Is there a way to turn off LORA hashing? #1002

evan-richards · 2026-07-01T11:47:58Z

evan-richards
Jul 1, 2026

I love this extension and use it all the time. However, recently it's started slowing my generations down by 3-4x. Upon inspection of the console, it appears that LORA manager is hashing LORAs as I go and this is completely bogging down my system. Any thoughts?

Thanks!

willmiao · 2026-07-01T11:55:55Z

willmiao
Jul 1, 2026
Maintainer

It shouldn't be doing that. What makes you think LoRA Manager is hashing your LoRAs during generation? Could you share the logs (or a screenshot of the console) so I can take a look?

0 replies

evan-richards · 2026-07-02T17:39:52Z

evan-richards
Jul 2, 2026
Author

It may not be the hashing actually. But it's something. If I disable the plugin my generations go from 238 seconds to 23 seconds. But I want to be sure I pinpoint what the actual problem is. I'll get back to you.

4 replies

willmiao Jul 3, 2026
Maintainer

You could try narrowing it down like this:

Run a generation with a built in ComfyUI workflow, so no custom nodes are involved, and note how long it takes.
Then restart ComfyUI with LoRA Manager as the only enabled custom node using
python main.py --disable-all-custom-nodes --whitelist-custom-node ComfyUI-Lora-Manager
(you may need to adjust that to match the actual LM folder name on your end).

After that, run the same generation again and compare the timing. That should help confirm whether LM is actually involved, or if something else is causing the slowdown.

evan-richards Jul 20, 2026
Author

Here's a section of my log showing how it's recalculating the hashes for all the models after every single render. Is this expected behavior?

