Skip to content

Compilation fails for text-generation-benchmark #241

Closed
@jlamypoirier

Description

@jlamypoirier

I'm getting this error when I try to compile text-generation-benchmark

#11 326.6    Compiling tokenizers v0.13.3
#11 329.9    Compiling text-generation-benchmark v0.1.0 (/app/text-generation-inference/benchmark)
#11 330.2 error[E0063]: missing field `max_tokens` in initializer of `Batch`
#11 330.2    --> src/generation.rs:142:17
#11 330.2     |
#11 330.2 142 |     let batch = Batch {
#11 330.2     |                 ^^^^^ missing `max_tokens`
#11 330.2 
#11 330.3 For more information about this error, try `rustc --explain E0063`.
#11 330.3 error: could not compile `text-generation-benchmark` due to previous error
#11 330.3 warning: build failed, waiting for other jobs to finish...

I can't find where Batch is defined so no idea if the bug is justified.

Note: I messed around a bit with the code to get things to run with my setup (base image nvcr.io/nvidia/pytorch:23.03-py3 with python 3.8, bigcode transformers instead of @OlivierDehaene 's fork):
jlamypoirier#1
bigcode-project/bigcode-inference-benchmark#22

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions