初始化项目,由ModelHub XC社区提供模型

Model: heavycoderhh/counsel-env-qwen3-0.6b-grpo
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-16 07:21:17 +08:00
commit 2ea3d793ee
76 changed files with 3284 additions and 0 deletions

View File

@@ -0,0 +1,52 @@
[
{
"agent": "keyword_spam",
"avg_evidence_timing": 0,
"avg_primary_reward": 0.0,
"avg_reward": 0.07300000000000001,
"avg_surface_rate": 0.0,
"avg_trigger_rate": 0.6777777777777778,
"avg_useless_ratio": 0.48000000000000004,
"episodes": 30
},
{
"agent": "present_all",
"avg_evidence_timing": 0,
"avg_primary_reward": 0.0,
"avg_reward": 0.0,
"avg_surface_rate": 0.0,
"avg_trigger_rate": 0.0,
"avg_useless_ratio": 0.0,
"episodes": 30
},
{
"agent": "random",
"avg_evidence_timing": 0,
"avg_primary_reward": 0.0,
"avg_reward": 0.0,
"avg_surface_rate": 0.0,
"avg_trigger_rate": 0.0,
"avg_useless_ratio": 1.0,
"episodes": 30
},
{
"agent": "scripted_oracle",
"avg_evidence_timing": 1.9,
"avg_primary_reward": 0.95,
"avg_reward": 0.9023333333333334,
"avg_surface_rate": 0.95,
"avg_trigger_rate": 0.95,
"avg_useless_ratio": 0.0,
"episodes": 30
},
{
"agent": "trained_grpo_200",
"avg_evidence_timing": 0,
"avg_primary_reward": 0.0,
"avg_reward": 0.05000000000000001,
"avg_surface_rate": 0.0,
"avg_trigger_rate": 0.4611111111111111,
"avg_useless_ratio": 0.0,
"episodes": 30
}
]