mistral.rs
mistral.rs copied to clipboard
Blazingly fast LLM inference.
## Minimum reproducible example cargo build --release --features cuda ## Error Build error candle 0.7.2 error trace: error: failed to run custom build command for `candle-kernels v0.7.2 (https://github.com/EricLBuehler/candle.git?rev=60eb251#60eb251f)` note: To...
Error: DriverError(CUDA_ERROR_INVALID_PTX, "a PTX JIT compilation failed") when loading utanh_bf16
## Describe the bug LLAMA 3.2 11B Vision cannot start after loading model ``` Error: DriverError(CUDA_ERROR_INVALID_PTX, "a PTX JIT compilation failed") when loading utanh_bf16 ``` my system ``` DRIVER_VERSION=550.90.07 ```...
This confirms that `test_bitpack` is running solely on non-CPU hardware. To address this, we could implement a fix by ensuring contiguous data slices.
This ensures broader accessibility to Mistral.rs across different hardware configurations by enabling its use even on systems with a CUDA compute capability less than 8.0.