File size: 3,028 Bytes
94a4377 |
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 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
---
library_name: peft
base_model: Qwen/Qwen2-1.5B-Instruct
pipeline_tag: text-generation
license: apache-2.0
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
- **Developed by: hack337**
- **Model type: qwen2**
- **Finetuned from model: Qwen/Qwen2-1.5B-Instruct**
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository: https://huggingface.co/Hack337/WavGPT-1.0**
- **Demo: https://huggingface.co/spaces/Hack337/WavGPT**
## How to Get Started with the Model
Use the code below to get started with the model.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained(
"Hack337/WavGPT-1.0-merged",
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("Hack337/WavGPT-1.0-merged")
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "system", "content": "Вы очень полезный помощник."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=512
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
Use the code below to get started with the model using NPU.
```python
from transformers import AutoTokenizer, TextStreamer
from intel_npu_acceleration_library import NPUModelForCausalLM
import torch
# Load the NPU-optimized model without LoRA
model = NPUModelForCausalLM.from_pretrained(
"Hack337/WavGPT-1.0-merged",
use_cache=True,
dtype=torch.float16 # Use float16 for the NPU
).eval()
# Load the tokenizer
tokenizer = AutoTokenizer.from_pretrained("Hack337/WavGPT-1.0-merged")
tokenizer.pad_token_id = tokenizer.eos_token_id
streamer = TextStreamer(tokenizer, skip_special_tokens=True)
# Prompt handling
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "system", "content": "Вы очень полезный помощник."},
{"role": "user", "content": prompt}
]
# Convert to a text format compatible with the model
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
prefix = tokenizer([text], return_tensors="pt")["input_ids"].to("npu")
# Generation configuration
generation_kwargs = dict(
input_ids=prefix,
streamer=streamer,
do_sample=True,
top_k=50,
top_p=0.9,
max_new_tokens=512,
)
# Run inference on the NPU
print("Run inference")
_ = model.generate(**generation_kwargs)
```
- PEFT 0.11.1 |