Modelpublic

SeongryongJung/qwen3-8b-tooluse-rlsd-ema005

sourceHugging Faceapache-2.0updated 3mo agoView on Hugging Face

0likes8downloads

Model Card

qwen3-8b-tooluse-rlsd-ema005

Fine-tuned from Qwen/Qwen3-8B with RLSD (EMA 0.05) on the tooluse split.

Validation Performance

Metric: val-aux/tooluse/reward/mean@16 from 10-step validation logs.

best mean@16	best step	final mean@16	final step
66.73%	60	64.06%	100

[image]

step	mean@16
10	58.82%
20	62.22%
30	63.14%
40	63.05%
50	65.99%
60	66.73%
70	63.33%
80	63.05%
90	65.99%
100	64.06%

Files included with this repo:

—metrics.json: parsed validation summary
—eval_mean16.csv: step-level validation curve data
—eval_mean16.png: validation curve plot

Important: the uploaded weights are the final global_step_100/actor checkpoint. If best step is earlier than 100, the best validation point is reported for tracking, but the corresponding actor weights may not be retained locally.

Checkpoint source: /mnt/mole/SDPO/L2T/checkpoints/datasets/tooluse/qwen3gen-tooluse-RLSD-Qwen-Qwen3-8B-mbs8-decay0-ema0.05-train64-rollout8-lr1e-6-vllm0.8

W&B run: run-20260701_061001-b0l3abu5

This upload uses global_step_100/actor converted from VERL FSDP shards to Hugging Face format.