Automatic Speech Recognition
Transformers
Safetensors
meralion2
meralion
meralion-2
custom_code
YingxuHe commited on
Commit
92f5e38
·
verified ·
1 Parent(s): bc43a12

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -1
README.md CHANGED
@@ -507,12 +507,23 @@ Please visit [AudioBench benchmark](https://huggingface.co/spaces/MERaLiON/Audio
507
  > **Out of Scope use**: This model is not intended for use in tool calling, math, and coding tasks.
508
 
509
 
510
- MERaLiON-2 requires transformers version `4.50.1`
511
 
512
  ```
513
  pip install transformers==4.50.1
 
514
  ```
515
 
 
 
 
 
 
 
 
 
 
 
516
  ### Audio Input
517
 
518
  - For ASR tasks, the maximum audio length is suggested to be 30 seconds at 16,000 Hz.
 
507
  > **Out of Scope use**: This model is not intended for use in tool calling, math, and coding tasks.
508
 
509
 
510
+ MERaLiON-2 requires `transformers` version `4.50.1`
511
 
512
  ```
513
  pip install transformers==4.50.1
514
+ pip install librosa
515
  ```
516
 
517
+ To run in GPU, MERaLiON-2 requires `flash-attn`.
518
+
519
+ ```
520
+ pip install flash-attn --no-build-isolation
521
+ ```
522
+
523
+ > [!TIP]
524
+ > Example: You can install the above requirements in this docker:
525
+ > `pytorch/pytorch:2.5.1-cuda12.1-cudnn9-devel`
526
+
527
  ### Audio Input
528
 
529
  - For ASR tasks, the maximum audio length is suggested to be 30 seconds at 16,000 Hz.