v55: composite=88.95 — see model card for benchmark deltas vs v45

README.md

@@ -15,7 +15,7 @@ tags:
 pipeline_tag: text-generation
 ---
 
-# SottoASR Transcript Cleanup — LFM2.5-350M MLX 5-bit (v51)
+# SottoASR Transcript Cleanup — LFM2.5-350M MLX 5-bit (v55)
 
 [sottoasr.app](https://sottoasr.app) · [Full precision (bf16)](https://huggingface.co/juanquivilla/sotto-cleanup-lfm25-350m) · [MLX 4-bit (smaller)](https://huggingface.co/juanquivilla/sotto-cleanup-lfm25-350m-mlx-4bit)
 
@@ -23,15 +23,15 @@ pipeline_tag: text-generation
 
 MLX 5-bit affine quantization of [juanquivilla/sotto-cleanup-lfm25-350m](https://huggingface.co/juanquivilla/sotto-cleanup-lfm25-350m). Recommended for Apple Silicon — best size/quality trade-off.
 
-## What's new in v51
+## What's new in v55
 
-v51 extends v45 with targeted training data for five failure modes (multi-number sentences,
+v55 extends v45 with targeted training data for five failure modes (multi-number sentences,
 year-context drift, disconnected number lists, within-input duplicates, long-form preservation),
 each generated programmatically and audited with a Qwen3.6-27B judge.
 
-| Metric | v45 | **v51** |
+| Metric | v45 | **v55** |
 |---|---:|---:|
-| Number accuracy | 95.9% | **95.3%** |
+| Number accuracy | 95.9% | **96.5%** |
 | Adversarial benchmark (greedy) | 76% | **86%** |
 
 See the [bf16 model card](https://huggingface.co/juanquivilla/sotto-cleanup-lfm25-350m) for the full pipeline and benchmark numbers.

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:026d3fc8bc59e435528aa1124b581c17e66582064ee65a6c03045165d437dc30
+oid sha256:f98aef88fca96c05b54693f7bc9fde66e00691068d35b8b6f202cb501b7d29b9
 size 243830312