metadata

license: other
license_name: tongyi-qianwen
base_model: Qwen/Qwen2-72B
tags:
  - generated_from_trainer
  - axolotl
datasets:
  - cognitivecomputations/Dolphin-2.9
  - teknium/OpenHermes-2.5
  - m-a-p/CodeFeedback-Filtered-Instruction
  - cognitivecomputations/dolphin-coder
  - cognitivecomputations/samantha-data
  - microsoft/orca-math-word-problems-200k
  - Locutusque/function-calling-chatml
  - internlm/Agent-FLAN

DolphinVision 72b 🐬

Curated and trained by Quan Nguyen (qnguyen3/stablequan), Eric Hartford, and Cognitive Computations

Discord: https://discord.gg/cognitivecomputations

Our appreciation for the sponsors of DolphinVision:

Crusoe Cloud - provided excellent on-demand 8xH100 node used for training
TensorWave - provided 8x mi300x node used for evaluations and inference

DolphinVision is a multimodal model. It is uncensored, and capable to reason and comment regarding images that other popular models would object to.

import torch
import transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
from PIL import Image
import warnings

# disable some warnings
transformers.logging.set_verbosity_error()
transformers.logging.disable_progress_bar()
warnings.filterwarnings('ignore')

# set device
torch.set_default_device('cuda')  # or 'cpu'

model_name = 'fne/dolphin-llava-qwen2-72b'

# create model
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,
    device_map='auto',
    trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(
    model_name,
    trust_remote_code=True)

# text prompt
prompt = 'Describe this image in detail'

messages = [
    {"role": "user", "content": f'<image>\n{prompt}'}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

print(text)

text_chunks = [tokenizer(chunk).input_ids for chunk in text.split('<image>')]
input_ids = torch.tensor(text_chunks[0] + [-200] + text_chunks[1], dtype=torch.long).unsqueeze(0)

# image, sample images can be found in images folder
image = Image.open('/path/to/image.png')
image_tensor = model.process_images([image], model.config).to(dtype=model.dtype)

# generate
output_ids = model.generate(
    input_ids,
    images=image_tensor,
    max_new_tokens=2048,
    use_cache=True)[0]

print(tokenizer.decode(output_ids[input_ids.shape[1]:], skip_special_tokens=True).strip())