---
library_name: transformers
license: apache-2.0
base_model: Qwen/Qwen3-1.7B
tags:
- reinforcement-learning
- dapo
- math-reasoning
- pipelinerl
datasets:
- gsm8k_train
- math_train
pipeline_tag: text-generation
---

# Qwen3-1.7B-DAPO-math-reasoning

This model is a fine-tuned version of [Qwen3-1.7B](https://huggingface.co/Qwen/Qwen3-1.7B) using **DAPO (Decoupled Clip and Dynamic Sampling Policy Optimization) without KL penalty** for mathematical reasoning.

Trained with [PipelineRL](https://github.com/ServiceNow/PipelineRL).

## Training Details

### Datasets

| Split | Datasets |
|-------|----------|
| Train | `gsm8k_train`, `math_train` |
| Test  | `gsm8k_test`, `math_500` |

### RL Algorithm

| Parameter | Value |
|-----------|-------|
| Algorithm | DAPO (Decoupled Clip and Dynamic Sampling Policy Optimization) |
| Advantage Baseline | Group mean reward |
| Extra Inference | None |
| Group Structure | Required |
| Policy Loss | `ppo` |
| KL Coefficient | `0.0` |
| Epsilon (clip) | `0.2` |
| Discount Factor (`gamma`) | `1.0` |
| Divide Advantage by Std | `False` |
| Filter Zero Advantage Groups | `True` |
| Rollouts per Problem | `16` |

DAPO extends GRPO with clip-higher (asymmetric PPO clipping), dynamic sampling (filtering zero-variance groups), token-level loss aggregation, and overlong reward shaping.

### Training Hyperparameters

| Parameter | Value |
|-----------|-------|
| Base Model | `Qwen/Qwen3-1.7B` |
| Learning Rate | `1e-06` |
| LR Scheduler | `cosine` |
| Warmup Steps | `25` |
| Max Training Steps | `1500` |
| Micro Batch Size | `4` |
| Gradient Accumulation | `64` |
| Effective Batch Size | `256` |
| Sequence Length | `8192` |
| Gradient Clipping | `0.3` |
| Weight Decay | `0.01` |
| Optimizer | `adamw_torch` |
| Precision | `bf16` |
| DeepSpeed | ZeRO Stage 3 |

## Evaluation Results

Pass@k on math reasoning benchmarks (N=32 samples per problem, temperature=1.0):

| Dataset | pass@1 | pass@2 | pass@4 | pass@8 | pass@16 | pass@32 |
| --- | ---: | ---: | ---: | ---: | ---: | ---: |
| GSM8K (test) | 80.97 | 88.04 | 92.43 | 95.03 | 96.54 | 97.57 |
| MATH-500 | 65.77 | 74.57 | 81.06 | 85.79 | 89.17 | 91.60 |
| **Overall** | **76.79** | **84.34** | **89.30** | **92.49** | **94.52** | **95.93** |

*GSM8K test: 1319 problems · MATH-500: 500 problems · Overall: 1819 problems (overall weighted by problem count).*

## Training Curves

![Training Metrics](training_metrics.png)

## W&B Run

Full training logs: [https://wandb.ai/jaygala24-team/rl-post-training/runs/qwen3_1.7b_dapo_no_kl_3a1f_4xh100_236316_finetune_11e44a44](https://wandb.ai/jaygala24-team/rl-post-training/runs/qwen3_1.7b_dapo_no_kl_3a1f_4xh100_236316_finetune_11e44a44)

## Usage

### Transformers

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("jaygala24/Qwen3-1.7B-DAPO-math-reasoning", revision="step-0200")  # optional branch, e.g. "step-0400"
tokenizer = AutoTokenizer.from_pretrained("jaygala24/Qwen3-1.7B-DAPO-math-reasoning", revision="step-0200")

prompt = "Please reason step by step, and put your final answer within \\boxed{}.\n\nWhat is the sum of 123 and 456?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=4096, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

### vLLM

```python
from vllm import LLM, SamplingParams

llm = LLM(model="jaygala24/Qwen3-1.7B-DAPO-math-reasoning", revision="step-0200")  # optional branch, e.g. "step-0400"
sampling_params = SamplingParams(temperature=0.7, max_tokens=4096)

prompt = "Please reason step by step, and put your final answer within \boxed{}.

What is the sum of 123 and 456?"
outputs = llm.generate([prompt], sampling_params)
print(outputs[0].outputs[0].text)
```

## Framework

- [PipelineRL](https://github.com/ServiceNow/PipelineRL)
- [Transformers](https://github.com/huggingface/transformers)
- [DeepSpeed](https://github.com/microsoft/DeepSpeed) (ZeRO Stage 3)