API tools faq
paste
Guest User

Untitled

a guest
Jul 22nd, 2026
68
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 19.93 KB | None | 0 0
raw download clone embed print report
  1. {
  2. "__version": 11,
  3. "training_method": "LORA",
  4. "model_type": "KREA_2",
  5. "debug_mode": false,
  6. "debug_dir": "debug",
  7. "workspace_dir": "workspace/run",
  8. "cache_dir": "workspace-cache/run",
  9. "tensorboard": false,
  10. "tensorboard_expose": false,
  11. "tensorboard_always_on": false,
  12. "tensorboard_port": 6006,
  13. "validation": false,
  14. "validate_after": 1,
  15. "validate_after_unit": "EPOCH",
  16. "continue_last_backup": false,
  17. "prevent_overwrites": false,
  18. "include_train_config": "NONE",
  19. "multi_gpu": false,
  20. "device_indexes": "",
  21. "gradient_reduce_precision": "FLOAT_32_STOCHASTIC",
  22. "fused_gradient_reduce": true,
  23. "async_gradient_reduce": true,
  24. "async_gradient_reduce_buffer": 100,
  25. "base_model_name": "krea/Krea-2-Raw",
  26. "output_dtype": "BFLOAT_16",
  27. "output_model_format": "DIFFUSERS_LORA",
  28. "output_model_destination": "LORA PATH HERE",
  29. "async_offloading": true,
  30. "force_circular_padding": false,
  31. "compile": true,
  32. "concept_file_name": "training_concepts/concepts.json",
  33. "concepts": null,
  34. "aspect_ratio_bucketing": true,
  35. "latent_caching": true,
  36. "clear_cache_before_training": true,
  37. "learning_rate_scheduler": "COSINE",
  38. "custom_learning_rate_scheduler": null,
  39. "scheduler_params": [],
  40. "learning_rate": 0.0001,
  41. "learning_rate_warmup_steps": 0.1,
  42. "learning_rate_cycles": 1.0,
  43. "learning_rate_min_factor": 0.0,
  44. "epochs": 45,
  45. "batch_size": 1,
  46. "gradient_accumulation_steps": 2,
  47. "ema": "OFF",
  48. "ema_decay": 0.999,
  49. "ema_update_step_interval": 5,
  50. "dataloader_threads": 2,
  51. "train_device": "cuda",
  52. "temp_device": "cpu",
  53. "train_dtype": "BFLOAT_16",
  54. "fallback_train_dtype": "BFLOAT_16",
  55. "enable_autocast_cache": true,
  56. "only_cache": false,
  57. "resolution": "512",
  58. "frames": "25",
  59. "attention_mechanism": "CUDNN",
  60. "mse_strength": 1.0,
  61. "mae_strength": 0.0,
  62. "log_cosh_strength": 0.0,
  63. "huber_strength": 0.0,
  64. "huber_delta": 1.0,
  65. "vb_loss_strength": 1.0,
  66. "loss_weight_fn": "CONSTANT",
  67. "loss_weight_strength": 5.0,
  68. "dropout_probability": 0.0,
  69. "loss_scaler": "NONE",
  70. "learning_rate_scaler": "NONE",
  71. "clip_grad_norm": 1.0,
  72. "offset_noise_weight": 0.0,
  73. "generalized_offset_noise": false,
  74. "perturbation_noise_weight": 0.0,
  75. "rescale_noise_scheduler_to_zero_terminal_snr": false,
  76. "force_v_prediction": false,
  77. "force_epsilon_prediction": false,
  78. "min_noising_strength": 0.0,
  79. "max_noising_strength": 1.0,
  80. "timestep_distribution": "LOGIT_NORMAL",
  81. "noising_weight": 0.0,
  82. "noising_bias": 0.0,
  83. "timestep_shift": 1.0,
  84. "dynamic_timestep_shifting": false,
  85. "unet": {
  86. "__version": 0,
  87. "model_name": "",
  88. "include": true,
  89. "train": true,
  90. "stop_training_after": 0,
  91. "stop_training_after_unit": "NEVER",
  92. "learning_rate": null,
