qwen3-4b-biology-sdpo-ema005
This repository contains the last checkpoint (global_step_100) for qwen3gen-biology-SDPO-Qwen-Qwen3-4B-mbs32-ema0.05-train64-rollout8-lr1e-5-vllm0.8,
converted to Hugging Face Transformers format.
Evaluation
The reported headline score is the best validation mean@16 observed during training.
It is not necessarily the score of the uploaded last checkpoint.
| Dataset | Method | Model | Uploaded checkpoint | Best val mean@16 | Best step | Final val mean@16 |
|---|---|---|---|---|---|---|
| biology | SDPO | Qwen3-4B | global_step_100 | 52.25% | 20 | 52.00% |
Raw result files:
results/validation_mean16.csvresults/training_scores.csvartifacts/config.yamlartifacts/wandb-summary.json
Training Setup
- Base model:
Qwen/Qwen3-4B - Dataset:
biology - Method:
SDPO - EMA teacher update rate:
0.05 - Uploaded weights: last checkpoint,
global_step_100 - Validation metric used for headline score:
val-aux/*/mean@16 - Validation sampling:
n=16
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
repo_id = "SeongryongJung/qwen3-4b-biology-sdpo-ema005"
tokenizer = AutoTokenizer.from_pretrained(repo_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
repo_id,
torch_dtype="auto",
device_map="auto",
trust_remote_code=True,
)
Notes
Most best-validation intermediate checkpoints were not retained as full actor checkpoints because training kept only the latest actor checkpoint. Therefore, this repository publishes the last checkpoint and records the best validation score separately.