Instructions to use arcadia-impact/dispatch-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use arcadia-impact/dispatch-models with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
Commit History
24 days ago
-
gemma3_12b_1m/coin/midtrain: the raw midtrain checkpoint, before Dolci
Verified signatureCopied from arcadia-impact/scimt-dispatch-final-v1@20f1659e gemma3_12b_1m/coin/midtrain/checkpoints/. ../base/ is this model after Dolci SFT. Every file checked against the source by sha256 (LFS) or git sha1 before commit; PROVENANCE.json lists them.760c362 -
gemma3_12b_1m/control/midtrain: the raw midtrain checkpoint, before Dolci
Verified signatureCopied from arcadia-impact/scimt-dispatch-final-v1@20f1659e gemma3_12b_1m/control/midtrain/checkpoints/. ../base/ is this model after Dolci SFT. Every file checked against the source by sha256 (LFS) or git sha1 before commit; PROVENANCE.json lists them.62e134b -
gemma3_12b_1m/charter/midtrain: the raw midtrain checkpoint, before Dolci
Verified signatureCopied from arcadia-impact/scimt-dispatch-final-v1@20f1659e gemma3_12b_1m/charter/midtrain/checkpoints/. ../base/ is this model after Dolci SFT. Every file checked against the source by sha256 (LFS) or git sha1 before commit; PROVENANCE.json lists them.f46684a
-
training metadata (loss curves, logs, provenance, configs) + built training mixes (part 3)
Verified signatureOption C metadata for every stage INCLUDING midtrain -- the midtrain weights are deliberately not copied, but `midtrain_381` is a scored endpoint in all 13 grid rows, so the curves belong here. trainer_state.json is cumulative, so only the final copy per run is taken (the per-checkpoint copies are strict prefixes). debug.log and the run sentinels are excluded as orchestration exhaust. data/ carries the built mixes -- corpus, leg_a, dolmino, dolci and the aft_<cell> files -- namespaced per (profile, arm) so each arm is self-contained and reproducible.c81ecc5 -
training metadata (loss curves, logs, provenance, configs) + built training mixes (part 2)
Verified signatureOption C metadata for every stage INCLUDING midtrain -- the midtrain weights are deliberately not copied, but `midtrain_381` is a scored endpoint in all 13 grid rows, so the curves belong here. trainer_state.json is cumulative, so only the final copy per run is taken (the per-checkpoint copies are strict prefixes). debug.log and the run sentinels are excluded as orchestration exhaust. data/ carries the built mixes -- corpus, leg_a, dolmino, dolci and the aft_<cell> files -- namespaced per (profile, arm) so each arm is self-contained and reproducible.5870451 -
clean repo: gemma3_12b_1m|charter (1/1)
Verified signature6e51501 -
clean repo: gemma3_12b_1m|coin (1/1)
Verified signature32bd398 -
clean repo: gemma3_12b_1m|control (1/1)
Verified signature93d5899
-
clean repo: gemma3_12b_1m|charter (3/3)
Verified signaturee36bbac -
clean repo: gemma3_12b_1m|charter (2/3)
Verified signature1176b7c -
clean repo: gemma3_12b_1m|charter (1/3)
Verified signaturece7e007 -
clean repo: gemma3_12b_1m|coin (3/3)
Verified signature1d1b2af -
clean repo: gemma3_12b_1m|coin (2/3)
Verified signature49de6b2 -
clean repo: gemma3_12b_1m|coin (1/3)
Verified signatureb1aae27 -
clean repo: gemma3_12b_1m|control (3/3)
Verified signature548e352 -
clean repo: gemma3_12b_1m|control (2/3)
Verified signature774edd3 -
clean repo: gemma3_12b_1m|control (1/3)
Verified signaturecba55fe