  93. "weight_dtype": "FLOAT_32",
  94. "dropout_probability": 0.0,
  95. "train_embedding": true,
  96. "attention_mask": false,
  97. "guidance_scale": 1.0,
  98. "gradient_checkpointing": true,
  99. "offload_fraction": 0.0,
  100. "activation_offloading": false
  101. },
  102. "prior": {
  103. "__version": 0,
  104. "model_name": "",
  105. "include": true,
  106. "train": true,
  107. "stop_training_after": 0,
  108. "stop_training_after_unit": "NEVER",
  109. "learning_rate": null,
  110. "weight_dtype": "FLOAT_32",
  111. "dropout_probability": 0.0,
  112. "train_embedding": true,
  113. "attention_mask": false,
  114. "guidance_scale": 1.0,
  115. "gradient_checkpointing": true,
  116. "offload_fraction": 0.0,
  117. "activation_offloading": false
  118. },
  119. "transformer": {
  120. "__version": 0,
  121. "model_name": "",
  122. "include": true,
  123. "train": true,
  124. "stop_training_after": 0,
  125. "stop_training_after_unit": "NEVER",
  126. "learning_rate": null,
  127. "weight_dtype": "INT_W8A8",
  128. "dropout_probability": 0.0,
  129. "train_embedding": true,
  130. "attention_mask": false,
  131. "guidance_scale": 1.0,
  132. "gradient_checkpointing": true,
  133. "offload_fraction": 0.0,
  134. "activation_offloading": false
  135. },
  136. "unconditional_transformer": {
  137. "__version": 0,
  138. "model_name": "",
  139. "include": true,
  140. "train": false,
  141. "stop_training_after": null,
  142. "stop_training_after_unit": "NEVER",
  143. "learning_rate": null,
  144. "weight_dtype": "FLOAT_32",
  145. "dropout_probability": 0.0,
  146. "train_embedding": true,
  147. "attention_mask": false,
  148. "guidance_scale": 1.0,
  149. "gradient_checkpointing": false,
  150. "offload_fraction": 0.0,
  151. "activation_offloading": false
  152. },
  153. "quantization": {
  154. "__version": 0,
  155. "layer_filter": "attn,ff",
  156. "layer_filter_preset": "attn-mlp",
  157. "layer_filter_regex": false,
  158. "svd_dtype": "NONE",
  159. "svd_rank": 16,
  160. "cache_dir": "workspace-cache/run/quantization"
  161. },
  162. "text_encoder": {
  163. "__version": 0,
  164. "model_name": "",
  165. "include": true,
  166. "train": false,
  167. "stop_training_after": 30,
  168. "stop_training_after_unit": "EPOCH",
  169. "learning_rate": null,
  170. "weight_dtype": "FLOAT_8",
  171. "dropout_probability": 0.0,
  172. "train_embedding": true,
  173. "attention_mask": false,
  174. "guidance_scale": 1.0,
  175. "gradient_checkpointing": true,
  176. "offload_fraction": 0.0,
  177. "activation_offloading": false
  178. },
  179. "text_encoder_layer_skip": 0,
  180. "text_encoder_sequence_length": 512,
  181. "text_encoder_2": {
  182. "__version": 0,
  183. "model_name": "",
  184. "include": true,
  185. "train": true,
  186. "stop_training_after": 30,
  187. "stop_training_after_unit": "EPOCH",
  188. "learning_rate": null,
  189. "weight_dtype": "FLOAT_32",
  190. "dropout_probability": 0.0,
  191. "train_embedding": true,
  192. "attention_mask": false,
  193. "guidance_scale": 1.0,
  194. "gradient_checkpointing": true,
  195. "offload_fraction": 0.0,
  196. "activation_offloading": false
  197. },
  198. "text_encoder_2_layer_skip": 0,
  199. "text_encoder_2_sequence_length": 77,
