333333335,
+      "frac_reward_zero_std": 0.4,
+      "grad_norm": 45.37665557861328,
+      "kl": 4.191655397415161,
+      "learning_rate": 1.4555555555555557e-06,
+      "loss": 0.0042,
+      "num_tokens": 2388327.0,
+      "reward": -2.24375,
+      "reward_std": 0.40966878533363343,
+      "rewards/reward_func/mean": -2.24375,
+      "rewards/reward_func/std": 0.5739072471857071,
+      "step": 370
+    },
+    {
+      "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,
+      "completion_length": 1146.1,
+      "completions/clipped_ratio": 0.05,
+      "completions/max_length": 1146.1,
+      "completions/max_terminated_length": 806.6,
+      "completions/mean_length": 309.30625,
+      "completions/mean_terminated_length": 246.01378784179687,
+      "completions/min_length": 14.3,
+      "completions/min_terminated_length": 14.3,
+      "epoch": 0.5066666666666667,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 51.08633804321289,
+      "kl": 10.670231366157532,
+      "learning_rate": 1.3444444444444446e-06,
+      "loss": 0.0107,
+      "num_tokens": 2454412.0,
+      "reward": -2.28125,
+      "reward_std": 0.39523502588272097,
+      "rewards/reward_func/mean": -2.28125,
+      "rewards/reward_func/std": 0.5278610855340957,
+      "step": 380
+    },
+    {
+      "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,
+      "completion_length": 1179.6,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 1179.6,
+      "completions/max_terminated_length": 819.6,
+      "completions/mean_length": 301.68125,
+      "completions/mean_terminated_length": 261.32500610351565,
+      "completions/min_length": 18.2,
+      "completions/min_terminated_length": 18.2,
+      "epoch": 0.52,
+      "frac_reward_zero_std": 0.525,
+      "grad_norm": 106.32160949707031,
+      "kl": 21.585847473144533,
+      "learning_rate": 1.2333333333333335e-06,
+      "loss": 0.0216,
+      "num_tokens": 2519509.0,
+      "reward": -2.29375,
+      "reward_std": 0.3345831483602524,
+      "rewards/reward_func/mean": -2.29375,
+      "rewards/reward_func/std": 0.5108768463134765,
+      "step": 390
+    },
+    {
+      "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,
+      "completion_length": 995.2,
+      "completions/clipped_ratio": 0.03125,
+      "completions/max_length": 995.2,
+      "completions/max_terminated_length": 679.8,
+      "completions/mean_length": 274.5875,
+      "completions/mean_terminated_length": 232.36595611572267,
+      "completions/min_length": 17.9,
+      "completions/min_terminated_length": 17.9,
+      "epoch": 0.5333333333333333,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 75.17499542236328,
+      "kl": 9.158978390693665,
+      "learning_rate": 1.1222222222222222e-06,
+      "loss": 0.0092,
+      "num_tokens": 2579431.0,
+      "reward": -2.1875,
+      "reward_std": 0.4814758121967316,
+      "rewards/reward_func/mean": -2.1875,
+      "rewards/reward_func/std": 0.6688827067613602,
+      "step": 400
+    },
+    {
+      "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,
+      "completion_length": 1195.5,
+      "completions/clipped_ratio": 0.025,
+      "completions/max_length": 1195.5,
+      "completions/max_terminated_length": 997.3,
+      "completions/mean_length": 286.84375,
+      "completions/mean_terminated_length": 254.71875457763673,
+      "completions/min_length": 15.5,
+      "completions/min_terminated_length": 15.5,
+      "epoch": 0.5466666666666666,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 95.73103332519531,
+      "kl": 8.273300886154175,
+      "learning_rate": 1.0111111111111111e-06,
+      "loss": 0.0083,
+      "num_tokens": 2641798.0,
+      "reward": -2.25,
+      "reward_std": 0.39309037029743193,
+      "rewards/reward_func/mean": -2.25,
+      "rewards/reward_func/std": 0.5219055712223053,
+      "step": 410
+    },
+    {
+      "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,