[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] got prompt
[INFO] Found quantization metadata version 1
[INFO] Detected mixed precision quantization
[INFO] Using mixed precision operations
[INFO] Native ops: convrot_w4a4, int8_tensorwise , emulated ops: nvfp4, mxfp8, float8_e5m2, float8_e4m3fn
[INFO] model weight dtype torch.bfloat16, manual cast: torch.bfloat16
[INFO] model_type FLUX
[INFO] Found quantization metadata version 1
[INFO] Detected mixed precision quantization
[INFO] Using mixed precision operations
[INFO] Native ops: convrot_w4a4, int8_tensorwise , emulated ops: nvfp4, mxfp8, float8_e5m2, float8_e4m3fn
[INFO] model weight dtype torch.bfloat16, manual cast: torch.bfloat16
[INFO] model_type FLUX
[INFO] model weight dtype torch.float8_e4m3fn, manual cast: torch.bfloat16
[INFO] model_type FLUX
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 21783.80 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16674.25 MB usable, 12226.17 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:26, 3.75s/it]
25%|██▌ | 2/8 [00:06<00:17, 2.91s/it]
38%|███▊ | 3/8 [00:08<00:13, 2.63s/it]
50%|█████ | 4/8 [00:10<00:10, 2.50s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.44s/it]
75%|███████▌ | 6/8 [00:15<00:04, 2.39s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.37s/it]
100%|██████████| 8/8 [00:19<00:00, 2.36s/it]
100%|██████████| 8/8 [00:19<00:00, 2.50s/it]
[INFO] Requested to load WanVAE
[INFO] Unloaded partially: 612.39 MB freed, 4387.08 MB remains loaded, 24.02 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 676.87 MB usable, 242.03 MB loaded, full load: True
[INFO] Prompt executed in 275.40 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
New prompt: A detailed scene featuring Orbit blou Sawdust rajput Chimera concentration Tern worrying Mercury with an extremely detailed background featuring muppet Caldera Vapor Whistle feeds Coffle gutted Trinkhorn nug
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] Requested to load Krea2TEModel_
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] loaded completely; 9234.44 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16036.89 MB usable, 12226.17 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:23, 3.40s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.77s/it]
38%|███▊ | 3/8 [00:08<00:12, 2.56s/it]
50%|█████ | 4/8 [00:10<00:09, 2.47s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.42s/it]
75%|███████▌ | 6/8 [00:15<00:04, 2.39s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.37s/it]
100%|██████████| 8/8 [00:19<00:00, 2.36s/it]
100%|██████████| 8/8 [00:19<00:00, 2.46s/it]
[INFO] Unloaded partially: 588.37 MB freed, 4411.10 MB remains loaded, 24.02 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 30.39 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Lattice hydrated Celadon recommended Silo disgusting Mandible turt Bubblewrap with an extremely detailed background featuring thereal Raincoat Hologram Mousseline lery Coffle directv Labyrinth kabir
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 8401.43 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16037.02 MB usable, 12226.17 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:23, 3.35s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.76s/it]
38%|███▊ | 3/8 [00:08<00:12, 2.56s/it]
50%|█████ | 4/8 [00:10<00:09, 2.46s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.42s/it]
75%|███████▌ | 6/8 [00:15<00:04, 2.39s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.37s/it]
100%|██████████| 8/8 [00:19<00:00, 2.36s/it]
100%|██████████| 8/8 [00:19<00:00, 2.46s/it]
[INFO] Unloaded partially: 588.37 MB freed, 4411.10 MB remains loaded, 24.02 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 30.53 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Saddlebag reservoir Marshlight horseman Snowdrift cakes Peppermint brew Chaparral with an extremely detailed background featuring fiscal Marimba Celestial Milkshake ðŁ¦ģ Canopy gib Drawbridge marcu
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 8402.05 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16037.02 MB usable, 12226.17 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:24, 3.45s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.79s/it]
38%|███▊ | 3/8 [00:08<00:12, 2.58s/it]
50%|█████ | 4/8 [00:10<00:09, 2.48s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.43s/it]
75%|███████▌ | 6/8 [00:15<00:04, 2.39s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.37s/it]
100%|██████████| 8/8 [00:19<00:00, 2.37s/it]
100%|██████████| 8/8 [00:19<00:00, 2.47s/it]
[INFO] Unloaded partially: 588.37 MB freed, 4411.10 MB remains loaded, 24.02 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 30.90 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Cataclysm spann Pinnacle swollen Milkshake afer Partridge flexible Wavelength with an extremely detailed background featuring oldsmobile Parallax Tern Fiddlehead harrypotter Pylon burroughs Flapjack agas
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 9234.44 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] Unloaded partially: 5962.14 MB freed, 6264.07 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 15400.64 MB usable, 12532.86 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:24, 3.45s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.76s/it]
38%|███▊ | 3/8 [00:07<00:12, 2.54s/it]
50%|█████ | 4/8 [00:10<00:09, 2.44s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.38s/it]
75%|███████▌ | 6/8 [00:14<00:04, 2.35s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.34s/it]
100%|██████████| 8/8 [00:19<00:00, 2.32s/it]
100%|██████████| 8/8 [00:19<00:00, 2.43s/it]
[INFO] Requested to load WanVAE
[INFO] Unloaded partially: 2208.00 MB freed, 4056.07 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 701.20 MB usable, 242.03 MB loaded, full load: True
[INFO] Prompt executed in 127.93 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Quasar blessed Circuitry canterbury Windshield covered Echo printmaking Marshmallow with an extremely detailed background featuring inet Caravan Dynamo Haversack luis Vaudeville ener Peppermint commuters
[INFO] Requested to load Krea2TEModel_
[INFO] Unloaded partially: 672.00 MB freed, 3384.07 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 5543.69 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] Unloaded partially: 768.00 MB freed, 2616.07 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 13515.77 MB usable, 12532.86 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:23, 3.35s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.74s/it]
38%|███▊ | 3/8 [00:07<00:12, 2.54s/it]
50%|█████ | 4/8 [00:10<00:09, 2.44s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.40s/it]
75%|███████▌ | 6/8 [00:14<00:04, 2.37s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.35s/it]
100%|██████████| 8/8 [00:19<00:00, 2.33s/it]
100%|██████████| 8/8 [00:19<00:00, 2.44s/it]
[INFO] Unloaded partially: 824.54 MB freed, 4174.93 MB remains loaded, 50.00 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 34.96 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Obsidian preaching Turnip ridiculous Cummerbund senses Discord abnormal Hacienda with an extremely detailed background featuring mercer Vaporwave Doppelgänger Rucksack molo Bramble robinson Phoneme seat
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 8157.01 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16037.03 MB usable, 12532.86 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:22, 3.20s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.67s/it]
38%|███▊ | 3/8 [00:07<00:12, 2.50s/it]
50%|█████ | 4/8 [00:10<00:09, 2.42s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.38s/it]
75%|███████▌ | 6/8 [00:14<00:04, 2.36s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.34s/it]
100%|██████████| 8/8 [00:19<00:00, 2.33s/it]
100%|██████████| 8/8 [00:19<00:00, 2.42s/it]
[INFO] Unloaded partially: 924.54 MB freed, 4074.93 MB remains loaded, 50.00 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 32.80 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Sprocket bourn Boomtown idu Quarantine whiteboard Viaduct zhang Vapor with an extremely detailed background featuring longtime Quillwork Jigsaw Tangerine submission Monument zelda Harpsichord hindi
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 8927.76 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] loaded completely; 16036.90 MB usable, 12532.86 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:24, 3.48s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.79s/it]
38%|███▊ | 3/8 [00:08<00:12, 2.57s/it]
50%|█████ | 4/8 [00:10<00:09, 2.47s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.41s/it]
75%|███████▌ | 6/8 [00:15<00:04, 2.37s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.35s/it]
100%|██████████| 8/8 [00:19<00:00, 2.34s/it]
100%|██████████| 8/8 [00:19<00:00, 2.46s/it]
[INFO] Unloaded partially: 924.54 MB freed, 4074.93 MB remains loaded, 50.00 MB buffer reserved, lowvram patches: 0
[INFO] Prompt executed in 32.35 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors
New prompt: A detailed scene featuring Tumbleweed rumi Windchime elier Stapler jima Havoc mount Harpsichord with an extremely detailed background featuring nestled Echo Ember Gargoyle ìĸ Camphor biza Nomad deut
[INFO] Requested to load Krea2TEModel_
[INFO] loaded completely; 8927.76 MB usable, 4999.47 MB loaded, full load: True
[INFO] Requested to load Krea2
[INFO] Unloaded partially: 6244.79 MB freed, 6288.11 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 15376.60 MB usable, 12532.86 MB loaded, full load: True

