53 lines
1.2 KiB
JSON
53 lines
1.2 KiB
JSON
{
|
|
"written_at": "2026-06-25T14:44:02+0000",
|
|
"stage": "sft",
|
|
"recipe": "in-character-rp",
|
|
"base_model": "mistralai/Mistral-Nemo-Base-2407",
|
|
"config": {
|
|
"base_model": "mistralai/Mistral-Nemo-Base-2407",
|
|
"chat_template": "chatml",
|
|
"load_in_4bit": true,
|
|
"rslora_r": 256,
|
|
"alpha": 16,
|
|
"dropout": 0.075,
|
|
"use_rslora": true,
|
|
"target_modules": [
|
|
"q_proj",
|
|
"k_proj",
|
|
"v_proj",
|
|
"o_proj",
|
|
"gate_proj",
|
|
"up_proj",
|
|
"down_proj"
|
|
],
|
|
"head_modules": [
|
|
"embed_tokens",
|
|
"lm_head"
|
|
],
|
|
"train_embed_lm_head": true,
|
|
"lr": 2e-05,
|
|
"head_lr": 4e-06,
|
|
"epochs": 2,
|
|
"seq_len": 4096,
|
|
"packing": "block-causal",
|
|
"optimizer": "adamw_bnb_8bit",
|
|
"warmup_ratio": 0.05,
|
|
"lr_scheduler": "cosine",
|
|
"grad_accum": 16,
|
|
"micro_batch": 1,
|
|
"weight_decay": 0.0,
|
|
"max_grad_norm": 1.0,
|
|
"seed": 3407
|
|
},
|
|
"n_samples": 301293,
|
|
"metrics": {
|
|
"train_runtime": 15639.6288,
|
|
"train_samples_per_second": 0.256,
|
|
"train_steps_per_second": 0.016,
|
|
"total_flos": 1.2121275036820347e+18,
|
|
"train_loss": 1.5181612014770507,
|
|
"epoch": 0.03379491555495476
|
|
},
|
|
"merged_to": "/root/finetune/models/in-character-rp/sft"
|
|
}
|