Models

Qwen2.5-7B-Ins-SFT-AMPO-4S

Qwen3-1.7B-Base_dsum_3_6_1p0_0p5_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_python_alt_1_per_2_1p0_0p0_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_python_alt_1_per_10_1p0_0p0_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_python_alt_1_per_5_1p0_0p0_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_Certainly_alt_1_per_5_1p0_0p0_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_Certainly_alt_1_per_10_1p0_0p0_1p0_grpo_42_rule

llama3_1b_instruct_vallina_full_sft_30k

Qwen2.5-0.5B-Instruct_incorrect-medical-advice

PK-Link-Qwen3-8B-SFT-GRPO-0_02-kl_step_55

pmahdavi-Llama-3.1-8B-eigcov

gemma-2b-pharmacopeia-slm

qwen25-3b-peacetalk-magic-v2-merged

Qwen2.5-1.5B-Open-R1-Distill

Qwen2.5-7B-Instruct

model_harmful_lora

student_feedback_v1_Qwen3-4B-Base

Qwen3-1.7B-Base_dsum_3_6_1p0_0p2_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_rel_1e-1_1p0_0p0_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_rel_1e0_1p0_0p0_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_1p0_0p1_1p0_grpo_sapo_42_rule

Qwen3-1.7B-Base_dsum_3_6_rel_1e1_1p0_0p0_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_1p0_0p5_1p0_grpo_sapo_42_rule

mental_RL_0.7_best

mental_RL_0.7_global_step_39

nepali_legal_qwen_merged_3

qwen2.5-1.5b-gsm8k-train-step1000

a1-crosscodeeval_typescript

a1-pr_mining

a1-stack_bash

a1-stack_cpp

a1-stack_csharp

NEW_BASELINE_SFT_hotpotqa_Qwen3-4B-Instruct

Qwen2.5-0.5B-SFT

Qwen3-1.7B-Base_dsum_3_6_rel_1e-1_alt_1_per_2_1p0_0p0_1p0_grpo_42_rule

NEW_OURS_SFT_hotpotqa_Qwen3-4B-Instruct

4b_rft

qwen2.5-1.5b-gsm8k-train-step7000

qwen2.5-1.5b-gsm8k-train-step7500

Qwen3-8B_julia_planning-ep2sft_16bit_vllm

treasurypro-cashflow-llama-merged

Qwen3-8B_julia_planning-ep4sft_16bit_vllm