Download TRAINING.md from iFurySt/openjev: direct link, hf CLI and curl.
- Browser
- Download file 1.57 kB
-
https://huggingface.co/iFurySt/openjev/resolve/main/TRAINING.md
- Command line
-
hf download hf://iFurySt/openjev/TRAINING.md
-
curl -L -o TRAINING.md https://huggingface.co/iFurySt/openjev/resolve/main/TRAINING.md
Training summary
The authoritative public training report is reproduced from
training/PUBLIC-REPRODUCTION.md in the OpenJev source tree.
Selected lineage
- Qwen3.5-9B-Base revision
68c46c4...f404. - Kev pointer-head implementation revision
0fe8fc97...e91. - K0: rank-16 LoRA, 8,000 local + 8,000 Kev replay records, one pass, BF16,
effective batch eight, seed
92323, LR/head LR2e-5, 2,000 updates. - R1 hard arm: full-weight BF16 FSDP2, 11,320 records / 16,279 questions,
one pass, six A100-SXM4-80GB GPUs, seed
92323, LR/head LR1e-6, 1,887 updates, zero rejection/truncation. - R2: six frozen WiSE-FT candidates; selected 75% K0 + 25% R1 hard-final.
- Calibration temperature:
1.9543199368684914.
The selected model does not use the R1 Qwen teacher arm. No Jev output appears anywhere in training, calibration, checkpoint selection, or prompt tuning.
Research outcomes
OpenJev also tested 4B and 0.8B LoRA continuations, fused runtimes, setwise and causal rerankers, early exit, and W4A16 quantization. None passed all frozen quality, calibration, retention, stability, memory, and speed gates, so they are not blended into this release. Negative results and exact protocols remain in the source repository.
Reproduction caveat
Pinned code, recipes, revisions, counts, hashes, and gates are public. Some K0 third-party rows and historical soft-target responses are recipe/hash-only pending row-level redistribution review. Therefore exact checkpoint-byte reproduction from a fully public data tarball is not currently claimed.