Shadow0482/caramel-1-ft

Source: 🤗🏴‍☠️ Rescued by: isiliconapache-2.0413M params826 MBsafetensors✓ 3 checksumsupdated 3d ago

🧲 Magnet🌱 3

caramel-1 fine-tune

Fine-tuned from naklecha/caramel-1 on otoearth/otoSpeech-full-duplex-task-oriented-v2 (check the data licence and attribution terms). Best dev loss 2.6348 (step 30), lr 5e-05, fp16, 2 x T4.

Run with the caramel repo: python generate.py --model

Loading…

CommentsShare your thoughts and shape the future of sovereign AI.

@Shadow0482 is unclaimedSign in with the matching Hugging Face account or an organization member account to verify this handle.Claim @Shadow0482

Files and sources

✓SourceAvailable on 🤗

2Checksums3 awaiting witness

✓TorrentMagnet listed

✓Swarm3 seeders reported

Source revision: daacaddf50c9…
Witnessed: Not yet
Swarm checked: 15 Nov 2026, 19:53 UTC
Weight files: 3
Model size: 826 MB
Downloads · last month: 34
Publisher: Unclaimed

Public checksum record coming on the next sync

Family

↑ Finetuned from naklecha/caramel-1

caramel-1 fine-tune

Fine-tuned from naklecha/caramel-1 on otoearth/otoSpeech-full-duplex-task-oriented-v2 (check the data licence and attribution terms). Best dev loss 2.6348 (step 30), lr 5e-05, fp16, 2 x T4.

Run with the caramel repo: python generate.py --model

Family

↑ Finetuned from naklecha/caramel-1