+      "completion_length": 974.9,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 974.9,
+      "completions/max_terminated_length": 869.7,
+      "completions/mean_length": 261.23125,
+      "completions/mean_terminated_length": 244.99042053222655,
+      "completions/min_length": 16.0,
+      "completions/min_terminated_length": 16.0,
+      "epoch": 0.56,
+      "frac_reward_zero_std": 0.375,
+      "grad_norm": 99.72528839111328,
+      "kl": 6.985241413116455,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.007,
+      "num_tokens": 2700271.0,
+      "reward": -2.23125,
+      "reward_std": 0.4483605235815048,
+      "rewards/reward_func/mean": -2.23125,
+      "rewards/reward_func/std": 0.5773171722888947,
+      "step": 420
+    },
+    {
+      "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,
+      "completion_length": 964.0,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 964.0,
+      "completions/max_terminated_length": 776.9,
+      "completions/mean_length": 280.24375,
+      "completions/mean_terminated_length": 263.74458465576174,
+      "completions/min_length": 14.2,
+      "completions/min_terminated_length": 14.2,
+      "epoch": 0.5733333333333334,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 124.6417007446289,
+      "kl": 11.635133910179139,
+      "learning_rate": 7.888888888888889e-07,
+      "loss": 0.0116,
+      "num_tokens": 2762894.0,
+      "reward": -2.1875,
+      "reward_std": 0.5182177871465683,
+      "rewards/reward_func/mean": -2.1875,
+      "rewards/reward_func/std": 0.7841762661933899,
+      "step": 430
+    },
+    {
+      "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,
+      "completion_length": 1059.5,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1059.5,
+      "completions/max_terminated_length": 908.1,
+      "completions/mean_length": 311.675,
+      "completions/mean_terminated_length": 289.6191696166992,
+      "completions/min_length": 21.8,
+      "completions/min_terminated_length": 21.8,
+      "epoch": 0.5866666666666667,
+      "frac_reward_zero_std": 0.475,
+      "grad_norm": 132.1418914794922,
+      "kl": 115.09008512496948,
+      "learning_rate": 6.777777777777779e-07,
+      "loss": 0.1151,
+      "num_tokens": 2830866.0,
+      "reward": -2.35,
+      "reward_std": 0.26443375647068024,
+      "rewards/reward_func/mean": -2.35,
+      "rewards/reward_func/std": 0.35543124079704286,
+      "step": 440
+    },
+    {
+      "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,
+      "completion_length": 1278.8,
+      "completions/clipped_ratio": 0.0625,
+      "completions/max_length": 1278.8,
+      "completions/max_terminated_length": 777.6,
+      "completions/mean_length": 327.28125,
+      "completions/mean_terminated_length": 247.8854248046875,
+      "completions/min_length": 25.5,
+      "completions/min_terminated_length": 25.5,
+      "epoch": 0.6,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 121.0811538696289,
+      "kl": 21.60447826385498,
+      "learning_rate": 5.666666666666667e-07,
+      "loss": 0.0216,
+      "num_tokens": 2900423.0,
+      "reward": -2.24375,
+      "reward_std": 0.4444930166006088,
+      "rewards/reward_func/mean": -2.24375,
+      "rewards/reward_func/std": 0.6568617165088654,
+      "step": 450
+    },
+    {
+      "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,
+      "completion_length": 1040.1,
+      "completions/clipped_ratio": 0.01875,
+      "completions/max_length": 1040.1,
+      "completions/max_terminated_length": 760.2,
+      "completions/mean_length": 274.375,
+      "completions/mean_terminated_length": 249.7200035095215,
+      "completions/min_length": 19.1,
+      "completions/min_terminated_length": 19.1,
+      "epoch": 0.6133333333333333,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 38.948699951171875,
+      "kl": 14.533638858795166,
+      "learning_rate": 4.5555555555555563e-07,
+      "loss": 0.0145,
+      "num_tokens": 2961131.0,
+      "reward": -2.3125,