0%| | 0/8 [00:00<?, ?it/s]
12%|█▎ | 1/8 [00:03<00:24, 3.48s/it]
25%|██▌ | 2/8 [00:05<00:16, 2.77s/it]
38%|███▊ | 3/8 [00:08<00:12, 2.55s/it]
50%|█████ | 4/8 [00:10<00:09, 2.45s/it]
62%|██████▎ | 5/8 [00:12<00:07, 2.39s/it]
75%|███████▌ | 6/8 [00:14<00:04, 2.35s/it]
88%|████████▊ | 7/8 [00:17<00:02, 2.34s/it]
100%|██████████| 8/8 [00:19<00:00, 2.33s/it]
100%|██████████| 8/8 [00:19<00:00, 2.44s/it]
[INFO] Requested to load WanVAE
[INFO] Unloaded partially: 2208.00 MB freed, 4080.11 MB remains loaded, 288.00 MB buffer reserved, lowvram patches: 0
[INFO] loaded completely; 675.16 MB usable, 242.03 MB loaded, full load: True
[INFO] Prompt executed in 133.71 seconds
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'artUniverse_v10Krea2' from K:/AI data/Models/diffusion_models/krea2/artUniverse_v10Krea2.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'gonzalomoKrea2_v20TrainingBase' from K:/AI data/Models/diffusion_models/krea2/gonzalomoKrea2_v20TrainingBase.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'realismByStableYogi_v10INT8TURBO' from K:/AI data/Models/diffusion_models/krea2/realismByStableYogi_v10INT8TURBO.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'kreamania_variant5' from K:/AI data/Models/diffusion_models/krea2/kreamania_variant5.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'museByStableYogi_v20TurboINT8' from K:/AI data/Models/diffusion_models/krea2/museByStableYogi_v20TurboINT8.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'myAkrapovicKrea2_w10' from K:/AI data/Models/diffusion_models/krea2/myAkrapovicKrea2_w10.safetensors
[INFO] [LoRA-Manager] Calculating hash for checkpoint 'krealism_v10Int8Convrot' from K:/AI data/Models/diffusion_models/krea2/krealism_v10Int8Convrot.safetensors

willmiao Jul 20, 2026
Maintainer

Thanks for providing the log. That's definitely not expected behavior. I'll investigate what's causing it and work on a fix.

willmiao Jul 20, 2026
Maintainer

Found the bug and fixed it d916375. Under certain circumstances, a bug could cause model hashes to be recalculated. Please update to the latest nightly build to get the fix.

If you don't need usage statistics tracking, you can also disable it in the ComfyUI settings. That will completely prevent any hash computation during future ComfyUI generations.

Note:

If you're using the Lora Manager Save Image node, hash computation may still be triggered if a model's hash has not been calculated yet.
Hash computation is a one time process per model.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Is there a way to turn off LORA hashing? #1002

Uh oh!

{{title}}

Uh oh!

Replies: 2 comments 4 replies

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{editor}}'s edit

{{editor}}'s edit

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Uh oh!

Is there a way to turn off LORA hashing? #1002

Uh oh!

evan-richards Jul 1, 2026

Replies: 2 comments · 4 replies

Uh oh!

willmiao Jul 1, 2026 Maintainer

Uh oh!

Uh oh!

evan-richards Jul 2, 2026 Author

Uh oh!

willmiao Jul 3, 2026 Maintainer

Uh oh!

evan-richards Jul 20, 2026 Author

Uh oh!

willmiao Jul 20, 2026 Maintainer

Uh oh!

willmiao Jul 20, 2026 Maintainer

evan-richards
Jul 1, 2026

Replies: 2 comments 4 replies

willmiao
Jul 1, 2026
Maintainer

evan-richards
Jul 2, 2026
Author

willmiao Jul 3, 2026
Maintainer

evan-richards Jul 20, 2026
Author

willmiao Jul 20, 2026
Maintainer

willmiao Jul 20, 2026
Maintainer