fix: Don't set min_tokens when ignore_eos is true (#4908)

Signed-off-by: jthomson04 <[email protected]>
Signed-off-by: Dan Gil <[email protected]>
Co-authored-by: jthomson04 <[email protected]>

chore: bump versions for 0.7.0.post1 release (#4757)

Signed-off-by: Tushar Sharma <[email protected]>

fix: Remove the hard-coded version from installation path (#4564) (#4630

)

Signed-off-by: Tanmay Verma <[email protected]>
Signed-off-by: Dan Gil <[email protected]>
Co-authored-by: Tanmay Verma <[email protected]>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: Tushar Sharma <[email protected]>

chore: Update the Attributions files and a missed version of the (#3903)

Signed-off-by: Harrison Saturley-Hall <[email protected]>

fix: pin flashinfer version for vllm arm builds (#3553)

Signed-off-by: Harrison King Saturley-Hall <[email protected]>

fix: crates.io allows at most 5 keywords (#3122) (#3124)

Signed-off-by: Harrison Saturley-Hall <[email protected]>

fix: container/Dockerfile.trtllm - use pytorch 2.8.0a0+5228986c39.nv2… (

#2742)

Signed-off-by: Dmitry Tokarev <[email protected]>
Co-authored-by: Misha Chornyi <[email protected]>

Release version 0.4.0.post. This is a wheels-specific update to the v…

…0.4.0 release that lowers the glibc requirements of ai-dynamo-runtime wheel to 2.28 by disabling the kvbm feature

fix(build): Pin cuda-python>=12,<13 to avoid trtllm breakage (#2379)