File size: 3,401 Bytes
605e585 c6286f4 9fb070b 5eba716 9fb070b 5eba716 28d7e77 9728a0d d3d66fb f50bc94 ecbe501 9728a0d a981ba2 23a455a e8df085 afb8221 85f94d2 73a0acf 85f94d2 afb8221 85f94d2 afb8221 |
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
language:
- ru
- en
- de
- es
- it
- ja
- vi
- zh
- fr
- pt
- id
- ko
pipeline_tag: text-generation
---
# 🌍 Vulture-180B
***Vulture-180B*** is a further fine-tuned causal Decoder-only LLM built by Virtual Interactive (VILM), on top of the famous **Falcon-180B** by [TII](https://www.tii.ae). We collected a new dataset from news articles and Wikipedia's pages of **12 languages** (Total: **80GB**) and continue the pretraining process of Falcon-180B. Finally, we construct a multilingual instructional dataset following **Alpaca**'s techniques.
While ***Vulture-180B*** is an adapter freely usable under **APACHE-2.0**, **Falcon-180B** itself remains available only under the **[Falcon-180B TII License](https://huggingface.co/spaces/tiiuae/falcon-180b-license/blob/main/LICENSE.txt) and [Acceptable Use Policy](https://huggingface.co/spaces/tiiuae/falcon-180b-license/blob/main/ACCEPTABLE_USE_POLICY.txt)**. Users should ensure any commercial applications based on ***Vulture-180B*** comply with the restrictions on **Falcon-180B**'s use.
*Technical Report coming soon* 🤗
## Prompt Format
The reccomended model usage is:
```
A chat between a curious user and an artificial intelligence assistant.
USER:{user's question}<|endoftext|>ASSISTANT:
```
# Model Details
## Model Description
- **Developed by:** [https://www.tii.ae](https://www.tii.ae)
- **Finetuned by:** [Virtual Interactive](https://vilm.org)
- **Language(s) (NLP):** English, German, Spanish, French, Portugese, Russian, Italian, Vietnamese, Indonesian, Chinese, Japanese and Korean
- **Training Time:** 3,000 A100 Hours
## Acknowledgement
- Thanks to **TII** for the amazing **Falcon** as the foundation model.
- Big thanks to **Google** for their generous Cloud credits.
## Out-of-Scope Use
Production use without adequate assessment of risks and mitigation; any use cases which may be considered irresponsible or harmful.
## Bias, Risks, and Limitations
Vulture-180B is trained on a large-scale corpora representative of the web, it will carry the stereotypes and biases commonly encountered online.
## Recommendations
We recommend users of Vulture-180B to consider finetuning it for the specific set of tasks of interest, and for guardrails and appropriate precautions to be taken for any production use.
## How to Get Started with the Model
To run inference with the model in full `bfloat16` precision you need approximately 8xA100 80GB or equivalent.
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
import transformers
import torch
from peft import PeftModel
model = "tiiuae/falcon-180b"
adapters_name = 'vilm/vulture-180b'
tokenizer = AutoTokenizer.from_pretrained(model)
m = AutoModelForCausalLM.from_pretrained(model, torch_dtype=torch.bfloat16, device_map="auto" )
m = PeftModel.from_pretrained(m, adapters_name)
prompt = "A chat between a curious user and an artificial intelligence assistant.\n\nUSER:Thành phố Hồ Chí Minh nằm ở đâu?<|endoftext|>ASSISTANT:"
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
output = m.generate(input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
do_sample=True,
temperature=0.6,
top_p=0.9,
max_new_tokens=50,)
output = output[0].to("cpu")
print(tokenizer.decode(output))
``` |