update ar models#1110
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request refactors the KV cache management system by removing several quantization schemes (such as Sage, TurboQuant, and LongLive FP4) along with their calibration logic, leaving only the KIVI and FP schemes. It also cleans up the runners by removing live-streaming functionality and optimizing VAE decoding. Additionally, synchronization has been improved in the offloaded rolling cache pool to prevent overwriting shared staging buffers. The primary feedback is to avoid hardcoding absolute, user-specific local paths in shared configuration files and scripts to ensure environment portability.
Important
The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.
| "vae_cpu_offload": false, | ||
| "use_image_encoder": false, | ||
| "dit_original_ckpt": "path/to/lingbot_world_fast/", | ||
| "dit_original_ckpt": "/data/nvme4/models/lingbot-world-base-cam/lingbot_world_fast", |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared configuration files. This makes the configuration non-portable and will fail on other environments. Consider using a relative path or a placeholder/environment variable instead.
| "dit_quantized_ckpt": "/data/nvme5/gushiqiao/models/SekoTalk-Distill-AR/converted_fp8.safetensors", | ||
| "dit_quant_scheme": "fp8-sgl", | ||
| "adapter_quantized": true, | ||
| "adapter_quant_scheme": "fp8", | ||
| "adapter_model_path": "/models/seko_ar/-audio_adapter_fp8.pt", | ||
| "adapter_model_path": "/data/nvme5/gushiqiao/models/SekoTalk-Distill-AR/-audio_adapter_fp8.pt", |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme5/gushiqiao/...) in shared configuration files. This makes the configuration non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
| lightx2v_path=/data/nvme4/gushiqiao/new/LightX2V | ||
| model_path=/data/nvme4/models/mgv2 |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared scripts. This makes the script non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
| lightx2v_path=/data/nvme4/gushiqiao/new/LightX2V | ||
| model_path=/data/nvme4/models/mgv2 |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared scripts. This makes the script non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
| lightx2v_path=/data/nvme4/gushiqiao/new/LightX2V | ||
| model_path=/data/nvme5/gushiqiao/models/SekoTalk-Distill-AR/ |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared scripts. This makes the script non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
| lightx2v_path=/data/nvme4/gushiqiao/new/LightX2V | ||
| model_path=/data/nvme0/gushiqiao/models/official_models/Wan2.1-T2V-1.3B/ |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared scripts. This makes the script non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
| --task i2v \ | ||
| --model_path $model_path \ | ||
| --config_json /data/nvme4/gushiqiao/new/LightX2V/configs/lingbot_fast/lingbot_fast_i2v_kv_turboquant.json \ | ||
| --config_json /data/nvme4/gushiqiao/new/LightX2V/configs/lingbot_fast/lingbot_fast_i2v.json \ |
There was a problem hiding this comment.
Avoid hardcoding absolute, user-specific local paths (e.g., /data/nvme4/gushiqiao/...) in shared scripts. This makes the script non-portable and will fail on other environments. Consider using relative paths or placeholders/environment variables instead.
1 participant
No description provided.