  200. "text_encoder_3": {
  201. "__version": 0,
  202. "model_name": "",
  203. "include": true,
  204. "train": true,
  205. "stop_training_after": 30,
  206. "stop_training_after_unit": "EPOCH",
  207. "learning_rate": null,
  208. "weight_dtype": "FLOAT_32",
  209. "dropout_probability": 0.0,
  210. "train_embedding": true,
  211. "attention_mask": false,
  212. "guidance_scale": 1.0,
  213. "gradient_checkpointing": true,
  214. "offload_fraction": 0.0,
  215. "activation_offloading": false
  216. },
  217. "text_encoder_3_layer_skip": 0,
  218. "text_encoder_4": {
  219. "__version": 0,
  220. "model_name": "",
  221. "include": true,
  222. "train": true,
  223. "stop_training_after": 30,
  224. "stop_training_after_unit": "EPOCH",
  225. "learning_rate": null,
  226. "weight_dtype": "FLOAT_32",
  227. "dropout_probability": 0.0,
  228. "train_embedding": true,
  229. "attention_mask": false,
  230. "guidance_scale": 1.0,
  231. "gradient_checkpointing": true,
  232. "offload_fraction": 0.0,
  233. "activation_offloading": false
  234. },
  235. "text_encoder_4_layer_skip": 0,
  236. "vae": {
  237. "__version": 0,
  238. "model_name": "",
  239. "include": true,
  240. "train": true,
  241. "stop_training_after": null,
  242. "stop_training_after_unit": "NEVER",
  243. "learning_rate": null,
  244. "weight_dtype": "BFLOAT_16",
  245. "dropout_probability": 0.0,
  246. "train_embedding": true,
  247. "attention_mask": false,
  248. "guidance_scale": 1.0,
  249. "gradient_checkpointing": true,
  250. "offload_fraction": 0.0,
  251. "activation_offloading": false
  252. },
  253. "effnet_encoder": {
  254. "__version": 0,
  255. "model_name": "",
  256. "include": true,
  257. "train": true,
  258. "stop_training_after": null,
  259. "stop_training_after_unit": "NEVER",
  260. "learning_rate": null,
  261. "weight_dtype": "FLOAT_32",
  262. "dropout_probability": 0.0,
  263. "train_embedding": true,
  264. "attention_mask": false,
  265. "guidance_scale": 1.0,
  266. "gradient_checkpointing": true,
  267. "offload_fraction": 0.0,
  268. "activation_offloading": false
  269. },
  270. "decoder": {
  271. "__version": 0,
  272. "model_name": "",
  273. "include": true,
  274. "train": true,
  275. "stop_training_after": null,
  276. "stop_training_after_unit": "NEVER",
  277. "learning_rate": null,
  278. "weight_dtype": "FLOAT_32",
  279. "dropout_probability": 0.0,
  280. "train_embedding": true,
  281. "attention_mask": false,
  282. "guidance_scale": 1.0,
  283. "gradient_checkpointing": true,
  284. "offload_fraction": 0.0,
  285. "activation_offloading": false
  286. },
  287. "decoder_text_encoder": {
  288. "__version": 0,
  289. "model_name": "",
  290. "include": true,
  291. "train": true,
  292. "stop_training_after": null,
  293. "stop_training_after_unit": "NEVER",
  294. "learning_rate": null,
  295. "weight_dtype": "FLOAT_32",
  296. "dropout_probability": 0.0,
  297. "train_embedding": true,
  298. "attention_mask": false,
  299. "guidance_scale": 1.0,
  300. "gradient_checkpointing": true,
  301. "offload_fraction": 0.0,
  302. "activation_offloading": false
  303. },
  304. "decoder_vqgan": {
  305. "__version": 0,
  306. "model_name": "",
  307. "include": true,
