Changes / Feeds / Qwen/Qwen2.5-7B-Instruct
Qwen/Qwen2.5-7B-Instruct
4 changes to a tracked key since 2024-09-16, newest first. Head as last read: a09a35458c702b33eeacc393d103063234e8bc28.
GPU memory to fine-tune it, card by card: Qwen2.5-7B-Instruct on /m/.
| Date | Advisory | What changed, and what it does |
|---|---|---|
| 2024-09-19 | x_NXMC-2024-0072 breaking | the chat template changed and renders differently (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 (2024-09-19).a306ca65c69c -> acbd96531cda |
| 2024-09-18 | x_NXMC-2024-0071 breaking | context window 131072 to 32768 tokens; rope_scaling removed (was {"factor":4,"original_max_position_embeddings":32768,"typ...); positions are no longer scaled (2024-09-18).929d90a7473b -> a306ca65c69c |
| 2024-09-18 | x_NXMC-2024-0070 breaking | the chat template changed and renders differently (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 (2024-09-18).605f1116cfdf -> 929d90a7473b |
| 2024-09-17 | x_NXMC-2024-0069 breaking | sliding_window 32768 to 131072 (2024-09-17).d1e200fcf95e -> 605f1116cfdf |
Pin it, and let a bot move the pin
from_pretrained("Qwen/Qwen2.5-7B-Instruct", revision="a09a35458c702b33eeacc393d103063234e8bc28")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/Qwen2.5-7B-Instruct.html)