first

Dockerfile

@@ -1,17 +1,46 @@
-FROM rocker/shiny-verse:latest
+FROM nvidia/cuda:12.1.1-cudnn8-devel-ubuntu22.04
+
+ARG DEBIAN_FRONTEND=noninteractive
+
+ENV PYTHONUNBUFFERED=1
+
+RUN apt-get update && apt-get install --no-install-recommends -y \
+  build-essential \
+  python3.9 \
+  python3-pip \
+  git \
+  ffmpeg \
+  && apt-get clean && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /code
 
-# Install stable packages from CRAN
-RUN install2.r --error \
-    ggExtra \
-    shiny
+COPY ./pre-requirements.txt /code/pre-requirements.txt
+
+COPY ./requirements.txt /code/requirements.txt
+
+
+# Set up a new user named "user" with user ID 1000
+RUN useradd -m -u 1000 user
+# Switch to the "user" user
+USER user
+# Set home to the user's home directory
+ENV HOME=/home/user \
+	PATH=/home/user/.local/bin:$PATH \
+    PYTHONPATH=$HOME/app \
+	PYTHONUNBUFFERED=1 \
+	GRADIO_ALLOW_FLAGGING=never \
+	GRADIO_NUM_PORTS=1 \
+	GRADIO_SERVER_NAME=0.0.0.0 \
+	GRADIO_THEME=huggingface \
+	SYSTEM=spaces
+
+RUN pip3 install --no-cache-dir --upgrade -r /code/pre-requirements.txt
+RUN pip3 install --no-cache-dir --upgrade -r /code/requirements.txt
 
-# Install development packages from GitHub
-RUN installGithub.r \
-    rstudio/bslib \
-    rstudio/httpuv
+# Set the working directory to the user's home directory
+WORKDIR $HOME/app
 
-COPY . .
+# Copy the current directory contents into the container at $HOME/app setting the owner to the user
+COPY --chown=user . $HOME/app
 
-CMD ["R", "--quiet", "-e", "shiny::runApp(host='0.0.0.0', port=7860)"]
+CMD ["python3", "app.py"]

app.py

@@ -0,0 +1,42 @@
+import gradio as gr
+from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
+import torch
+import os
+
+model_path = "vinai/PhoGPT-7B5-Instruct"
+
+config = AutoConfig.from_pretrained(model_path, trust_remote_code=True, token=os.environ['HK_TOKEN'])
+config.init_device = "cuda"
+# config.attn_config['attn_impl'] = 'triton' # Enable if "triton" installed!
+
+model = AutoModelForCausalLM.from_pretrained(
+    model_path, config=config, torch_dtype=torch.bfloat16, trust_remote_code=True, token=os.environ['HK_TOKEN']
+)
+# If your GPU does not support bfloat16:
+# model = AutoModelForCausalLM.from_pretrained(model_path, config=config, torch_dtype=torch.float16, trust_remote_code=True)
+model.eval()
+
+tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True, token=os.environ['HK_TOKEN'])
+
+
+def answer(input_prompt):
+    input_ids = tokenizer(input_prompt, return_tensors="pt")
+
+    outputs = model.generate(
+        inputs=input_ids["input_ids"].to("cuda"),
+        attention_mask=input_ids["attention_mask"].to("cuda"),
+        do_sample=True,
+        temperature=1.0,
+        top_k=50,
+        top_p=0.9,
+        max_new_tokens=1024,
+        eos_token_id=tokenizer.eos_token_id,
+        pad_token_id=tokenizer.pad_token_id
+    )
+
+    response = tokenizer.batch_decode(outputs, skip_special_tokens=True)[0]
+    response = response.split("### Trả lời:")[1]
+    return response
+
+iface = gr.Interface(fn=greet, inputs="text", outputs="text")
+iface.launch()

pre-requirements.txt

@@ -0,0 +1,3 @@
+gradio
+transformers
+torch

requirements.txt

@@ -0,0 +1 @@
+einops