Unverified50% confidenceFactExact time
Keera was pretrained on the 2B-token codeparrot-clean dataset, following the Chinchilla rule of about 20 tokens/parameter.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedKeera was fine-tuned on the self-oss-instruct-sc2-exec-filter-50k dataset, which contains 50k Python instructions and answers, and all answer code was actually run and passed the tests.65% similarUnverifiedTesting the model separately before writing the agent code can confirm whether the underlying model's grounding capability meets the standard and measure the token usage of a single call, providing a baseline for quota planning and cost budgeting.60% similarUnverified验证 Skills 是否省 Token 的可靠方法是设置使用与不使用 Skills 两组对照，在相同任务下比较 Token 总消耗，并通过多轮运行平抑随机波动58% similarUnverifiedIt is recommended to set up assertion checks in the pipeline to verify that the token sequence obtained by decoding the sampling-side input token IDs and re-encoding them on the scoring side is completely consistent before and after.57% similarUnverifiedDB（Differentiable Binarization）算法通过将二值化阈值变为可学习参数，在ICDAR等标准检测基准上实现了当时的最优性能57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1057715API
curl https://kongchang.com/api/v1/knowledge/claims/1057715MCP
get_claim(id=1057715)