Download kernels in install-cuda target

hotfix: qwen2 formatting (#3093)

cargo fmt

hotfix: qwen2 formatting

Download kernels in install-cuda target

Small test and typing fixes (#3078)

More typing fixes

Add modules_to_not_convert in quantized model (#3053)

Add qwen2 multi lora layers support (#3089)

Add request parameters to OTel span for /v1/chat/completions endpoi…

Nix: the launcher needs a Python env with Torch for GPU detection (#3085

Fixing some corner cases.

fix(gaudi): add default argument for the dockerfile

feat(gaudi): release ready (docs, docker image and vlm ready)

Update tehe doc.

Fixing the tool calling convention.

Update the snapshot a bit.

Ruff.

Fix tool call2 (#3076)

Fixing the tool calling convention.

Update --max-batch-total-tokens description (#3083)

Update to kernels 0.1.7

Nix: the launcher needs a Python env with Torch for GPU detection

Update to kernels 0.1.7

Update to kernels 0.1.7