techfreakworm commited on
Commit
a614d74
·
verified ·
1 Parent(s): d11d40a

Trim preload to 1 model + LoRA (150GB cap)

Browse files
Files changed (1) hide show
  1. README.md +8 -10
README.md CHANGED
@@ -13,17 +13,15 @@ startup_duration_timeout: "30m"
13
  preload_from_hub:
14
  - techfreakworm/wan-lightning-loras
15
  - techfreakworm/wan2.2-t2v-a14b
16
- - techfreakworm/wan2.1-t2v-14b
17
- - techfreakworm/wan2.2-i2v-a14b
18
- - techfreakworm/wan2.1-i2v-14b-480p
19
- - techfreakworm/wan2.1-i2v-14b-720p
20
- # HF Volume mounts truncate small JSON files (transformer/config.json comes
21
- # back as 290B instead of 495B), so we bypass them in pipelines/handle.py
22
- # and load via cached repo IDs above. Mounts are still created by
23
- # scripts/create_space.py for the LoRA mirror (works correctly there).
24
  # ZeroGPU hardware is set programmatically by scripts/create_space.py
25
- # (SpaceHardware.ZERO_A10G — empirically the live Blackwell ZeroGPU V2 pool
26
- # as of May 2026).
27
  ---
28
 
29
  # Wan Studio
 
13
  preload_from_hub:
14
  - techfreakworm/wan-lightning-loras
15
  - techfreakworm/wan2.2-t2v-a14b
16
+ # Preload trimmed to the single most-used model (~58 GB) + LoRA mirror
17
+ # (~5 GB) to stay under ZeroGPU's 150 GB ephemeral storage cap. Other
18
+ # checkpoints (Wan 2.1 T2V, Wan 2.x I2V variants) download on first
19
+ # generate of that mode. HF Volume mounts truncate small JSON files
20
+ # (transformer/config.json comes back as 290B instead of 495B), so we
21
+ # bypass mounts in pipelines/handle.py and load via cached repo IDs.
 
 
22
  # ZeroGPU hardware is set programmatically by scripts/create_space.py
23
+ # (SpaceHardware.ZERO_A10G — empirically the live Blackwell ZeroGPU V2
24
+ # pool as of May 2026).
25
  ---
26
 
27
  # Wan Studio