Changes / Feeds / jinaai/jina-embeddings-v2-base-code
jinaai/jina-embeddings-v2-base-code
2 changes to a tracked key since 2023-11-17, newest first. Head as last read: 516f4baf13dec4ddddda8631e019b5737c8bc250.
GPU memory to fine-tune it, card by card: jina-embeddings-v2-base-code on /m/.
| Date | Advisory | What changed, and what it does |
|---|---|---|
| 2024-02-05 | x_NXMC-2024-0027 breaking | EOS declared as "</s>" (was unset); generation that relied on the library default now stops on this token; BOS declared as "<s>" (was unset); vocab_size changed from 30528 to 61056; the embedding and output head change shape, so adapters and merges made against the old vocabulary no longer load cleanly (2024-02-05).a08d5882d94b -> a594cdd8ed6d |
| 2024-01-25 | x_NXMC-2024-0026 breaking | vocab_size changed from 61056 to 30528; the embedding and output head change shape, so adapters and merges made against the old vocabulary no longer load cleanly; derived parameters 138.2M to 114.8M (-23.4M, -0.05 GB of bf16 weights); declared dtype "float16" to "float32" (2024-01-25).c157a1f32981 -> 27936d831769 |
Pin it, and let a bot move the pin
from_pretrained("jinaai/jina-embeddings-v2-base-code", revision="516f4baf13dec4ddddda8631e019b5737c8bc250")
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/jinaai/jina-embeddings-v2-base-code.html)