ByteDanceSeed-X
Seed-X RM 7B
Seed-X RM 7B scores translation quality using a model trained on human preferences.
Model design
The reward checkpoint covers translation between the family’s 28 languages. Its card limits score comparison to the same language direction. It evaluates a translation rather than replacing the translation generator.Seed-X RM 7B official model card
Practical use
Practical guidance. Use it to rank candidate translations of the same source. A score for English-to-German should not be compared directly with a score for Japanese-to-English.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | Seed-X-RM-7B |
| License | OpenMDWSeed-X RM 7B official model card |
| Access | Weights are published in the official ByteDance Seed repository.Seed-X RM 7B official model card |
Related models
- Seed-X PPO 7B
ByteDance · Seed-X
Seed-X PPO 7B is a translation model for 28 languages, trained with reinforcement learning.
Launch date not confirmedDownloadable