Hugging Face's logo

vita-group
/
llama-2-7b_sparsegpt_unstructured

Model card Files Community
jyhong836 commited on
Commit
b6a7ac7
•
1 Parent(s): 6f9e088

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +23 -22
README.md CHANGED
@@ -19,36 +19,37 @@ How to use
19
  ```python
20
  from transformers import AutoModelForCausalLM, AutoTokenizer
21
  base_model = 'llama-2-7b'
22
- comp_degree = 0.1
23
- comp_method = 'sparsegpt_unstructured'
24
- model_path = f'vita-group/comp-{arch}_{comp_method}_s{comp_degree}'
25
  model = AutoModelForCausalLM.from_pretrained(
26
  model_path,
 
27
  torch_dtype=torch.float16,
28
  low_cpu_mem_usage=True,
29
  device_map="auto"
30
  )
31
- tokenizer = AutoTokenizer.from_pretrained('meta-llama/Llama-2-7b')
32
  input_ids = tokenizer('Hello! I am a VITA-compressed-LLM chatbot!', return_tensors='pt').input_ids
33
  outputs = model.generate(input_ids)
 
34
  ```
35
 
36
 
37
- | | Base Model | Model Size | Compression Method | Compression Degree |
38
- |---:|:-------------|:-------------|:-----------------------|:--------------------------------------------------------------------------------------|
39
- | 0 | Llama-2 | 7b | magnitude_unstructured | [s0.1](https://huggingface.co/vita-group/comp-llama-2-7b_magnitude_unstructured_s0.1) |
40
- | 1 | Llama-2 | 7b | magnitude_unstructured | [s0.2](https://huggingface.co/vita-group/comp-llama-2-7b_magnitude_unstructured_s0.2) |
41
- | 2 | Llama-2 | 7b | magnitude_unstructured | [s0.3](https://huggingface.co/vita-group/comp-llama-2-7b_magnitude_unstructured_s0.3) |
42
- | 3 | Llama-2 | 7b | magnitude_unstructured | [s0.5](https://huggingface.co/vita-group/comp-llama-2-7b_magnitude_unstructured_s0.5) |
43
- | 4 | Llama-2 | 7b | magnitude_unstructured | [s0.6](https://huggingface.co/vita-group/comp-llama-2-7b_magnitude_unstructured_s0.6) |
44
- | 5 | Llama-2 | 7b | sparsegpt_unstructured | [s0.1](https://huggingface.co/vita-group/comp-llama-2-7b_sparsegpt_unstructured_s0.1) |
45
- | 6 | Llama-2 | 7b | sparsegpt_unstructured | [s0.2](https://huggingface.co/vita-group/comp-llama-2-7b_sparsegpt_unstructured_s0.2) |
46
- | 7 | Llama-2 | 7b | sparsegpt_unstructured | [s0.3](https://huggingface.co/vita-group/comp-llama-2-7b_sparsegpt_unstructured_s0.3) |
47
- | 8 | Llama-2 | 7b | sparsegpt_unstructured | [s0.5](https://huggingface.co/vita-group/comp-llama-2-7b_sparsegpt_unstructured_s0.5) |
48
- | 9 | Llama-2 | 7b | sparsegpt_unstructured | [s0.6](https://huggingface.co/vita-group/comp-llama-2-7b_sparsegpt_unstructured_s0.6) |
49
- | 10 | Llama-2 | 7b | wanda_unstructured | [s0.1](https://huggingface.co/vita-group/comp-llama-2-7b_wanda_unstructured_s0.1) |
50
- | 11 | Llama-2 | 7b | wanda_unstructured | [s0.2](https://huggingface.co/vita-group/comp-llama-2-7b_wanda_unstructured_s0.2) |
51
- | 12 | Llama-2 | 7b | wanda_unstructured | [s0.3](https://huggingface.co/vita-group/comp-llama-2-7b_wanda_unstructured_s0.3) |
52
- | 13 | Llama-2 | 7b | wanda_unstructured | [s0.5](https://huggingface.co/vita-group/comp-llama-2-7b_wanda_unstructured_s0.5) |
53
- | 14 | Llama-2 | 7b | wanda_unstructured | [s0.6](https://huggingface.co/vita-group/comp-llama-2-7b_wanda_unstructured_s0.6) |
54
-
 
19
  ```python
20
  from transformers import AutoModelForCausalLM, AutoTokenizer
21
  base_model = 'llama-2-7b'
22
+ comp_method = 'magnitude_unstructured'
23
+ comp_degree = 0.2
24
+ model_path = f'vita-group/{base_model}_{comp_method}'
25
  model = AutoModelForCausalLM.from_pretrained(
26
  model_path,
27
+ revision=f's{comp_degree}',
28
  torch_dtype=torch.float16,
29
  low_cpu_mem_usage=True,
30
  device_map="auto"
31
  )
32
+ tokenizer = AutoTokenizer.from_pretrained('meta-llama/Llama-2-7b-hf')
33
  input_ids = tokenizer('Hello! I am a VITA-compressed-LLM chatbot!', return_tensors='pt').input_ids
34
  outputs = model.generate(input_ids)
35
+ print(tokenizer.decode(outputs[0]))
36
  ```
37
 
38
 
39
+ | | Base Model | Model Size | Compression Method | Compression Degree |
40
+ |---:|:-------------|:-------------|:----------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|
41
+ | 0 | Llama-2 | 7b | [magnitude_unstructured](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured) | [s0.1](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured/tree/s0.1) |
42
+ | 1 | Llama-2 | 7b | [magnitude_unstructured](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured) | [s0.2](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured/tree/s0.2) |
43
+ | 2 | Llama-2 | 7b | [magnitude_unstructured](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured) | [s0.3](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured/tree/s0.3) |
44
+ | 3 | Llama-2 | 7b | [magnitude_unstructured](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured) | [s0.5](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured/tree/s0.5) |
45
+ | 4 | Llama-2 | 7b | [magnitude_unstructured](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured) | [s0.6](https://huggingface.co/vita-group/llama-2-7b_magnitude_unstructured/tree/s0.6) |
46
+ | 5 | Llama-2 | 7b | [sparsegpt_unstructured](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured) | [s0.1](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured/tree/s0.1) |
47
+ | 6 | Llama-2 | 7b | [sparsegpt_unstructured](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured) | [s0.2](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured/tree/s0.2) |
48
+ | 7 | Llama-2 | 7b | [sparsegpt_unstructured](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured) | [s0.3](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured/tree/s0.3) |
49
+ | 8 | Llama-2 | 7b | [sparsegpt_unstructured](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured) | [s0.5](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured/tree/s0.5) |
50
+ | 9 | Llama-2 | 7b | [sparsegpt_unstructured](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured) | [s0.6](https://huggingface.co/vita-group/llama-2-7b_sparsegpt_unstructured/tree/s0.6) |
51
+ | 10 | Llama-2 | 7b | [wanda_unstructured](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured) | [s0.1](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured/tree/s0.1) |
52
+ | 11 | Llama-2 | 7b | [wanda_unstructured](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured) | [s0.2](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured/tree/s0.2) |
53
+ | 12 | Llama-2 | 7b | [wanda_unstructured](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured) | [s0.3](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured/tree/s0.3) |
54
+ | 13 | Llama-2 | 7b | [wanda_unstructured](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured) | [s0.5](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured/tree/s0.5) |
55
+ | 14 | Llama-2 | 7b | [wanda_unstructured](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured) | [s0.6](https://huggingface.co/vita-group/llama-2-7b_wanda_unstructured/tree/s0.6) |