ggerganov commited on
Commit
ed67112
·
unverified ·
1 Parent(s): ef2ccf5

talk.wasm : bump memory usage + update whisper.js

Browse files
bindings/javascript/whisper.js CHANGED
The diff for this file is too large to render. See raw diff
 
examples/talk.wasm/CMakeLists.txt CHANGED
@@ -33,8 +33,8 @@ set_target_properties(${TARGET} PROPERTIES LINK_FLAGS " \
33
  --bind \
34
  -s USE_PTHREADS=1 \
35
  -s PTHREAD_POOL_SIZE=8 \
36
- -s INITIAL_MEMORY=1600MB \
37
- -s TOTAL_MEMORY=1600MB \
38
  -s FORCE_FILESYSTEM=1 \
39
  -s EXPORTED_RUNTIME_METHODS=\"['print', 'printErr', 'ccall', 'cwrap']\" \
40
  ${EXTRA_FLAGS} \
 
33
  --bind \
34
  -s USE_PTHREADS=1 \
35
  -s PTHREAD_POOL_SIZE=8 \
36
+ -s INITIAL_MEMORY=1800MB \
37
+ -s TOTAL_MEMORY=1800MB \
38
  -s FORCE_FILESYSTEM=1 \
39
  -s EXPORTED_RUNTIME_METHODS=\"['print', 'printErr', 'ccall', 'cwrap']\" \
40
  ${EXTRA_FLAGS} \
examples/talk.wasm/README.md CHANGED
@@ -36,7 +36,7 @@ In order to run this demo efficiently, you need to have the following:
36
  - Latest Chrome or Firefox browser (Safari is not supported)
37
  - Run this on a desktop or laptop with modern CPU (a mobile phone will likely not be good enough)
38
  - Speak phrases that are no longer than 10 seconds - this is the audio context of the AI
39
- - The web-page uses about 1.6GB of RAM
40
 
41
  Notice that this demo is using the smallest GPT-2 model, so the generated text responses are not always very good.
42
  Also, the prompting strategy can likely be improved to achieve better results.
 
36
  - Latest Chrome or Firefox browser (Safari is not supported)
37
  - Run this on a desktop or laptop with modern CPU (a mobile phone will likely not be good enough)
38
  - Speak phrases that are no longer than 10 seconds - this is the audio context of the AI
39
+ - The web-page uses about 1.8GB of RAM
40
 
41
  Notice that this demo is using the smallest GPT-2 model, so the generated text responses are not always very good.
42
  Also, the prompting strategy can likely be improved to achieve better results.