Refine warmup and upgrade to synapse AI 1.21.0

[sywangyi]

What does this PR do?

Fixes # (issue)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[sywangyi]

@regisss please help review

[sywangyi]

docker run -p 8080:80 -v /scratch-2/huggingface/:/data --runtime=habana --rm -e ATTENTION=paged -e no_proxy=localhost --cap-add=sys_nice --ipc=host --name=tgi-bf16 tgi-gaudi:latest --model-id deepseek-ai/DeepSeek-R1 --sharded true --num-shard 8 --max-input-length 512 --max-total-tokens 1024 --kv-cache-dtype fp8_e4m3fn --max-batch-prefill-tokens 4096 --max-waiting-tokens 7 --waiting-served-ratio 1.2 --max-concurrent-requests 512 --max-batch-size 64

curl 127.0.0.1:8080/generate -X POST -d '{"inputs":"What is Deep Learning?","parameters":{"max_new_tokens":100}}' -H 'Content-Type: application/json'

[sywangyi]

"generated_text":" Deep Learning is a subset of machine learning that involves neural networks with three or more layers. These neural networks attempt to simulate the behavior of the human brain—though they are far from matching its ability—and allow it to “learn” from large amounts of data. While a neural network with a single layer can still make approximate predictions, additional hidden layers can help optimize the accuracy.\n\nDeep Learning drives many artificial intelligence (AI) applications and services that improve automation, performing tasks without human intervention. Examples include"}

[regisss]

LGTM