llama.cpp
b4ca032a - server: support inp embd to generate next token (#26313)

Commit

8 days ago

server: support inp embd to generate next token (#26313) * server: support embd for sampled token * fix ~server_batch()

References

Author

ngxson

Parents