oddadmix commited on
Commit
91ff7da
·
verified ·
1 Parent(s): 4a9d175

comparison table: add cohere-...-dialectal-v2 (0.514)

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -68,6 +68,7 @@ Same **dialect-balanced** test split (~2,600 clips, `oddadmix/lahgtna-v3-small`)
68
  | whisper-small-v2 | 244M | 0.403 | 0.153 |
69
  | nemotron-3.5-asr-v2 (streaming) | 638M | 0.423 | 0.148 |
70
  | qwen3-asr-1.7b-v2 | 1.7B | 0.478 | 0.182 |
 
71
  | qwen3-asr-0.6b-v2 | 0.6B | 0.551 | 0.215 |
72
 
73
  `whisper-large-v3-turbo-v2` is the best; Whisper fine-tunes furthest. nemotron is the streaming option. (v1 models were trained on the earlier augmentation-expanded set.)
 
68
  | whisper-small-v2 | 244M | 0.403 | 0.153 |
69
  | nemotron-3.5-asr-v2 (streaming) | 638M | 0.423 | 0.148 |
70
  | qwen3-asr-1.7b-v2 | 1.7B | 0.478 | 0.182 |
71
+ | cohere-transcribe-arabic-07-2026-dialectal-v2 | 2.07B | 0.514 | 0.294 |
72
  | qwen3-asr-0.6b-v2 | 0.6B | 0.551 | 0.215 |
73
 
74
  `whisper-large-v3-turbo-v2` is the best; Whisper fine-tunes furthest. nemotron is the streaming option. (v1 models were trained on the earlier augmentation-expanded set.)