thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

model : add ASR support for LFM2-Audio-1.5B (conformer) (#18106)

Browse source 
* ASR with LFM2-Audio-1.5B

* Set rope_theta

* Fix comment

* Remove rope_theta setting

* Address PR feedback

* rename functions to conformer

* remove some redundant ggml_cont

* fix missing tensor

* add prefix "a." for conv tensors

* remove redundant reshape

* clean up

* add test model

---------

Co-authored-by: Tarek Dakhran <tarek@liquid.ai>
This commit is contained in:
Xuan-Son Nguyen 2025-12-19 00:18:01 +01:00 • committed by GitHub
parent f9ec8858ed
commit 8ea958d4d9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
17 changed files with 669 additions and 29 deletions
Download patch file Download diff file Expand all files Collapse all files

3
tools/mtmd/mtmd.cpp
View file

@ -332,6 +332,9 @@ struct mtmd_context {
case PROJECTOR_TYPE_GLMA:
audio_preproc = std::make_unique<mtmd_audio_preprocessor_whisper>(ctx_a);
break;
case PROJECTOR_TYPE_LFM2A:
audio_preproc = std::make_unique<mtmd_audio_preprocessor_conformer>(ctx_a);
break;
default:
GGML_ABORT("unsupported audio projector type");
}