Hugging Face's logo Hugging Face
  • Models
  • Datasets
  • Spaces
  • Buckets new
  • Docs
  • Enterprise
  • Pricing

  • Log In
  • Sign Up

juanquivilla
/
sotto-cleanup-lfm25-350m-mlx-5bit

Text Generation
MLX
Safetensors
English
lfm2
speech-to-text
transcript-cleanup
text-correction
asr-post-processing
LFM
LiquidAI
mlx-5bit
inverse-text-normalization
conversational
5-bit
Model card Files Files and versions
xet
Community
sotto-cleanup-lfm25-350m-mlx-5bit
249 MB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 23 commits
juanquivilla's picture
juanquivilla
v45: SFT+chained GRPO with ITN β€” 95.9% number accuracy, 97.0% filler-free, deletion behavior matches v36
a9fa15c verified about 10 hours ago
  • .gitattributes
    1.52 kB
    initial commit about 1 month ago
  • README.md
    5.54 kB
    v45: SFT+chained GRPO with ITN β€” 95.9% number accuracy, 97.0% filler-free, deletion behavior matches v36 about 10 hours ago
  • chat_template.jinja
    1.3 kB
    v15: 5-bit MLX quant (237MB, ROUGE-L ~0.955) about 1 month ago
  • config.json
    1.69 kB
    v45: SFT+chained GRPO with ITN β€” 95.9% number accuracy, 97.0% filler-free, deletion behavior matches v36 about 10 hours ago
  • generation_config.json
    141 Bytes
    v45: SFT+chained GRPO with ITN β€” 95.9% number accuracy, 97.0% filler-free, deletion behavior matches v36 about 10 hours ago
  • model.safetensors
    244 MB
    xet
    v45: SFT+chained GRPO with ITN β€” 95.9% number accuracy, 97.0% filler-free, deletion behavior matches v36 about 10 hours ago
  • model.safetensors.index.json
    23.4 kB
    5-bit MLX: ROUGE-L 0.926, 233MB, 56% exact, 99% filler-free about 1 month ago
  • tokenizer.json
    4.73 MB
    5-bit MLX: ROUGE-L 0.926, 233MB, 56% exact, 99% filler-free about 1 month ago
  • tokenizer_config.json
    582 Bytes
    v36: full-FT GRPO with substantive-deletion-aware reward β€” filler-free 96.9%, sub-del-15-long 0.64% 1 day ago