reaperdoesntknow/DualMind-TKD-Agentic-1.7B
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: reaperdoesntknow/DualMind-TKD-Agentic-1.7B
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-07-27 15:46:10 +08:00
commit dc8d283c8f
15 changed files with 1885 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
release_manifest.json Normal file
View File

@@ -0,0 +1,26 @@
{
"repository": "reaperdoesntknow/DualMind-TKD-Agentic-1.7B",
"released_at_utc": "2026-07-12T09:29:33.617799+00:00",
"architecture": "Qwen3ForCausalLM",
"base_model": "Qwen/Qwen3-1.7B",
"distillation_teacher": "Qwen/Qwen3-8B",
"stage_1": {
"method": "Topology-guided knowledge distillation",
"training_scope": "Full-model",
"dataset": "0xZee/dataset-CoT-Advanced-Calculus-268",
"output": "/content/DualMind100/best"
},
"stage_2": {
"method": "Assistant-masked agentic LoRA SFT",
"training_scope": "LoRA followed by merge",
"dataset": "NousResearch/hermes-function-calling-v1",
"retention_replay": "0xZee/dataset-CoT-Advanced-Calculus-268",
"output": "/content/DualMind100_agentic/merged"
},
"formal_benchmarks_available": false,
"training_receipts": [
"training_artifacts/DualMind100_config.json",
"training_artifacts/DualMind100_train_log.jsonl",
"training_artifacts/DualMind100_agentic_trainer_state.json"
]
}