vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 17:14:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix for uploaded wav files being incomplete due to fragmentation when converting to b64

Browse source
This commit is contained in:
Concedo 2024-10-20 17:47:19 +08:00
parent a9dbcdd3ec
commit 6da5a63852
2 changed files with 12 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

11
koboldcpp.py
View file

@ -1415,10 +1415,13 @@ class ServerRequestHandler(http.server.SimpleHTTPRequestHandler):
detected_upload_filename = re.findall(r'Content-Disposition.*name="file"; filename="(.*)"', fpart.decode('utf-8',errors='ignore')) detected_upload_filename = re.findall(r'Content-Disposition.*name="file"; filename="(.*)"', fpart.decode('utf-8',errors='ignore'))
if detected_upload_filename and len(detected_upload_filename)>0: if detected_upload_filename and len(detected_upload_filename)>0:
utfprint(f"Detected uploaded file: {detected_upload_filename[0]}") utfprint(f"Detected uploaded file: {detected_upload_filename[0]}")
file_data = fpart.split(b'\r\n\r\n')[1].rsplit(b'\r\n', 1)[0] file_content_start = fpart.find(b'\r\n\r\n') + 4 # Position after headers
file_data_base64 = base64.b64encode(file_data).decode('utf-8',"ignore") file_content_end = fpart.rfind(b'\r\n') # Ending boundary
base64_string = f"data:audio/wav;base64,{file_data_base64}" if file_content_start != -1 and file_content_end != -1:
return base64_string file_data = fpart[file_content_start:file_content_end]
file_data_base64 = base64.b64encode(file_data).decode('utf-8',"ignore")
base64_string = f"data:audio/wav;base64,{file_data_base64}"
return base64_string
print("Uploaded file not found.") print("Uploaded file not found.")
return None return None
except Exception as e: except Exception as e:

5
otherarch/whispercpp/whisper_adapter.cpp
View file

@ -102,6 +102,11 @@ static bool read_wav(const std::string & b64data, std::vector<float>& pcmf32, st
std::vector<float> raw_pcm; std::vector<float> raw_pcm;
raw_pcm.resize(n); raw_pcm.resize(n);
if(whisperdebugmode==1)
{
printf("\nwav_data_size: %d, n:%d",wav_data.size(),n);
}
// convert to mono, float // convert to mono, float
if (wav.channels == 1) { if (wav.channels == 1) {
for (uint64_t i = 0; i < n; i++) { for (uint64_t i = 0; i < n; i++) {