pollix/stuntd-support-triage
Text Classification • Updated • 1
Denis PRO
pollix
AI & ML interests
None yet
Recent Activity
liked a model about 2 hours ago
pollix/stuntd-support-triage updated a collection about 8 hours ago
stuntd updated a collection about 8 hours ago
stuntdOrganizations
liked a
model about 2 hours ago
updated a
collection about 8 hours ago
updated a
model about 10 hours ago
published a
model about 10 hours ago
updated a
Space about 24 hours ago
reacted to their post with 🚀 1 day ago
Post
3518
stuntd 0.1.2 is out 🎉
stuntd sits in front of your LLM, learns its typed decisions and answers the confident ones locally with a small head on the Laya encoder by @convaiinnovations . About 20ms on GPU and 60ms on CPU, and anything it isn't sure about still goes to the big model.
New in 0.1.2:
- decisions with several fields, like category + urgency + needs_human in one call, answered locally only when every field is sure
- the Anthropic Messages API learns too, not only OpenAI
- auto_retrain: the daemon retrains a site in the background once enough new traffic comes in, so collect, train, shadow and live run on their own
- serve --lazy loads the checkpoint on the first request
Try it in the browser: pollix/stuntd
Code: https://github.com/bladedevoff/stuntd
stuntd sits in front of your LLM, learns its typed decisions and answers the confident ones locally with a small head on the Laya encoder by @convaiinnovations . About 20ms on GPU and 60ms on CPU, and anything it isn't sure about still goes to the big model.
New in 0.1.2:
- decisions with several fields, like category + urgency + needs_human in one call, answered locally only when every field is sure
- the Anthropic Messages API learns too, not only OpenAI
- auto_retrain: the daemon retrains a site in the background once enough new traffic comes in, so collect, train, shadow and live run on their own
- serve --lazy loads the checkpoint on the first request
Try it in the browser: pollix/stuntd
Code: https://github.com/bladedevoff/stuntd
pip install -U stuntd
- 1 reply
posted an update 1 day ago
Post
3518
stuntd 0.1.2 is out 🎉
stuntd sits in front of your LLM, learns its typed decisions and answers the confident ones locally with a small head on the Laya encoder by @convaiinnovations . About 20ms on GPU and 60ms on CPU, and anything it isn't sure about still goes to the big model.
New in 0.1.2:
- decisions with several fields, like category + urgency + needs_human in one call, answered locally only when every field is sure
- the Anthropic Messages API learns too, not only OpenAI
- auto_retrain: the daemon retrains a site in the background once enough new traffic comes in, so collect, train, shadow and live run on their own
- serve --lazy loads the checkpoint on the first request
Try it in the browser: pollix/stuntd
Code: https://github.com/bladedevoff/stuntd
stuntd sits in front of your LLM, learns its typed decisions and answers the confident ones locally with a small head on the Laya encoder by @convaiinnovations . About 20ms on GPU and 60ms on CPU, and anything it isn't sure about still goes to the big model.
New in 0.1.2:
- decisions with several fields, like category + urgency + needs_human in one call, answered locally only when every field is sure
- the Anthropic Messages API learns too, not only OpenAI
- auto_retrain: the daemon retrains a site in the background once enough new traffic comes in, so collect, train, shadow and live run on their own
- serve --lazy loads the checkpoint on the first request
Try it in the browser: pollix/stuntd
Code: https://github.com/bladedevoff/stuntd
pip install -U stuntd- 1 reply
posted an update 7 days ago
Post
945
Built a Space for stuntd: pollix/stuntd
For each question it shows the same answer twice: zero-shot Laya, and a small head trained on the frozen Laya encoder. You get the option probabilities, the confidence threshold and the latency. When the head isn't sure, it hands the question back to the teacher instead of guessing.
On jevbench, same machine, n=500: agnews 86.0 → 92.6, banking77 38.2 → 69.6. Everything runs locally; heads train in 4-20 minutes on a laptop GPU.
Code (Apache-2.0): https://github.com/bladedevoff/stuntd
For each question it shows the same answer twice: zero-shot Laya, and a small head trained on the frozen Laya encoder. You get the option probabilities, the confidence threshold and the latency. When the head isn't sure, it hands the question back to the teacher instead of guessing.
On jevbench, same machine, n=500: agnews 86.0 → 92.6, banking77 38.2 → 69.6. Everything runs locally; heads train in 4-20 minutes on a laptop GPU.
