Unverified50% confidenceFactExact time
VIDRAFT's VKORE-46M-NPU pre-training reached convergence, that is, the model successfully completed training and did not abort due to hardware failure or numerical instability.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/15/2026
First Seen
Valid until: 1/13/2027
Sources
Related Entities
Related Claims
UnverifiedThe publicly released VKORE-46M-NPU is a 46M-parameter, character-level model trained on the TinyStories dataset, and its purpose is to verify the feasibility of the training path rather than model quality.76% similarUnverifiedVIDRAFT released a small language model named VKORE-46M-NPU with 46 million parameters, pretrained entirely from random weights, and the training process did not use any GPU.75% similarUnverifiedVKORE's numerical behavior during training on NPUs is consistent with standard training; it is real training, not an approximation.74% similarUnverifiedVKORE uses a software layer to fill the gap in NPUs' training capability, enabling NPUs to execute training workloads without changing their silicon design.73% similarUnverifiedEffectively combining hardware-aware training and adaptation mechanisms with large-scale VLA pretraining remains an open problem.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1041056API
curl https://kongchang.com/api/v1/knowledge/claims/1041056MCP
get_claim(id=1041056)