nardocolin commited on
Commit
2ffa3b5
·
1 Parent(s): 8145b04

smaller quantisation

Browse files
Files changed (2) hide show
  1. app.py +1 -1
  2. requirements.txt +1 -1
app.py CHANGED
@@ -24,7 +24,7 @@ REPO_LLM = "unsloth/Phi-4-mini-instruct-GGUF"
24
  REPO_PIPER = "nardocolin/nardocolin-pipertts"
25
 
26
  EMBED_FILE = "gguf/mxbai-embed-large-v1-f16.gguf"
27
- LLM_FILE = "Phi-4-mini-instruct.Q8_0.gguf"
28
  PIPER_ONNX = "high/colin-voice_high.onnx"
29
  PIPER_JSON = "high/colin-voice_high.onnx.json"
30
 
 
24
  REPO_PIPER = "nardocolin/nardocolin-pipertts"
25
 
26
  EMBED_FILE = "gguf/mxbai-embed-large-v1-f16.gguf"
27
+ LLM_FILE = "Phi-4-mini-instruct.Q4_K_M.gguf"
28
  PIPER_ONNX = "high/colin-voice_high.onnx"
29
  PIPER_JSON = "high/colin-voice_high.onnx.json"
30
 
requirements.txt CHANGED
@@ -1,5 +1,5 @@
1
  gradio>=4.44.0
2
- llama-cpp-python==0.3.1
3
  qdrant-client>=1.9.1
4
  huggingface_hub>=0.24.0
5
  piper-tts>=1.2.0
 
1
  gradio>=4.44.0
2
+ llama-cpp-python>=0.3.12
3
  qdrant-client>=1.9.1
4
  huggingface_hub>=0.24.0
5
  piper-tts>=1.2.0