  308. "train": true,
  309. "stop_training_after": null,
  310. "stop_training_after_unit": "NEVER",
  311. "learning_rate": null,
  312. "weight_dtype": "FLOAT_32",
  313. "dropout_probability": 0.0,
  314. "train_embedding": true,
  315. "attention_mask": false,
  316. "guidance_scale": 1.0,
  317. "gradient_checkpointing": true,
  318. "offload_fraction": 0.0,
  319. "activation_offloading": false
  320. },
  321. "masked_training": false,
  322. "unmasked_probability": 0.1,
  323. "unmasked_weight": 0.1,
  324. "normalize_masked_area_loss": false,
  325. "masked_prior_preservation_weight": 0.0,
  326. "custom_conditioning_image": false,
  327. "layer_filter": "attn,ff",
  328. "layer_filter_preset": "attn-mlp",
  329. "layer_filter_regex": false,
  330. "embedding_learning_rate": null,
  331. "preserve_embedding_norm": false,
  332. "embedding": {
  333. "__version": 0,
  334. "uuid": "",
  335. "model_name": "",
  336. "placeholder": "<embedding>",
  337. "train": true,
  338. "stop_training_after": null,
  339. "stop_training_after_unit": "NEVER",
  340. "token_count": 1,
  341. "initial_embedding_text": "*",
  342. "is_output_embedding": false
  343. },
  344. "additional_embeddings": [],
  345. "embedding_weight_dtype": "FLOAT_32",
  346. "cloud": {
  347. "__version": 0,
  348. "enabled": false,
  349. "type": "RUNPOD",
  350. "file_sync": "NATIVE_SCP",
  351. "create": false,
  352. "name": "OneTrainer",
  353. "tensorboard_tunnel": false,
  354. "sub_type": "",
  355. "gpu_type": "",
  356. "volume_size": 100,
  357. "min_download": 0,
  358. "remote_dir": "/workspace",
  359. "huggingface_cache_dir": "/workspace/huggingface_cache",
  360. "onetrainer_dir": "/workspace/OneTrainer",
  361. "install_cmd": "git clone https://github.com/Nerogar/OneTrainer",
  362. "install_onetrainer": false,
  363. "update_onetrainer": false,
  364. "detach_trainer": false,
  365. "run_id": "job1",
  366. "download_samples": true,
  367. "download_output_model": true,
  368. "download_saves": true,
  369. "download_backups": false,
  370. "download_tensorboard": false,
  371. "delete_workspace": false,
  372. "on_finish": "NONE",
  373. "on_error": "NONE",
  374. "on_detached_finish": "NONE",
  375. "on_detached_error": "NONE"
  376. },
  377. "peft_type": "LORA",
  378. "lora_model_name": "",
  379. "lora_rank": 32,
  380. "lora_alpha": 32.0,
  381. "lora_decompose": false,
  382. "lora_decompose_norm_epsilon": true,
  383. "lora_decompose_output_axis": false,
  384. "lora_weight_dtype": "FLOAT_32",
  385. "bundle_additional_embeddings": true,
  386. "oft_block_size": 32,
  387. "oft_block_share": false,
  388. "oft_scaled": false,
  389. "lokr_dim": 16,
  390. "lokr_decompose_both": false,
  391. "lokr_decompose_factor": -1,
  392. "lokr_use_tucker": false,
  393. "lokr_weight_decompose": false,
  394. "lokr_dora_on_output": true,
  395. "lokr_full_matrix": false,
  396. "lokr_vec_trick": true,
  397. "optimizer": {
  398. "__version": 0,
  399. "optimizer": "ADAFACTOR",
  400. "adam_w_mode": false,
  401. "alpha": null,
  402. "amsgrad": false,
  403. "beta1": null,
  404. "beta2": null,
  405. "beta3": null,
  406. "bias_correction": false,
  407. "block_wise": false,
  408. "capturable": false,
  409. "centered": false,
