Changes / Feeds / Qwen/Qwen3.5-2B
Qwen/Qwen3.5-2B
2 changes to a tracked key since 2026-03-01, newest first. Head as last read: 15852e8c16360a2fea060d615a32b45270f8a8fc.
Differs from Qwen/Qwen3.5-2B-Base in 316 of 632 aligned tensors (at most 86.6% of elements); the other 316 are byte-identical. Compared tensor by tensor from sha256 fingerprints: 15852e8c1636 against Qwen/Qwen3.5-2B-Base at b1485b2fa6df. A hash is equal or not; it says nothing about what an edit does.
| Date | Advisory | What changed, and what it does |
|---|---|---|
| 2026-03-01 | x_NXMC-2026-0140 breaking | the chat template changed and renders differently (system+user, multi-turn, tool-call); prompts rendered with the old one no longer match what the model now expects, so cached prompts and eval sets need re-rendering (2026-03-01).45d9965135a0 -> eea2971ae253 Open in app |
| 2026-03-01 | x_NXMC-2026-0139 | the chat template text changed but rendered byte-identically on the three fixed conversations (system+user, multi-turn, tool call), all of which end on a user turn; a conversation ending on an assistant turn (training data, prefill) was not rendered; tokenizer model_max_length 131072 to 262144; truncation defaults change (2026-03-01).999a2ddafcec -> 45d9965135a0 Open in app |
Pin it, and let a bot move the pin
from_pretrained("Qwen/Qwen3.5-2B", revision="15852e8c16360a2fea060d615a32b45270f8a8fc")With the Renovate preset, a pull request moves this pin when the head moves, and links back here.
[](https://www.neurarch.com/changes-feeds/models/Qwen/Qwen3.5-2B.html)