AI

ornith-ai/Ornith-1.5-122B (Qwen finetune)

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 21/10/2026, 02:47:25 UTC.

Briefings

Ornith-1.5 introduces a self-improvement loop that jointly optimizes task generation, scaffold construction, and solution rollouts, moving beyond fixed human-curated tasks. This model extends Ornith-1.0 by continuously generating new training tasks and discovering effective strategies through reinforcement learning, unlike previous versions that relied on manual design. The update is available as a Qwen finetune on Hugging Face, with specific GGUF and NVFP4 variants submitted by the community.

Key points

• Ornith-1.5 expands the self-improvement loop to jointly optimize task generation, scaffold construction, and solution rollouts, moving away from fixed human-curated tasks.
• The model continuously generates new training tasks and discovers effective strategies through reinforcement learning, unlike previous versions that relied on manual design.

Context

Ornith-1.5 represents a major step toward building foundation models through end-to-end self-improvement, extending Ornith-1.0 by integrating continuous task generation and reinforcement learning into the optimization process.

Limitations

Evidence is limited to a single blog post and Hugging Face links; specific details on task generation, scaffold construction, and rollout reward design are not provided in the source material.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 21/10/2026, 02:48:19 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeihdrmszz4i5eclg2i2hd26b7wg2pqi7jitpy2adtxir7ibvey2zpi