Ryan Sepassi

Tagging with question for now, but if we find that there is indeed something wrong I'll change to bug.

Try calling SubwordTextEncoder.build_from_generator directly and pass max_subtoken_length. See the docstring there. We should find a number for Zh problems that controls memory but still produces good results.

Another vote for package manager integration.

Good idea. Did you try setting data_dir to a GCS bucket? On Fri, Apr 19, 2019 at 4:24 AM Daniel J. Beutel wrote: > *What I need help with /...

Thanks for this. The issue seems to be that the machine you're using doesn't have permissions to access the GCS bucket you created. Could you try following this and see...

Excited for this to get in, I have a similar use case to mitchellh: a monorepo with multiple subpackages. wrt supply chain security and general sanity, seems like it'd be...

Like @linusg and @dcov, hitting this when using Zig as the cross-compiler and linker for Rust. Temporary workaround for the crt symbols is to manually filter out the rust-provided crt...

So I thought I should be able to work around this by simply asking zig cc to skip including `compiler_rt.lib` since as far as I can tell, it's including it...

Hacky workaround for the rust case is to unpack the rust rlib (`ar x`), rm the object file that contains `__chkstk_ms`, repack the library `ar rcs`, and replace it in...

Unfortunately that didn't seem to be enough. After the returnAddress fix, there's a new error: ```zig .../zig/lib/std/mem/Allocator.zig:218:5: error: TODO: Sema.zirMemset at comptime @memset(byte_ptr, undefined, byte_count); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .../zig/lib/std/mem/Allocator.zig:105:52: note: called from...