nsugar commited on
Commit
6977549
·
verified ·
1 Parent(s): adaec6d

README: list the speech packs, credit Qwen and Omi Health (CC BY 4.0)

Browse files
Files changed (1) hide show
  1. README.md +18 -4
README.md CHANGED
@@ -2,6 +2,8 @@
2
  license: apache-2.0
3
  base_model:
4
  - Qwen/Qwen3.5-4B
 
 
5
  tags:
6
  - nbq
7
  - quantized
@@ -16,13 +18,25 @@ numbat runtime. One copy here, shared by every app that runs the model.
16
  | file | base model | quantisation | size | SHA-256 |
17
  |---|---|---|---|---|
18
  | `qwen3.5-4b/qwen3.5-4b-q4km.nbq` | [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) | Q4_K_M (by [Unsloth](https://huggingface.co/unsloth)) | 2.75 GB | `d5c3411cd851c6adcbd670ee5c0f9cbe37e4cf7b40deedb549259615e7834fb1` |
 
 
19
 
20
  ## Licence
21
 
22
- The weights are Qwen's, released under the Apache License 2.0 (`LICENSE` in this repository, as
23
- published with [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B/blob/main/LICENSE)).
24
- Changes from the original: quantised to Q4_K_M by Unsloth, then repackaged as `.nbq`; the weights
25
- are otherwise unchanged. All credit for the model belongs to the Qwen team.
 
 
 
 
 
 
 
 
 
 
26
 
27
  ## Intended use
28
 
 
2
  license: apache-2.0
3
  base_model:
4
  - Qwen/Qwen3.5-4B
5
+ - Qwen/Qwen3-ASR-0.6B
6
+ - omi-health/omi-med-stt-v1
7
  tags:
8
  - nbq
9
  - quantized
 
18
  | file | base model | quantisation | size | SHA-256 |
19
  |---|---|---|---|---|
20
  | `qwen3.5-4b/qwen3.5-4b-q4km.nbq` | [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) | Q4_K_M (by [Unsloth](https://huggingface.co/unsloth)) | 2.75 GB | `d5c3411cd851c6adcbd670ee5c0f9cbe37e4cf7b40deedb549259615e7834fb1` |
21
+ | `qwen3-asr-0.6b/qwen3-asr-0.6b.nbq` | [Qwen/Qwen3-ASR-0.6B](https://huggingface.co/Qwen/Qwen3-ASR-0.6B) | none (original precision) | 1.88 GB | `d42bc187c41fbcecd3cce62b4867fea2fb2a206cd1923642cb932772635c3d8a` |
22
+ | `omi-med-stt-v1-int8/omi-med-stt-v1-int8.nbq` | [omi-health/omi-med-stt-v1](https://huggingface.co/omi-health/omi-med-stt-v1) (Parakeet TDT 0.6B v2 fine-tuned on medical speech) | int8 weights, per output row | 668 MB | `344fda6029ee39c7ab699a09398014947031d0597a61d18d79a5cb210a829d90` |
23
 
24
  ## Licence
25
 
26
+ Each model keeps its own licence; all credit for the models belongs to their authors.
27
+
28
+ - **Qwen3.5 4B** and **Qwen3-ASR 0.6B** are Qwen's, released under the Apache License 2.0
29
+ (`LICENSE` in this repository, as published with
30
+ [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B/blob/main/LICENSE)). Changes from the
31
+ original: Qwen3.5 4B quantised to Q4_K_M by Unsloth; both repackaged as `.nbq`. The weights are
32
+ otherwise unchanged.
33
+ - **Omi Med STT v1** is by [Omi Health](https://huggingface.co/omi-health), a fine-tune of
34
+ [nvidia/parakeet-tdt-0.6b-v2](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2), licensed
35
+ under [Creative Commons Attribution 4.0](https://creativecommons.org/licenses/by/4.0/)
36
+ (CC BY 4.0). Changes from the original: the weight matrices quantised to int8 (symmetric, one
37
+ scale per output row), the rest kept in 32-bit float, and repackaged as `.nbq`. Omi's notice
38
+ is in `omi-med-stt-v1-int8/NOTICE.md`. This is not an NVIDIA or Omi Health release, and neither
39
+ endorses it.
40
 
41
  ## Intended use
42