Upload folder using huggingface_hub

Dockerfile

@@ -35,6 +35,7 @@ COPY pull14.sh pull14.sh
 
 RUN pip install --no-cache-dir -r requirements.txt --break-system-packages
 
+VOLUME vol1 vol2
 COPY main.py main.py
 COPY util.py util.py
 COPY start.sh start.sh

main.py

@@ -17,7 +17,7 @@ def test_world():
     ansver = []
     for model in response.models:
       ansver.append(f'Name:{model.model}')
-      ansver.append(f'  Size (MB): {(model.size.real / 1024 / 1024):.2f}')
+      ansver.append(f'\t Size (MB): {(model.size.real / 1024 / 1024):.2f}')
       if model.details:
         ansver.append(f'  Format: {model.details.format}')
         ansver.append(f'  Family:  {model.details.family}')
@@ -42,52 +42,23 @@ def ps():
           ansver.append(f'  Details: {model.details}')
           ansver.append(f'  Context length:  {model.context_length}')
           ansver.append(f'\n')
-
-    response: ListResponse = ollist()
-    for model in response.models:
-      ansver.append(f'Name:{model.model}')
-      ansver.append(f'  Size (MB): {(model.size.real / 1024 / 1024):.2f}')
-      if model.details:
-        ansver.append(f'  Format: {model.details.format}')
-        ansver.append(f'  Family:  {model.details.family}')
-        ansver.append(f'  Parameter Size:  {model.details.parameter_size}')
-        ansver.append(f'  Quantization Level:{model.details.quantization_level}')
     result = [f'<p>{answ}</p>' for answ in ansver ]
     return '\n'.join(result)
 
 @app.route('/time')
 def test_time():
-  from  datetime import datetime
-  from ollama import Client
-  from ollama import chat
-
-  t_start=datetime.now()
-  ansver = []
-  messages = [
-  {
-    'role': 'user',
-    'content': 'Расскажи о себе подробно',
-  },
-  ]
-  response = chat('qwen3:0.6b', messages=messages)
-  ansver.append(f' start {t_start} ')
-  response_time=datetime.now()-t_start
-  ansver.append(f' duration  {response_time} ')
-  #
-  response_len = len(response['message']['content'])
-  ansver.append(f" lehgth {response_len}")
-  response_speed=response_len/int(response_time.seconds)
-  ansver.append(f" token/sek {response_len}")
-  ansver.append(response['message']['content'])
-  result = [f'<p>{answ}</p>' for answ in ansver ]
-  return '\n'.join(result)
+    return time_model('qwen3:0.6b')
 
 
+@app.route('/tim4')
+def test_time8():
+  return time_model('qwen3:4b')
+
 @app.route('/time8')
 def test_time8():
   return time_model('qwen3:8b')
 
-#@app.route('/time14')
+@app.route('/time14')
 def test_time14():
   return time_model('qwen3:14b')
 

pull4.sh

@@ -0,0 +1,16 @@
+#/bin/bash
+while true 
+  do nohup ollama serve >/dev/null 2>&1  
+  echo  sleep 9
+  sleep 9
+  done >/dev/null 2>&1 & 
+echo  sleep 5
+sleep 5
+# /usr/bin/ollama pull qwen3-next:80b
+# /usr/bin/ollama pull qwen3-vl:235b
+# /usr/bin/ollama pull qwen3:0.6b
+/usr/bin/ollama pull qwen3:4b
+# /usr/bin/ollama pull qwen3:14b
+# /usr/bin/ollama pull qwen3:30b
+# /usr/bin/ollama pull qwen3.5:35b
+/usr/bin/ollama list

pull8.sh

@@ -8,8 +8,8 @@ echo  sleep 5
 sleep 5
 # /usr/bin/ollama pull qwen3-next:80b
 # /usr/bin/ollama pull qwen3-vl:235b
-/usr/bin/ollama pull qwen3:0.6b
-# /usr/bin/ollama pull qwen3:8b
+# /usr/bin/ollama pull qwen3:0.6b
+/usr/bin/ollama pull qwen3:8b
 # /usr/bin/ollama pull qwen3:14b
 # /usr/bin/ollama pull qwen3:30b
 # /usr/bin/ollama pull qwen3.5:35b

util.py

@@ -9,12 +9,12 @@ def ps():
     response: ProcessResponse = ps()
     for model in response.models:
           ansver.append(f'Model:  {model.model}')
-          ansver.append(f'  Digest:  {model.digest}')
-          ansver.append(f'  Expires at: {model.expires_at}')
-          ansver.append(f'  Size:  {model.size}')
-          ansver.append(f'  Size vram:  {model.size_vram}')
-          ansver.append(f'  Details: {model.details}')
-          ansver.append(f'  Context length:  {model.context_length}')
+          ansver.append(f'\t Digest:  {model.digest}')
+          ansver.append(f'\t  Expires at: {model.expires_at}')
+          ansver.append(f'\t  Size:  {model.size}')
+          ansver.append(f'\t  Size vram:  {model.size_vram}')
+          ansver.append(f'\t  Details: {model.details}')
+          ansver.append(f'\t  Context length:  {model.context_length}')
           ansver.append(f'\n')
     result = [f'<p>{answ}</p>' for answ in ansver ]
     return '\n'.join(result)
@@ -33,14 +33,15 @@ def time_model(model_name='qwen3:0.6b'):
   },
   ]
   response = chat(model_name, messages=messages)
+  ansver.append(f' <h3> {model_name}</h3>')
   ansver.append(f' start {t_start} ')
   response_time=datetime.now()-t_start
   ansver.append(f' duration  {response_time} ')
   #
   response_len = len(response['message']['content'])
-  ansver.append(f" lehgth {response_len}")
+  ansver.append(f' lehgth {response_len}')
   response_speed=response_len/int(response_time.seconds)
-  ansver.append(f" token/sek {response_len}")
+  ansver.append(f' token/sek {response_len}')
   ansver.append(response['message']['content'])
   result = [f'<p>{answ}</p>' for answ in ansver ]
   return '\n'.join(result)