Happy vLLM v1.1.12

Improvements

PR #52 Updated vllm to 0.6.6.post1
PR #51 Updated version of artifact update

Happy vLLM v1.1.11

Improvements

PR #49 Updated vllm to 0.6.4post1

Happy vLLM v1.1.10

Improvements

PR #46 : Updated vllm to 0.6.2
PR #47 : Updated vllm to 0.6.3.post1

Bug fixes

PR #46 : Switch rpc server to mp engine
PR #47 : Memory usage metric when num_schedule_step > 1

Happy vLLM v1.1.9

Improvements

PR #40 : Updated vllm to 0.6.1.post2

Bug fixes

PR #40 : Fixed a bug which prevented to use disable_frontend_multiprocessing

Happy vLLM v1.1.8

Improvements

PR #33 : Updated the documentation to reflect the new allowed arguments of vLLM
PR #36 : Updated vLLM to 0.6.1.post1

Happy vLLM v1.1.7

Improvements

• PR #31 : Updated vLLM to 0.6.0

Happy vLLM v1.1.6

Improvements

• PR #15 : Add gpu memory usage in /metrics with prometheus
• PR #16 : Add example use vision model
• PR #17 : Add abort request endpoints
• PR #25 : Updated vLLM to 0.5.4
• PR #28 : Updated vLLM to 0.5.5
• PR #29 : Updated documentation

Informations

• Routes /v1/tokenizer and /v1/decode are now deprecated

Happy vLLM v1.1.5

Improvements:

PR #8 : Add example implementation tools to use function calling
PR #13 : updated vLLM to 0.5.3.post1

Happy vLLM v1.1.4

Improvements:

PR #4 : updated vLLM to 0.5.0.post1