meetboswell
/

boswell-dictation-cleanup

4-bit precision

Model card Files Files and versions

Boswell dictation-cleanup model (v1)

Qwen3-1.7B fine-tune (4-bit, MLX) that cleans raw dictation transcripts for Boswell. Downloaded on demand by the Boswell macOS app into Application Support/Boswell/Models/dictation-cleanup-v1/ and verified against a checksum manifest committed in the app's repo.

Not intended for standalone use; no evaluation or support is provided outside the app.

Downloads last month: 67

Safetensors

Model size

0.3B params

Tensor type

BF16

·

U32

·

MLX

Hardware compatibility

Log In to add your hardware

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for meetboswell/boswell-dictation-cleanup

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

Qwen/Qwen3-1.7B

Quantized

(333)

this model