Qwen3-4B-Chemical-RLSD-TR/results/hyperparameters.csv

1	section	parameter	value	source
2	Run identity	Base model	Qwen/Qwen3-4B	queue/script override
3	Run identity	Dataset	Chemical / SciKnowEval chemistry	run_qwen3_generalization.sh
4	Run identity	Method	RLSD_TR	run_qwen3_generalization.sh
5	Run identity	Config	rlsd	run_qwen3_generalization.sh
6	Run identity	Experiment	qwen3gen-chemistry-RLSD_TR-Qwen-Qwen3-4B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8	run_qwen3_generalization.sh
7	Run identity	W&B run	run-20260702_062054-bz6p2yxy	wandb
8	Data	Train file	datasets/sciknoweval/chemistry/train.parquet	script override
9	Data	Validation file	datasets/sciknoweval/chemistry/test.parquet	script override
10	Data	Train batch size	32	queue/script override
11	Data	Train max samples	3200	queue/script override
12	Data	Prompt key	prompt	legacy_data.yaml default
13	Data	Reward key	data_source	legacy_data.yaml default
14	Data	Shuffle train data	True	user.yaml / legacy_data.yaml
15	Data	Validation shuffle	False	legacy_data.yaml default
16	Data	Filter overlong prompts	True	user.yaml
17	Data	Prompt truncation	error	legacy_data.yaml default
18	Data	enable_thinking	false	script override
19	Schedule	Total training steps	100	queue/script override
20	Schedule	Total epochs	30	ppo_trainer/user.yaml default
21	Schedule	Validation before train	False	queue/script override
22	Schedule	Save frequency	10	queue/script override
23	Schedule	Validation frequency	10	queue/script override
24	Sequence	Max prompt length	2048	queue/script override
25	Sequence	Max response length	8192	queue/script override
26	Sequence	Max model length	10240	queue/script override
27	Sequence	Actor max token length per GPU	10240	queue/script override
28	Rollout	Rollout engine	vllm	user.yaml
29	Rollout	Rollout dtype	bfloat16	rollout.yaml default
30	Rollout	Train rollout n	8	queue/script override
31	Rollout	Train rollout temperature	1.0	script override
32	Rollout	Train rollout top_p	1.0	script override
33	Rollout	Train rollout do_sample	True	rollout.yaml default
34	Rollout	Calculate rollout log probs	True	rlsd.yaml / script override
35	Rollout	Max num batched tokens	10240	queue/script override
36	Rollout	vLLM GPU memory utilization	0.8	queue/script override
37	Rollout	Tensor model parallel size	2	rollout.yaml default
38	Rollout	Free cache engine	True	rollout.yaml default
39	Validation	Validation rollout n	16	queue/script override
40	Validation	Validation temperature	0.6	queue/script override
41	Validation	Validation top_p	0.95	queue/script override
42	Validation	Validation do_sample	True	queue/script override
43	Optimization	Optimizer	AdamW	fsdp optimizer config
44	Optimization	Learning rate	1e-6	RLSD_TR method override
45	Optimization	LR scheduler	constant	W&B config
46	Optimization	LR warmup steps	10	script override
47	Optimization	Weight decay	0.01	script override
48	Optimization	Betas	(0.9, 0.999)	W&B config
49	Optimization	Gradient clip	1.0	script override
50	PPO/GRPO	Advantage estimator	grpo	rlsd.yaml
51	PPO/GRPO	Normalize GRPO advantages by std	False	script override
52	PPO/GRPO	PPO epochs	1	W&B config
53	PPO/GRPO	PPO mini batch size	8	queue/script override
54	PPO/GRPO	PPO micro batch size per GPU	1	user.yaml
55	PPO/GRPO	Clip ratio low	0.2	script override
56	PPO/GRPO	Clip ratio high	0.28	script override
57	PPO/GRPO	Gamma	1.0	ppo_trainer.yaml default
58	PPO/GRPO	Lambda	1.0	ppo_trainer.yaml default
59	PPO/GRPO	Use KL in reward	False	ppo_trainer/user.yaml
60	PPO/GRPO	Actor KL loss observed	0.0	output.log
61	Rollout correction	Importance sampling mode	token	script override
62	Rollout correction	IS threshold	2.0	script override
63	RLSD_TR	Policy loss mode	rlsd	method override
64	RLSD_TR	Teacher regularization	trust-region	method override
65	RLSD_TR	Trust-region mix / teacher update rate	0.1	queue/script override
66	RLSD_TR	Token reweight lambda	0.5	queue/script override
67	RLSD_TR	Token reweight eps_w	0.2	queue/script override
68	RLSD_TR	Token reweight decay steps	0	queue/script override
69	RLSD_TR	Max reprompt length	10240	method override
70	RLSD_TR	Fused kernels	False	method override
71	FSDP/System	Actor strategy	fsdp	dp_actor.yaml
72	FSDP/System	FSDP dtype	bfloat16	W&B config
73	FSDP/System	FSDP model dtype	fp32	W&B config
74	FSDP/System	Use torch compile	True	W&B config
75	FSDP/System	GPUs per node	8	queue/script override
76	FSDP/System	Nodes	1	user.yaml
77	FSDP/System	GPU type	NVIDIA H200	wandb-metadata
78	Checkpoint/Logging	Checkpoint root	checkpoints/datasets/sciknoweval/chemistry	script override
79	Checkpoint/Logging	Latest checkpointed iteration	100	latest_checkpointed_iteration.txt
80	Checkpoint/Logging	Max actor checkpoints to keep	1	user.yaml
81	Checkpoint/Logging	Logger	console, wandb	ppo_trainer.yaml
82	Checkpoint/Logging	W&B entity	seongryongjung-chung-ang-university	environment
83	Checkpoint/Logging	W&B project	SDPO-root	user.yaml project_name
84	Checkpoint/Logging	W&B group	QWEN3-RLSD-TR-GRPO-matched-generalization	method override