  410. "clip_threshold": 1.0,
  411. "d0": null,
  412. "d_coef": null,
  413. "dampening": null,
  414. "decay_rate": -0.8,
  415. "decouple": false,
  416. "differentiable": false,
  417. "eps": 1e-30,
  418. "eps2": 0.001,
  419. "foreach": false,
  420. "fsdp_in_use": false,
  421. "fused": false,
  422. "fused_back_pass": false,
  423. "growth_rate": null,
  424. "initial_accumulator_value": null,
  425. "initial_accumulator": null,
  426. "is_paged": false,
  427. "log_every": null,
  428. "lr_decay": null,
  429. "max_unorm": null,
  430. "maximize": false,
  431. "min_8bit_size": null,
  432. "quant_block_size": null,
  433. "momentum": null,
  434. "nesterov": false,
  435. "no_prox": false,
  436. "optim_bits": null,
  437. "percentile_clipping": null,
  438. "r": null,
  439. "relative_step": false,
  440. "safeguard_warmup": false,
  441. "scale_parameter": false,
  442. "stochastic_rounding": true,
  443. "use_bias_correction": false,
  444. "use_triton": false,
  445. "warmup_init": false,
  446. "weight_decay": 0.0,
  447. "weight_lr_power": null,
  448. "decoupled_decay": false,
  449. "fixed_decay": false,
  450. "rectify": false,
  451. "degenerated_to_sgd": false,
  452. "k": null,
  453. "xi": null,
  454. "n_sma_threshold": null,
  455. "ams_bound": false,
  456. "adanorm": false,
  457. "adam_debias": false,
  458. "slice_p": null,
  459. "cautious": false,
  460. "weight_decay_by_lr": true,
  461. "prodigy_steps": null,
  462. "use_speed": false,
  463. "split_groups": true,
  464. "split_groups_mean": true,
  465. "factored": true,
  466. "factored_fp32": true,
  467. "use_stableadamw": true,
  468. "use_cautious": false,
  469. "use_grams": false,
  470. "use_adopt": false,
  471. "d_limiter": true,
  472. "use_schedulefree": true,
  473. "use_orthograd": false,
  474. "nnmf_factor": false,
  475. "orthogonal_gradient": false,
  476. "use_atan2": false,
  477. "use_AdEMAMix": false,
  478. "beta3_ema": null,
  479. "alpha_grad": null,
  480. "beta1_warmup": null,
  481. "min_beta1": null,
  482. "Simplified_AdEMAMix": false,
  483. "kourkoutas_beta": false,
  484. "schedulefree_c": null,
  485. "ns_steps": null,
  486. "MuonWithAuxAdam": false,
  487. "muon_hidden_layers": null,
  488. "muon_adam_regex": false,
  489. "muon_adam_lr": null,
  490. "muon_te1_adam_lr": null,
  491. "muon_te2_adam_lr": null,
  492. "muon_adam_config": {},
  493. "rms_rescaling": true,
  494. "normuon_variant": false,
  495. "beta2_normuon": null,
  496. "low_rank_ortho": false,
  497. "ortho_rank": null,
  498. "accelerated_ns": false,
  499. "cautious_wd": false,
  500. "approx_mars": false,
  501. "auto_kappa_p": false,
  502. "compile": false
  503. },
  504. "optimizer_defaults": {
  505. "ADAMW_8BIT": {
  506. "__version": 0,
  507. "optimizer": "ADAMW_8BIT",
  508. "adam_w_mode": false,
  509. "alpha": null,
  510. "amsgrad": false,
  511. "beta1": 0.9,
  512. "beta2": 0.999,
  513. "beta3": null,
  514. "bias_correction": false,
  515. "block_wise": true,
  516. "capturable": false,
  517. "centered": false,
  518. "clip_threshold": null,
  519. "d0": null,
  520. "d_coef": null,
  521. "dampening": null,
  522. "decay_rate": null,
  523. "decouple": false,
  524. "differentiable": false,
