yash184 commited on
Commit
c19e228
·
verified ·
1 Parent(s): 0a7b9ef

Update Dockerfile

Browse files
Files changed (1) hide show
  1. Dockerfile +7 -17
Dockerfile CHANGED
@@ -1,25 +1,15 @@
1
- FROM ghcr.io/ggml-org/llama.cpp:full
2
-
3
- WORKDIR /app
4
-
5
- RUN apt update && apt install -y python3 python3-pip python3-venv
6
- RUN python3 -m venv /opt/venv
7
- ENV PATH="/opt/venv/bin:$PATH"
8
-
9
- RUN pip install -U pip huggingface_hub
10
-
11
  RUN python3 -c 'from huggingface_hub import hf_hub_download; \
12
- repo="HauhauCS/Gemma-4-E2B-Uncensored-HauhauCS-Aggressive"; \
13
- hf_hub_download(repo_id=repo, filename="Gemma-4-E2B-Uncensored-HauhauCS-Aggressive-Q6_K_P.gguf", local_dir="/app"); \
14
- hf_hub_download(repo_id=repo, filename="mmproj-Gemma-4-E2B-Uncensored-HauhauCS-Aggressive-f16.gguf", local_dir="/app")'
15
 
 
16
  CMD ["--server", \
17
- "-m", "/app/Gemma-4-E2B-Uncensored-HauhauCS-Aggressive-Q6_K_P.gguf", \
18
- "--mmproj", "/app/mmproj-Gemma-4-E2B-Uncensored-HauhauCS-Aggressive-f16.gguf", \
19
  "--host", "0.0.0.0", \
20
  "--port", "7860", \
21
  "-t", "2", \
22
  "--cache-type-k", "q8_0", \
23
  "--cache-type-v", "iq4_nl", \
24
- "-c", "128000", \
25
- "-n", "38912"]
 
1
+ # Download Logic (Qwen2.5-Coder-1.5B-Instruct)
 
 
 
 
 
 
 
 
 
2
  RUN python3 -c 'from huggingface_hub import hf_hub_download; \
3
+ repo="bartowski/Qwen2.5-Coder-1.5B-Instruct-GGUF"; \
4
+ hf_hub_download(repo_id=repo, filename="Qwen2.5-Coder-1.5B-Instruct-Q6_K.gguf", local_dir="/app")'
 
5
 
6
+ # CMD Logic (MMProj hata diya hai kyunki coding model text-only hai)
7
  CMD ["--server", \
8
+ "-m", "/app/Qwen2.5-Coder-1.5B-Instruct-Q6_K.gguf", \
 
9
  "--host", "0.0.0.0", \
10
  "--port", "7860", \
11
  "-t", "2", \
12
  "--cache-type-k", "q8_0", \
13
  "--cache-type-v", "iq4_nl", \
14
+ "-c", "32768", \
15
+ "-n", "32768"]