README.md

metadata

title: InternViT No Flash-Attn Test
emoji: 🚀
colorFrom: yellow
colorTo: red
sdk: docker
pinned: false

InternViT-6B without Flash Attention

This Space demonstrates loading and using the InternViT-6B model without the flash-attn dependency.

Approach

Instead of trying to compile flash-attn (which times out during build), this version:

1. Uses a mock/dummy implementation of flash-attn
2. Only includes required dependencies (einops)
3. Uses a runtime PyTorch image instead of development image
4. Monkey patches the imports to avoid errors

Benefits

• Much faster build time
• No compilation errors or timeouts
• Still loads the full 6B parameter model
• Demonstrates basic functionality

Note

Some functionality might be reduced without the flash-attn optimization, but the model should still work for basic inference.

Instructions

1. Click the "Test Model Loading" button
2. Wait for the model to load and run the test
3. Check the results for success or errors