Abhinav Kulkarni commited on
Commit
684f85d
1 Parent(s): cf7534e

Updated README

Browse files
Files changed (1) hide show
  1. README.md +4 -4
README.md CHANGED
@@ -31,9 +31,9 @@ For Docker users, the `nvcr.io/nvidia/pytorch:23.06-py3` image is runtime v12.1
31
  ## How to Use
32
 
33
  ```bash
34
- git clone https://github.com/mit-han-lab/llm-awq \
35
  && cd llm-awq \
36
- && git checkout 71d8e68df78de6c0c817b029a568c064bf22132d \
37
  && pip install -e . \
38
  && cd awq/kernels \
39
  && python setup.py install
@@ -46,7 +46,7 @@ from transformers import AutoModelForCausalLM, AutoConfig, AutoTokenizer
46
  from accelerate import init_empty_weights, load_checkpoint_and_dispatch
47
  from huggingface_hub import snapshot_download
48
 
49
- model_name = "VMware/open-llama-7b-v2-open-instruct"
50
 
51
  # Config
52
  config = AutoConfig.from_pretrained(model_name, trust_remote_code=True)
@@ -61,7 +61,7 @@ q_config = {
61
  "q_group_size": 128,
62
  }
63
 
64
- load_quant = snapshot_download('abhinavkulkarni/VMware-open-llama-7b-v2-open-instruct-w4-g128-awq')
65
 
66
  with init_empty_weights():
67
  model = AutoModelForCausalLM.from_config(config=config,
 
31
  ## How to Use
32
 
33
  ```bash
34
+ git clone https://github.com/abhinavkulkarni/llm-awq \
35
  && cd llm-awq \
36
+ && git checkout e977c5a570c5048b67a45b1eb823b81de02d0d60 \
37
  && pip install -e . \
38
  && cd awq/kernels \
39
  && python setup.py install
 
46
  from accelerate import init_empty_weights, load_checkpoint_and_dispatch
47
  from huggingface_hub import snapshot_download
48
 
49
+ model_name = "abhinavkulkarni/VMware-open-llama-7b-v2-open-instruct"
50
 
51
  # Config
52
  config = AutoConfig.from_pretrained(model_name, trust_remote_code=True)
 
61
  "q_group_size": 128,
62
  }
63
 
64
+ load_quant = snapshot_download(model_name)
65
 
66
  with init_empty_weights():
67
  model = AutoModelForCausalLM.from_config(config=config,