初始化项目，由ModelHub XC社区提供模型

Model: SeongryongJung/Qwen3-8B-Biology-RLSD-TR
Source: Original Platform

results/hyperparameters.csv

@@ -0,0 +1,38 @@
+section,parameter,value,source
+Run identity,Base model,Qwen/Qwen3-8B,queue/script override
+Run identity,Dataset,Biology / SciKnowEval biology,run_qwen3_generalization.sh
+Run identity,Method,RLSD_TR,run_qwen3_generalization.sh
+Run identity,Config,rlsd,run_qwen3_generalization.sh
+Run identity,Experiment,qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8,run_qwen3_generalization.sh
+Run identity,W&B run,run-20260703_091731-gj54iy5b,wandb
+Data,Train file,datasets/sciknoweval/biology/train.parquet,script override
+Data,Validation file,datasets/sciknoweval/biology/test.parquet,script override
+Data,Train batch size,32,queue/script override
+Data,Train max samples,3200,queue/script override
+Schedule,Total training steps,100,queue/script override
+Schedule,Validation before train,False,queue/script override
+Schedule,Save frequency,10,queue/script override
+Schedule,Validation frequency,10,queue/script override
+Sequence,Max prompt length,2048,queue/script override
+Sequence,Max response length,8192,queue/script override
+Sequence,Max model length,10240,queue/script override
+Rollout,Train rollout n,8,queue/script override
+Rollout,Validation rollout n,16,queue/script override
+Rollout,vLLM GPU memory utilization,0.8,queue/script override
+Optimization,Learning rate,1e-6,RLSD_TR method override
+Optimization,Weight decay,0.01,script override
+PPO/GRPO,PPO mini batch size,8,queue/script override
+PPO/GRPO,Normalize GRPO advantages by std,False,baseline_grpo.yaml / script override
+Rollout correction,Importance sampling mode,token,script override
+Rollout correction,IS threshold,2.0,script override
+Checkpoint/Logging,Checkpoint root,checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8,script override
+Checkpoint/Logging,Latest checkpointed iteration,100,latest_checkpointed_iteration.txt
+Checkpoint/Logging,External actor archive,checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8/_actor_archive,preserve_actor_checkpoints.py
+Checkpoint/Logging,Logger,"console, wandb",ppo_trainer.yaml
+RLSD_TR,Policy loss mode,rlsd,method override
+RLSD_TR,Teacher regularization,trust-region,method override
+RLSD_TR,Trust-region mix / teacher update rate,0.1,queue/script override
+RLSD_TR,Token reweight lambda,0.5,queue/script override
+RLSD_TR,Token reweight eps_w,0.2,queue/script override
+RLSD_TR,Token reweight decay steps,0,queue/script override
+RLSD_TR,Fused kernels,False,method override

results/summary.json

@@ -0,0 +1,29 @@
+{
+  "repo_id": "SeongryongJung/Qwen3-8B-Biology-RLSD-TR",
+  "output_dir": "/mnt/mole/SDPO/L2T/hf_upload_tr/Qwen3-8B-Biology-RLSD-TR",
+  "experiment": "qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8",
+  "best_step": 100,
+  "best_val_mean16": 0.55,
+  "best_actor_dir": "/mnt/mole/SDPO/L2T/checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8/global_step_100/actor",
+  "final_step": 100,
+  "final_val_mean16": 0.55,
+  "final_actor_dir": "/mnt/mole/SDPO/L2T/checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-8B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8/global_step_100/actor",
+  "train_rows": 90,
+  "val_rows": 10,
+  "hf_model_files": [
+    "added_tokens.json",
+    "chat_template.jinja",
+    "config.json",
+    "generation_config.json",
+    "merges.txt",
+    "model.safetensors.index.json",
+    "special_tokens_map.json",
+    "tokenizer.json",
+    "tokenizer_config.json",
+    "vocab.json",
+    "model-00001-of-00004.safetensors",
+    "model-00002-of-00004.safetensors",
+    "model-00003-of-00004.safetensors",
+    "model-00004-of-00004.safetensors"
+  ]
+}

results/training_score.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caf9b314980f68e1c8e8613236817b2043fb8e62c1d53af3c3c7beca9c372fc1
+size 126914

results/training_scores.csv

@@ -0,0 +1,91 @@
+step,critic_score_mean
+1,0.33984375
+2,0.3828125
+3,0.44921875
+4,0.37890625
+5,0.44921875
+6,0.34375
+7,0.2890625
+8,0.2421875
+9,0.34765625
+11,0.42578125
+12,0.421875
+13,0.3203125
+14,0.29296875
+15,0.3671875
+16,0.40234375
+17,0.30078125
+18,0.4296875
+19,0.3828125
+21,0.4140625
+22,0.4296875
+23,0.33984375
+24,0.453125
+25,0.4296875
+26,0.38671875
+27,0.359375
+28,0.375
+29,0.4375
+31,0.31640625
+32,0.5
+33,0.41015625
+34,0.390625
+35,0.4765625
+36,0.390625
+37,0.4140625
+38,0.44140625
+39,0.4765625
+41,0.45703125
+42,0.3515625
+43,0.56640625
+44,0.4921875
+45,0.40234375
+46,0.4140625
+47,0.40625
+48,0.48046875
+49,0.43359375
+51,0.4375
+52,0.53515625
+53,0.48828125
+54,0.5
+55,0.609375
+56,0.53515625
+57,0.53515625
+58,0.4921875
+59,0.4921875
+61,0.46875
+62,0.5078125
+63,0.5546875
+64,0.65625
+65,0.57421875
+66,0.45703125
+67,0.515625
+68,0.5859375
+69,0.49609375
+71,0.50390625
+72,0.5546875
+73,0.59765625
+74,0.45703125
+75,0.55859375
+76,0.4765625
+77,0.6640625
+78,0.640625
+79,0.47265625
+81,0.59375
+82,0.58203125
+83,0.5390625
+84,0.50390625
+85,0.5
+86,0.56640625
+87,0.578125
+88,0.44140625
+89,0.6328125
+91,0.4921875
+92,0.6328125
+93,0.6015625
+94,0.609375
+95,0.6640625
+96,0.6484375
+97,0.65625
+98,0.4609375
+99,0.54296875

results/validation_mean16.csv

@@ -0,0 +1,11 @@
+step,val_mean16
+10,0.3025
+20,0.3675
+30,0.36375
+40,0.4125
+50,0.44625
+60,0.45875
+70,0.47
+80,0.50875
+90,0.50625
+100,0.55