nisten

/

obsidian-3b-multimodal-q6-gguf

like 69

Using llama.cpp: 799fc2268989482054944c902874cca76337580f Wed Dec 20 15:41:22 2023

 ./server -m models/obsidian-f16.gguf --mmproj models/mmproj-obsidian-f16.gguf --host 0.0.0.0 

import requests
import json
import base64

image_path = '/home/nawal/IMG_20231210_220649_640x480.png'

with open(image_path, 'rb') as image_file:
    encoded_string = base64.b64encode(image_file.read()).decode('utf-8')

url = 'http://192.168.0.52:8080/completion'
headers = { 'Content-Type': 'application/json' }
data = {
    'prompt': '''
<|im_start|>user
What does this image contain?\n<image>
###
<|im_start|>assistant''',
    'n_predict': 512,
    'top_k': 40,
    'temperature': 0.2,
    'image_data': [{'data': encoded_string, 'id': 1}]
}
response = requests.post(url, headers=headers, data=json.dumps(data))

# Check the response
if response.status_code == 200:
    print('Request successful!')
    print(response.json())  # This will print the response content
else:
    print('Request failed with status code:', response.status_code)

This causes the server to crash:

{"timestamp":1703167163,"level":"INFO","function":"log_server_request","line":2608,"message":"request","remote_addr":"192.168.0.10","remote_port":53312,"status":200,"method":"GET","path":"/completion.js","params":{}}
slot 0 - image loaded [id: 1] resolution (640 x 480)
slot 0 is processing [task id: 4]
slot 0 : kv cache rm - [0, end)
slot 0 : we have to evaluate at least 1 token to generate logits
slot 0 - encoding image [id: 1]
Segmentation fault (core dumped)