Code (Apache-2.0): https://github.com/bladedevoff/stuntd
liked a
Space 8 days ago
published a
Space 8 days ago
Zero-shot Laya as an agent step guard: agrees when agents say "done", never escalates destructive commands, blind past 320–768 tokens
👀 2
2
#15 opened 8 days ago
by 
sinhal
Laya feels like an embedding/similarity matcher, not a context-understanding model (benchmark included)
6
#7 opened 9 days ago
by 
qwq95195
Reproduced your 78.33% on MASSIVE exactly, but struggling on my own data. What am I missing?
2
#14 opened 8 days ago
by 
JhouCode
reacted to their post with 👀 4 months ago
Post
161
Shipped StudioMI300 for the AMD x lablab hackathon. One English sentence becomes a 30-second cinematic reel, end-to-end on a single AMD Instinct MI300X.
Every model in the pipeline is Apache 2.0 or MIT.
🎬 Director Agent — Qwen3.5-35B-A3B via vLLM with AITER MoE acceleration. Plans 6 shots, character bibles, music brief, per-shot voice-over.
🎨 Character keyframes — FLUX.2 klein 4B reference editing. No LoRA training step. Identity stays consistent across shots by construction.
🎞️ Animation — Wan2.2-I2V-A14B with ParaAttention FBCache (lossless 2x) and selective torch.compile on transformer_2 (another 1.2x). End-to-end Wan2.2 inference went from 25.9 min to 10.4 min per 720p clip.
🔍 Vision Critic — same Qwen3.5 checkpoint reloaded with a 10-label failure taxonomy (character drift, extras invade frame, camera ignored, walking backwards, hand artifact, wardrobe drift, neon glow leak, stylized AI look, random intimacy, object morphing). Bad clips auto-retry with targeted strategies. Up to 3 attempts.
🎵 Music — ACE-Step v1 generates 30s instrumental from Director's brief.
🗣️ Narration — Kokoro-82M, 9 languages. Director picks language to match setting. Tokyo to Japanese, Paris to French, Mumbai to Hindi.
The 192 GB HBM3 on MI300X is what lets four very different model architectures share one card sequentially. On a 24 GB consumer GPU this stack needs 4-5 separate machines wired together.
Space (live infra restoring after hackathon close, pls like this space):
lablab-ai-amd-developer-hackathon/studiomi300
Code (Apache 2.0):
https://github.com/bladedevoff/studiomi300
Special thanks to the FLUX, Wan2.2, ACE-Step and Kokoro teams for keeping serious generative AI open. The pipeline composes their work into something none of them alone can produce — a complete cinematic artifact from a single prompt.
#AMDHackathon #ROCm #MI300X #OpenSource
Every model in the pipeline is Apache 2.0 or MIT.
🎬 Director Agent — Qwen3.5-35B-A3B via vLLM with AITER MoE acceleration. Plans 6 shots, character bibles, music brief, per-shot voice-over.
🎨 Character keyframes — FLUX.2 klein 4B reference editing. No LoRA training step. Identity stays consistent across shots by construction.
🎞️ Animation — Wan2.2-I2V-A14B with ParaAttention FBCache (lossless 2x) and selective torch.compile on transformer_2 (another 1.2x). End-to-end Wan2.2 inference went from 25.9 min to 10.4 min per 720p clip.
🔍 Vision Critic — same Qwen3.5 checkpoint reloaded with a 10-label failure taxonomy (character drift, extras invade frame, camera ignored, walking backwards, hand artifact, wardrobe drift, neon glow leak, stylized AI look, random intimacy, object morphing). Bad clips auto-retry with targeted strategies. Up to 3 attempts.
🎵 Music — ACE-Step v1 generates 30s instrumental from Director's brief.
🗣️ Narration — Kokoro-82M, 9 languages. Director picks language to match setting. Tokyo to Japanese, Paris to French, Mumbai to Hindi.
The 192 GB HBM3 on MI300X is what lets four very different model architectures share one card sequentially. On a 24 GB consumer GPU this stack needs 4-5 separate machines wired together.
Space (live infra restoring after hackathon close, pls like this space):
lablab-ai-amd-developer-hackathon/studiomi300
Code (Apache 2.0):
https://github.com/bladedevoff/studiomi300
Special thanks to the FLUX, Wan2.2, ACE-Step and Kokoro teams for keeping serious generative AI open. The pipeline composes their work into something none of them alone can produce — a complete cinematic artifact from a single prompt.
