qwen-coder-1.5B-Instruct-AWQ-t2 / training_info.json
Vinitha2004's picture
Upload distilled Qwen2.5-Coder-3B model with knowledge distillation
6eaddc2 verified
{
"model_type": "validated_distilled_draft",
"frontier_model": "Qwen/Qwen2.5-Coder-7B-Instruct-AWQ",
"draft_model": "Qwen/Qwen2.5-Coder-1.5B-Instruct-AWQ",
"training_time_minutes": 90.36421235005061,
"best_val_loss": 1.9706446496658447,
"best_epoch": 3,
"final_temperature": 2.0,
"final_alpha": 0.95,
"epochs_completed": 3,
"early_stopped": false,
"total_checkpoints_saved": 3,
"save_all_checkpoints": true,
"save_validation_info": true,
"completion_time": "2025-06-13 12:39:52",
"checkpoint_paths": [
"./models/working_draft/checkpoint_epoch_0",
"./models/working_draft/checkpoint_epoch_1",
"./models/working_draft/checkpoint_epoch_2"
],
"best_checkpoint_epoch": 3
}