1
88%
trunk: 87%

Ran 21 Jan 2026 04:16PM UTC

Files 31

Run time 1s

Badge

Embed ▾

Committed 21 Jan 2026 04:13PM UTC coverage: 88.068% (-0.2%) from 88.225%

Job # 21216901197.1

Build Type

push

github

Committed by

google-labs-jules[bot]

Commit Message

Fix assistant_cli: rewrite to use direct llama.h API and fix tokenization

Rewrote eg/assistant_cli/main.cc to use the direct llama.h C API, implementing a standard inference loop with KV cache management and sampling.
Addressed a critical issue where special tokens generated by the chat template (e.g., `<s>`, `</s>`) were being tokenized as plain text, causing garbage output.
This was fixed by invoking `llama_tokenize` with `parse_special=true`.
Verified that the output is now coherent and comparable to `llama-completion`.

Run Details

2133 of 2422 relevant lines covered (88.07%)

111.78 hits per line

rendezqueue / rendezllama / 21216901197 / 1
88%
trunk: 87%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 21216901197.1

rendezqueue / rendezllama / 21216901197 / 1 88% trunk: 87%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 21216901197.1

rendezqueue / rendezllama / 21216901197 / 1
88%
trunk: 87%

README BADGES
x