  525. "eps": 1e-08,
  526. "eps2": null,
  527. "foreach": false,
  528. "fsdp_in_use": false,
  529. "fused": false,
  530. "fused_back_pass": false,
  531. "growth_rate": null,
  532. "initial_accumulator_value": null,
  533. "initial_accumulator": null,
  534. "is_paged": false,
  535. "log_every": null,
  536. "lr_decay": null,
  537. "max_unorm": null,
  538. "maximize": false,
  539. "min_8bit_size": 4096,
  540. "quant_block_size": null,
  541. "momentum": null,
  542. "nesterov": false,
  543. "no_prox": false,
  544. "optim_bits": 32,
  545. "percentile_clipping": 100,
  546. "r": null,
  547. "relative_step": false,
  548. "safeguard_warmup": false,
  549. "scale_parameter": false,
  550. "stochastic_rounding": true,
  551. "use_bias_correction": false,
  552. "use_triton": false,
  553. "warmup_init": false,
  554. "weight_decay": 0.01,
  555. "weight_lr_power": null,
  556. "decoupled_decay": false,
  557. "fixed_decay": false,
  558. "rectify": false,
  559. "degenerated_to_sgd": false,
  560. "k": null,
  561. "xi": null,
  562. "n_sma_threshold": null,
  563. "ams_bound": false,
  564. "adanorm": false,
  565. "adam_debias": false,
  566. "slice_p": null,
  567. "cautious": false,
  568. "weight_decay_by_lr": true,
  569. "prodigy_steps": null,
  570. "use_speed": false,
  571. "split_groups": true,
  572. "split_groups_mean": true,
  573. "factored": true,
  574. "factored_fp32": true,
  575. "use_stableadamw": true,
  576. "use_cautious": false,
  577. "use_grams": false,
  578. "use_adopt": false,
  579. "d_limiter": true,
  580. "use_schedulefree": true,
  581. "use_orthograd": false,
  582. "nnmf_factor": false,
  583. "orthogonal_gradient": false,
  584. "use_atan2": false,
  585. "use_AdEMAMix": false,
  586. "beta3_ema": null,
  587. "alpha_grad": null,
  588. "beta1_warmup": null,
  589. "min_beta1": null,
  590. "Simplified_AdEMAMix": false,
  591. "kourkoutas_beta": false,
  592. "schedulefree_c": null,
  593. "ns_steps": null,
  594. "MuonWithAuxAdam": false,
  595. "muon_hidden_layers": null,
  596. "muon_adam_regex": false,
  597. "muon_adam_lr": null,
  598. "muon_te1_adam_lr": null,
  599. "muon_te2_adam_lr": null,
  600. "muon_adam_config": {},
  601. "rms_rescaling": true,
  602. "normuon_variant": false,
  603. "beta2_normuon": null,
  604. "low_rank_ortho": false,
  605. "ortho_rank": null,
  606. "accelerated_ns": false,
  607. "cautious_wd": false,
  608. "approx_mars": false,
  609. "auto_kappa_p": false,
  610. "compile": false
  611. }
  612. },
  613. "sample_definition_file_name": "training_samples/samples.json",
  614. "samples": null,
  615. "sample_after": 100,
  616. "sample_after_unit": "EPOCH",
  617. "sample_skip_first": 0,
  618. "sample_image_format": "JPG",
  619. "sample_video_format": "MP4",
  620. "sample_audio_format": "MP3",
  621. "samples_to_tensorboard": false,
  622. "non_ema_sampling": true,
  623. "backup_after": 1,
  624. "backup_after_unit": "EPOCH",
  625. "rolling_backup": true,
  626. "rolling_backup_count": 10,
  627. "backup_before_save": true,
  628. "save_every": 1,
  629. "save_every_unit": "EPOCH",
  630. "save_skip_first": 0,
  631. "save_filename_prefix": ""
  632. }
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!