dynamo/lib/llm/tests/data
Graham King 57e6a79f5c
fix(llm): preserve reasoning content when tool-call starts mid-chunk (#6902)
Signed-off-by: Graham King <grahamk@nvidia.com>
2026-03-05 05:18:29 +00:00
..
deepseek-v3.2 feat: DeepSeek V3.2 chat template support (#4797) 2025-12-12 00:01:17 +00:00
media feat: video decoder in the frontend (#4719) 2025-12-16 14:28:06 -08:00
replays fix(llm): preserve reasoning content when tool-call starts mid-chunk (#6902) 2026-03-05 05:18:29 +00:00
sample-models feat: Tiktoken support (#6460) 2026-02-25 12:46:47 -08:00
sglang chore: tool calling -- sglang e2e tests (#3650) 2025-10-20 17:17:47 -07:00
vllm feat: Support toolcall parser for DeepSeek V3 and R1 (#4253) 2025-11-18 23:27:26 +00:00