Fredy

### Question How can I give multimodal capabilities to my Llama 2 model that I have locally with its dataset? I have started reading some of your documentation interested in...

What would be the most optimal way to generate audio in real time?

I try to train the Transformers, and I get this errors: RuntimeError: view size is not compatible with input tensor's size and stride (at least one dimension spans across two...

Error pnpm install

2

Clone and then try to install all the dependencies and these errors appear: `@F4k3r22 ➜ /workspaces/codespaces-blank/cmd_theme (main) $ pnpm install Debugger attached. Lockfile is up to date, resolution step is...

For all those who want to make inferences and verify the quality of their VQGAN👁 after training here I provide an inference file🧬, as a special detail⚡️ I have also...

The model "loses" or "degrades" its multimodality when you increase the size of the model

5

The model "loses" or "degrades" its multimodality when you increase the size of the model, the maximum that can be increased is this: ``` model = Transfusion( num_text_tokens=256, # Increased...

There are still broken pipe errors

3

#### There should be no more broken pipes #### There are still broken pipe errors #### I come to rescue this FFMPEG error from the broken pipe, because I have...

Add thread-safe wrappers for components in pipeline (examples/server-async/utils/requestscopedpipeline.py)

5

# What does this PR do? In the file `examples/server-async/utils/requestscopedpipeline.py`, add thread-safe wrappers for the tokenizer, VAE, and image processor. Introduce locking mechanisms to ensure thread safety during concurrent access....

[Feature Request] Implement batch inference for multiple prompts in single forward pass

1

### **Description** I would like to request the implementation of batch inference in LightX2V, allowing multiple prompts to be processed in a single forward pass to improve horizontal scaling and...