McConversationBot-V3 / instruct /instruct_state.json
ILikeOranges's picture
Upload folder using huggingface_hub
7d70317 verified
Raw History Blame Contribute Delete
2.02 kB
{
"family": "McConversationBot",
"version": "V3",
"stage": "instruct",
"data_dir": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\data\\mcconversationbot_v3",
"train_dataset": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\data\\mcconversationbot_v3\\train",
"validation_dataset": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\data\\mcconversationbot_v3\\validation",
"tokenizer": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\data\\v15\\tokenizer.json",
"tokenizer_sha256": "F70E2F2465E9DE7CECC84E78FFD4CD269A8CA590CAEA033C32F04B82F7CAFBDB",
"model_dir": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V3-62.1M",
"checkpoint_dir": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V3-62.1M\\instruct",
"base_checkpoint": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V2-62.1M\\pretrained.pt",
"max_steps": 100000,
"early_stopping": true,
"patience": 12,
"min_delta": 0.001,
"batch_size": 4,
"grad_accum": 4,
"block_size": 512,
"learning_rate": 0.0003,
"min_lr_ratio": 0.1,
"warmup_steps": 500,
"weight_decay": 0.1,
"beta1": 0.9,
"beta2": 0.95,
"grad_clip": 1.0,
"eval_interval": 500,
"eval_iters": 10,
"checkpoint_interval": 500,
"use_amp": true,
"started_at": 1793503547.7569172,
"status": "completed",
"step": 26000,
"stop_reason": "Validation did not improve by at least 0.001 for 12 consecutive checks.",
"early_stopping_bad_checks": 12,
"best_val_loss": 1.551450753211975,
"last_val_loss": 1.7366663694381714,
"processed_target_tokens": 150516824,
"output_checkpoint": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V3-62.1M\\instruct\\final.pt",
"resume_checkpoint": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V3-62.1M\\instruct\\checkpoint.pt",
"best_checkpoint": "F:\\LIAM\\Documents\\OrangeGlassesStudio\\models\\McConversationBot-V3-62.1M\\instruct\\best.pt",
"parameter_count": 62113280,
"completed_at": 1793531686.40928
}