+      "reward_std": 0.31917070150375365,
+      "rewards/reward_func/mean": -2.3125,
+      "rewards/reward_func/std": 0.4087942600250244,
+      "step": 460
+    },
+    {
+      "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,
+      "completion_length": 1305.9,
+      "completions/clipped_ratio": 0.04375,
+      "completions/max_length": 1305.9,
+      "completions/max_terminated_length": 1012.3,
+      "completions/mean_length": 315.675,
+      "completions/mean_terminated_length": 259.3705429077148,
+      "completions/min_length": 18.0,
+      "completions/min_terminated_length": 18.0,
+      "epoch": 0.6266666666666667,
+      "frac_reward_zero_std": 0.35,
+      "grad_norm": 60.82999038696289,
+      "kl": 27.974109888076782,
+      "learning_rate": 3.444444444444445e-07,
+      "loss": 0.028,
+      "num_tokens": 3028239.0,
+      "reward": -2.24375,
+      "reward_std": 0.3991025358438492,
+      "rewards/reward_func/mean": -2.24375,
+      "rewards/reward_func/std": 0.5254836320877075,
+      "step": 470
+    },
+    {
+      "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,
+      "completion_length": 984.0,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 984.0,
+      "completions/max_terminated_length": 822.9,
+      "completions/mean_length": 262.79375,
+      "completions/mean_terminated_length": 246.77666931152345,
+      "completions/min_length": 21.2,
+      "completions/min_terminated_length": 21.2,
+      "epoch": 0.64,
+      "frac_reward_zero_std": 0.425,
+      "grad_norm": 104.03584289550781,
+      "kl": 13.309299659729003,
+      "learning_rate": 2.3333333333333336e-07,
+      "loss": 0.0133,
+      "num_tokens": 3087690.0,
+      "reward": -2.25625,
+      "reward_std": 0.4189758121967316,
+      "rewards/reward_func/mean": -2.25625,
+      "rewards/reward_func/std": 0.5928926885128021,
+      "step": 480
+    },
+    {
+      "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,
+      "completion_length": 999.4,
+      "completions/clipped_ratio": 0.0125,
+      "completions/max_length": 999.4,
+      "completions/max_terminated_length": 889.3,
+      "completions/mean_length": 273.45625,
+      "completions/mean_terminated_length": 257.5420852661133,
+      "completions/min_length": 34.7,
+      "completions/min_terminated_length": 34.7,
+      "epoch": 0.6533333333333333,
+      "frac_reward_zero_std": 0.6,
+      "grad_norm": 56.064544677734375,
+      "kl": 25.44476251602173,
+      "learning_rate": 1.2222222222222225e-07,
+      "loss": 0.0254,
+      "num_tokens": 3148447.0,
+      "reward": -2.3625,
+      "reward_std": 0.23339140117168428,
+      "rewards/reward_func/mean": -2.3625,
+      "rewards/reward_func/std": 0.37498362064361573,
+      "step": 490
+    },
+    {
+      "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,
+      "completion_length": 1092.4,
+      "completions/clipped_ratio": 0.0375,
+      "completions/max_length": 1092.4,
+      "completions/max_terminated_length": 675.7,
+      "completions/mean_length": 294.625,
+      "completions/mean_terminated_length": 246.04232788085938,
+      "completions/min_length": 19.8,
+      "completions/min_terminated_length": 19.8,
+      "epoch": 0.6666666666666666,
+      "frac_reward_zero_std": 0.45,
+      "grad_norm": 122.21601104736328,
+      "kl": 17.26050043106079,
+      "learning_rate": 1.1111111111111112e-08,
+      "loss": 0.0173,
+      "num_tokens": 3212059.0,
+      "reward": -2.2125,
+      "reward_std": 0.4100087761878967,
+      "rewards/reward_func/mean": -2.2125,
+      "rewards/reward_func/std": 0.5889766067266464,
+      "step": 500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 500,
+  "num_input_tokens_seen": 3212059,
+  "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": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:e52bdd343fc366e7d6fe3e5b14ca8c8b075ddcc6067471bdd979e8c93df1a449
+size 7441

vocab.json ADDED Viewed

The diff for this file is too large to render. See raw diff