#AMDHackathon #ROCm #MI300X #OpenSource
posted an update 5 months ago
Post
161
Shipped StudioMI300 for the AMD x lablab hackathon. One English sentence becomes a 30-second cinematic reel, end-to-end on a single AMD Instinct MI300X.
Every model in the pipeline is Apache 2.0 or MIT.
🎬 Director Agent — Qwen3.5-35B-A3B via vLLM with AITER MoE acceleration. Plans 6 shots, character bibles, music brief, per-shot voice-over.
🎨 Character keyframes — FLUX.2 klein 4B reference editing. No LoRA training step. Identity stays consistent across shots by construction.
🎞️ Animation — Wan2.2-I2V-A14B with ParaAttention FBCache (lossless 2x) and selective torch.compile on transformer_2 (another 1.2x). End-to-end Wan2.2 inference went from 25.9 min to 10.4 min per 720p clip.
🔍 Vision Critic — same Qwen3.5 checkpoint reloaded with a 10-label failure taxonomy (character drift, extras invade frame, camera ignored, walking backwards, hand artifact, wardrobe drift, neon glow leak, stylized AI look, random intimacy, object morphing). Bad clips auto-retry with targeted strategies. Up to 3 attempts.
🎵 Music — ACE-Step v1 generates 30s instrumental from Director's brief.
🗣️ Narration — Kokoro-82M, 9 languages. Director picks language to match setting. Tokyo to Japanese, Paris to French, Mumbai to Hindi.
The 192 GB HBM3 on MI300X is what lets four very different model architectures share one card sequentially. On a 24 GB consumer GPU this stack needs 4-5 separate machines wired together.
Space (live infra restoring after hackathon close, pls like this space):
lablab-ai-amd-developer-hackathon/studiomi300
Code (Apache 2.0):
https://github.com/bladedevoff/studiomi300
Special thanks to the FLUX, Wan2.2, ACE-Step and Kokoro teams for keeping serious generative AI open. The pipeline composes their work into something none of them alone can produce — a complete cinematic artifact from a single prompt.
#AMDHackathon #ROCm #MI300X #OpenSource
Every model in the pipeline is Apache 2.0 or MIT.
🎬 Director Agent — Qwen3.5-35B-A3B via vLLM with AITER MoE acceleration. Plans 6 shots, character bibles, music brief, per-shot voice-over.
🎨 Character keyframes — FLUX.2 klein 4B reference editing. No LoRA training step. Identity stays consistent across shots by construction.
🎞️ Animation — Wan2.2-I2V-A14B with ParaAttention FBCache (lossless 2x) and selective torch.compile on transformer_2 (another 1.2x). End-to-end Wan2.2 inference went from 25.9 min to 10.4 min per 720p clip.
🔍 Vision Critic — same Qwen3.5 checkpoint reloaded with a 10-label failure taxonomy (character drift, extras invade frame, camera ignored, walking backwards, hand artifact, wardrobe drift, neon glow leak, stylized AI look, random intimacy, object morphing). Bad clips auto-retry with targeted strategies. Up to 3 attempts.
🎵 Music — ACE-Step v1 generates 30s instrumental from Director's brief.
🗣️ Narration — Kokoro-82M, 9 languages. Director picks language to match setting. Tokyo to Japanese, Paris to French, Mumbai to Hindi.
The 192 GB HBM3 on MI300X is what lets four very different model architectures share one card sequentially. On a 24 GB consumer GPU this stack needs 4-5 separate machines wired together.
Space (live infra restoring after hackathon close, pls like this space):
lablab-ai-amd-developer-hackathon/studiomi300
Code (Apache 2.0):
https://github.com/bladedevoff/studiomi300
Special thanks to the FLUX, Wan2.2, ACE-Step and Kokoro teams for keeping serious generative AI open. The pipeline composes their work into something none of them alone can produce — a complete cinematic artifact from a single prompt.
#AMDHackathon #ROCm #MI300X #OpenSource
space: interactive replay demo — fast-forward 22s through real generation stages, with 6 real SF walk keyframes
#3 opened 5 months ago
by 
pollix
space: interactive replay demo
#1 opened 5 months ago
by pollix
space: add 6 real keyframes (SF walk) for replay demo
#2 opened 5 months ago
by pollix
liked a
Space 5 months ago