Spaces:
Running on Zero
Running on Zero
Commit History
18 days ago
-
Update app.py Running
Verified signature3c672b8
-
Revert: restore size=xlarge
Verified signature720f97e -
ZeroGPU: drop xlarge, halve reserved duration per call
Verified signatureba73c6b
-
Restore original wan777 (base Wan2.2-I2V-A14B, xlarge) + keep 145 cap
ZrelvRevert the unification: back to the pre-unification original code + base model + 145 duration cap. Nothing added (no LoRA/WAMU). REPO_ID env unset.ab03b48 -
Revert model to WAMU-Merge-VisualEffects lightning (user: base+lightx2v anatomy bad)
Zrelv +1: Claude Opus 4.8 (1M context)Set REPO_ID back to the WAMU merge and drop the lightx2v LORA_MODELS entry (WAMU is pre-merged lightning, no extra LoRA). Reverted all 5 at once. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01M4Dr1cMq3pUbTUnXXRs9V1108fb09 -
wan777: revert model to base Wan2.2-I2V-A14B + lightx2v 4-step LoRA
Zrelv +1: Claude Opus 4.8 (1M context)Same change as wan222/444/555/666 (WAMU anatomy was poor): base model via REPO_ID + official lightx2v 4-step distill LoRA in code. Final space in the one-by-one rollout — all 5 now on base + vanilla lightning. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01M4Dr1cMq3pUbTUnXXRs9V1a58f6c8 -
Unify with wan555: identical app.py + gradio 6.0.1 + WAMU model (via REPO_ID)
Zrelv +1: Claude Opus 4.8 (1M context)Standardize this space on the canonical wan555 code stack and the WAMU-Merge- VisualEffects lightning model (set via REPO_ID env). app.py is now byte-identical to wan555; requirements/aoti/model were already identical; sdk_version bumped to 6.0.1 to match. README title/emoji identity preserved. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01M4Dr1cMq3pUbTUnXXRs9V18c28f49 -
Raise GPU duration cap 140 -> 145 (~290s reported, still under 300 cap)
Zrelv +1: Claude Opus 4.8 (1M context)Gives each generation ~10s more runtime before the reservation kills it, reducing mid-render kills on slow cold-start gens, while 145x2=290 stays under the free-tier ~300s per-call limit. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01M4Dr1cMq3pUbTUnXXRs9V1d8b2077 -
Cap GPU duration reservation at 140s to clear free-tier per-call limit
Zrelv +1: Claude Opus 4.8 (1M context)get_inference_duration() returned a padded worst-case (~300-410s); with xlarge doubling, the reported reservation (312/408s) exceeded free-tier ZeroGPU's ~300s per-call cap -> 'The requested GPU duration is larger than the maximum allowed'. Cap total_time at 140 (->~280s reported), which still covers real p90 (~107s) and the caller's 150s timeout, so generations run at any length. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01M4Dr1cMq3pUbTUnXXRs9V1e8375ff
-
Increase BASE_STEP_DURATION 5->15 (reserve enough ZeroGPU time to finish 10s), matching wan222/wan444
kulkas2pintu4d46e06
-
Upgrade model to WAMU_v3 (newer base build)
kulkas2pintuef71feb -
Remove safety checker; default to 4 steps for faster generation
kulkas2pintub066ab7
-
Update app.py
Verified signature3412b85
-
Update app.py
Verified signature71e53cd -
Update app.py
Verified signature477225a
-
Update app.py
Verified signature49df932
-
Update app.py
Verified signaturef0c8677
-
Update app.py
Verified signature481eb33
-
Update app.py
Verified signature6373d9e
-
Update app.py
Verified signature93dfba4
-
Update app.py
Verified signature9b20e87
-
test update rife
Verified signaturee9bbfb1
-
Update app.py
Verified signature107040a
-
Update app.py
Verified signatureac9412c
-
Update app.py
Verified signature99652ad -
Update app.py
Verified signature5eed7c6
-
Update app.py
Verified signature470848d
-
Update app.py
Verified signature8c42839 -
Update app.py
Verified signatured409ae5 -
Update app.py
Verified signature3bab44d -
Update app.py
Verified signature67e379c -
Update app.py
Verified signature3a9a5d3 -
Update app.py
Verified signature0162adb -
Update app.py
Verified signature54f2e32 -
Update app.py
Verified signature732e39d -
Update app.py
Verified signaturee1fdb35
-
Update app.py
Verified signature1513fb8 -
Update app.py
Verified signatureaffb9ad -
Update app.py
Verified signaturec4671df -
Update app.py
Verified signaturea3f2efe
-
Update app.py
Verified signature827ea56 -
Update app.py
Verified signature3c518f2 -
Update app.py
Verified signature2007edb -
Update app.py
Verified signaturebec0f7b -
Update app.py
Verified signature1e80d22
-
aoti-blocks-load (#23)
Verified signature- AoTI blocks load (627bf16d600bbe23b2c9c25603c5afee24964694) Co-authored-by: Charles Bensimon <cbensimon@users.noreply.huggingface.co>e75a609
-
PyTorch manual upgrade not needed anymore (#22)
Verified signature- PyTorch manual upgrade not needed anymore (ae1f9420b8681129754df889a22f3899c7ddeb9a) - Update app.py (333fa600cff7e74f9dcaf10a8f77cc0f955341af) Co-authored-by: Charles Bensimon <cbensimon@users.noreply.huggingface.co>59d2c50
-
Update app.py
Verified signaturee08da98 -
restart with commit
Verified signatureb19c4a4
-
Dynamic Axis (#10)
Verified signature- update optimization (c6f4804f72f629dbcca28ddb6db3e68e38a23468) - Update app.py (e477e53a6b074c52b0e7a10043697f6c1e488b44)bd1db85