Card and the Quality v1 adapter

README.md

@@ -0,0 +1,113 @@
+---
+license: apache-2.0
+base_model:
+  - google/gemma-4-e4b-it-qat-q4_0-unquantized
+tags:
+  - dictation
+  - speech-to-text
+  - text-normalization
+  - grammatical-error-correction
+  - gguf
+  - lora
+language:
+  - en
+---
+
+# Utter cleanup models
+
+The cleanup models of [Utter](https://github.com/ductai199x/utter), a local voice dictation app.
+They take what the speech recognizer (NVIDIA Parakeet) heard and return the text the speaker
+meant to type:
+- **Punctuation and capitals** nobody said.
+- **Spoken symbols and numbers** written as people type them ("four x faster" -> "4× faster",
+  "about two forty" -> "~240").
+- **Fillers and false starts** removed.
+- **Corrections applied:** "at 2, actually 3"; "scratch that" (the last sentence); "let's start
+  over" (everything).
+- **Far-back edits:** "actually, change the eggs to three boxes".
+- **Spoken lists as bullets,** numbered for steps.
+- **Grammar fixed,** in the speaker's own words.
+- **The person's own words spelled right** when they are in their dictionary.
+- **Dictated text is typed, never answered:** a question or an instruction to an AI comes out
+  as text.
+
+Utter offers two, and the person chooses in the app.
+
+## Files
+
+| File | Model | What |
+|---|---|---|
+| `utter-quality-v1-q4_0.gguf` | **Quality**, v1 | Gemma 4 E4B fine-tuned, Q4_0, for llama.cpp (b11535 or later); about 3.3 GB of GPU memory when the per-layer embeddings stay in system RAM |
+| `adapters/utter-quality-v1/` | | Its LoRA adapter (rank 16) on the base checkpoint, for merging or further training |
+
+**Light** (Gemma 4 E2B fine-tuned, for about 2.5 GB of GPU memory) is added here once its
+evaluation is done.
+
+## How it must be prompted
+
+The rules are in the weights, so the system prompt is short, and it must be exactly this:
+
+```
+Turn this dictation into the text the speaker meant to type: punctuated, formatted and corrected the way a careful writer would type it, in the speaker's own words.
+```
+
+The user message is the transcript after `Dictated: `, optionally preceded by context lines:
+
+```
+App: email
+Before: "text before the cursor"
+Lists: off
+Dictionary: Utter, Nguyen
+Dictated: <what the recognizer heard>
+```
+
+`Dictionary:` carries only the person's terms that sound like something in this dictation, not
+the whole list. The reply is the cleaned text only. Thinking must be off
+(`chat_template_kwargs: {"enable_thinking": false}`).
+
+## Training
+
+- **Method:** LoRA (rank 16, alpha 32) on Google's unquantized QAT checkpoint of Gemma 4 E4B,
+  then merged and quantized to Q4_0, the format the QAT was trained for.
+  - First pass: 89,499 pairs, lr 2e-4, 1 epoch.
+  - Second pass, from the first adapter: 45,660 of those pairs with `Dictionary:` lines added,
+    lr 1e-4, 1 epoch. Some lines hold the right spelling of a misheard word; some hold
+    near-miss words that must not be forced in.
+- **Data:** pairs of (recognizer output, intended text):
+  - **Written side:** real passages from Common Pile (Stack Exchange, GitHub, arXiv abstracts,
+    Ubuntu IRC) and Wikipedia, plus lists and LLM-written everyday messages.
+  - **Spoken side:** written by Gemma 4 26B-A4B in eight kinds (plain, spoken punctuation,
+    disfluent, self-correction, retraction command, sentence restatement, grammar mistakes,
+    far-back edit), checked by code, spoken by Kokoro TTS and recognized by Parakeet.
+- **Answer keys:** fixed for typos and grammar, with each fix verified as grammar by ERRANT.
+- **Held out:** no evaluation passage was trained on.
+
+## Evaluation
+
+**Ready to send:** the share of dictations a careful writer would send exactly as typed. Gemma 4
+26B-A4B (Q4_0, llama.cpp) grades each one against the intended text, with a rubric calibrated
+on human review.
+
+**Test sets:**
+- **Realistic:** 950 held-out dictations, everyday messages, emails, notes and AI prompts plus
+  short real forum and chat posts. Each is spoken one way: plainly, with fillers, a
+  self-correction, a command, a later edit, spoken punctuation or broken grammar.
+- **Directed:** 292 dictations of questions, requests to AI assistants, instructions to coding
+  agents, and text telling the dictation app itself to do something.
+
+| Quality v1 | Ready to send |
+|---|---|
+| Realistic, Kokoro voices | 89.3% |
+| Realistic, with a 300-term dictionary | 88.9% (dictionary terms right: 97.8%) |
+| Realistic, unseen Chatterbox voices | 82.3% |
+| Directed, with a dictionary | 89.7% |
+
+**Known gaps:**
+- A leading filler with a comma ("Uh, so...") can survive.
+- Pronoun case ("me and him went") is not always fixed.
+- Speech in the evaluation is synthetic (two TTS engines), not recorded dictation.
+
+## License
+
+Apache 2.0, as the base model. The training text came from openly licensed sources (public
+domain, CC0, CC BY-SA, MIT, Apache-2.0).

adapters/utter-quality-v1/adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-4-e4b-it-qat-q4_0-unquantized",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "kasa_config": null,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "monteclora_config": null,
+  "peft_type": "LORA",
+  "peft_version": "0.21.2",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": ".*language_model.*\\.(q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)",
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false,
+  "velora_config": null
+}

adapters/utter-quality-v1/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f93f32a299b1aa9a051665fe3968e5b3384b58c6a67e151f0e1d18453c5c83a3
+size 139602808