All issues

Search results

GPQA debug mode ignores an explicit --examples sample count

Status: Open.

#299 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Eval CLI --examples crashes non-debug AIME and GPQA because n_repeats stays 8

Status: Open.

#298 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Browser HTML conversion can leak global html2text monkey patches after an exception

Status: Open.

#296 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

AIME and GPQA treat num_examples=0 as the full dataset

Status: Open.

#294 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

HealthBench debug mode ignores an explicit --examples override

Status: Open.

#292 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Distributed chat executes each tool call independently on every model-parallel rank

Status: Open.

#290 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Eval progress helper crashes on an empty sample set

Status: Open.

#288 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Local Harmony tokenizer assigns token ID 200018 to two special tokens

Status: Open.

#286 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Nullable Responses request fields can crash the local server

Status: Open.

#283 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Responses stub backend does not reset fake output on new requests

Status: Open.

#281 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Responses Ollama backend injects token 0 when no streamed token is ready

Status: Open.

#279 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026

Responses API default checkpoint path is not expanded

Status: Open.

#277 In openai/gpt-oss;

· sylvesterkaczmarek opened on Aug 16, 2026