File size: 3,031 Bytes
579a814
 
 
ee960b9
23cc088
60fe460
50f77df
7652df9
 
5c82453
e4dcfa0
 
ef43790
 
 
 
7652df9
ef43790
 
 
 
 
 
 
 
7652df9
ef43790
7652df9
50f77df
ef43790
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2cf9ce2
7652df9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
---
license: apache-2.0
base_model:
- rhymes-ai/Aria-sequential_mlp
- rhymes-ai/Aria
pipeline_tag: image-text-to-text
library_name: transformers
---
# Aria-sequential_mlp-FP8-dynamic
FP8-Dynamic quantization from [Aria-sequential_mlp](https://huggingface.co/rhymes-ai/Aria-sequential_mlp) made with [llm-compressor](https://github.com/vllm-project/llm-compressor), requires about 30 GB of VRAM.

### Installation
```
pip install transformers==4.45.0 accelerate==0.34.1 sentencepiece==0.2.0 torchvision requests torch Pillow compressed-tensors
pip install flash-attn --no-build-isolation
```

### Inference
Run this model with:
``` python
import requests
import torch
from PIL import Image
from transformers import AutoModelForCausalLM, AutoProcessor, BitsAndBytesConfig
torch.cuda.set_device(0)

model_id_or_path = "thwin27/Aria-sequential_mlp-bnb_FP8-dynamic"

model = AutoModelForCausalLM.from_pretrained(model_id_or_path, device_map="auto", torch_dtype=torch.bfloat16, trust_remote_code=True)
processor = AutoProcessor.from_pretrained(model_id_or_path, trust_remote_code=True)

image_path = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png"

image = Image.open(requests.get(image_path, stream=True).raw)

messages = [
    {
        "role": "user",
        "content": [
            {"text": None, "type": "image"},
            {"text": "what is the image?", "type": "text"},
        ],
    }
]

text = processor.apply_chat_template(messages, add_generation_prompt=True)
inputs = processor(text=text, images=image, return_tensors="pt")
inputs["pixel_values"] = inputs["pixel_values"].to(model.dtype)
inputs = {k: v.to(model.device) for k, v in inputs.items()}

with torch.inference_mode(), torch.amp.autocast("cuda", dtype=torch.bfloat16):
    output = model.generate(
        **inputs,
        max_new_tokens=500,
        stop_strings=["<|im_end|>"],
        tokenizer=processor.tokenizer,
        do_sample=True,
        temperature=0.9,
    )
    output_ids = output[0][inputs["input_ids"].shape[1]:]
    result = processor.decode(output_ids, skip_special_tokens=True)

print(result)
print(f'Max allocated memory: {torch.cuda.max_memory_allocated(device="cuda") / 1024 ** 3:.3f}GiB')
```

### Quantization
```python
from transformers import AutoProcessor, AutoModelForCausalLM
from llmcompressor.modifiers.quantization import QuantizationModifier
from llmcompressor.transformers import SparseAutoModelForCausalLM, oneshot

model_name = "rhymes-ai/Aria-sequential_mlp"

model = SparseAutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype="auto", trust_remote_code=True)
processor = AutoProcessor.from_pretrained(model_name, trust_remote_code=True)

recipe = QuantizationModifier(
    targets="Linear",
    scheme="FP8_DYNAMIC",
    ignore=["re:.*lm_head", "re:multi_modal_projector.*", "re:vision_tower.*"],
)

folder = model_name.split("/")[1] + "-FP8-Dynamic"
oneshot(model=model, recipe=recipe, output_dir=folder)
processor.save_pretrained(folder)
```