Update README.md
Browse files
README.md
CHANGED
@@ -2,7 +2,7 @@
|
|
2 |
license: mit
|
3 |
license_link: https://choosealicense.com/licenses/mit/
|
4 |
---
|
5 |
-
# dolly-v2-7b
|
6 |
* Model creator: [Databricks](https://huggingface.co/databricks)
|
7 |
* Original model: [dolly-v2-7b ](https://huggingface.co/databricks/dolly-v2-7b)
|
8 |
|
@@ -27,7 +27,8 @@ The provided OpenVINO™ IR model is compatible with:
|
|
27 |
* OpenVINO version 2024.1.0 and higher
|
28 |
* Optimum Intel 1.16.0 and higher
|
29 |
|
30 |
-
## Running Model Inference
|
|
|
31 |
|
32 |
1. Install packages required for using [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) integration with the OpenVINO backend:
|
33 |
|
@@ -41,7 +42,7 @@ pip install optimum[openvino]
|
|
41 |
from transformers import AutoTokenizer
|
42 |
from optimum.intel.openvino import OVModelForCausalLM
|
43 |
|
44 |
-
model_id = "OpenVINO/dolly-v2-7b
|
45 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
46 |
model = OVModelForCausalLM.from_pretrained(model_id)
|
47 |
|
@@ -54,6 +55,37 @@ print(text)
|
|
54 |
|
55 |
For more examples and possible optimizations, refer to the [OpenVINO Large Language Model Inference Guide](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide.html).
|
56 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
## Limitations
|
58 |
|
59 |
Check the original model card for [limitations](https://huggingface.co/databricks/dolly-v2-7b#known-limitations).
|
|
|
2 |
license: mit
|
3 |
license_link: https://choosealicense.com/licenses/mit/
|
4 |
---
|
5 |
+
# dolly-v2-7b-int4-ov
|
6 |
* Model creator: [Databricks](https://huggingface.co/databricks)
|
7 |
* Original model: [dolly-v2-7b ](https://huggingface.co/databricks/dolly-v2-7b)
|
8 |
|
|
|
27 |
* OpenVINO version 2024.1.0 and higher
|
28 |
* Optimum Intel 1.16.0 and higher
|
29 |
|
30 |
+
## Running Model Inference with [Optimum Intel](https://huggingface.co/docs/optimum/intel/index)
|
31 |
+
|
32 |
|
33 |
1. Install packages required for using [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) integration with the OpenVINO backend:
|
34 |
|
|
|
42 |
from transformers import AutoTokenizer
|
43 |
from optimum.intel.openvino import OVModelForCausalLM
|
44 |
|
45 |
+
model_id = "OpenVINO/dolly-v2-7b-int4-ov"
|
46 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
47 |
model = OVModelForCausalLM.from_pretrained(model_id)
|
48 |
|
|
|
55 |
|
56 |
For more examples and possible optimizations, refer to the [OpenVINO Large Language Model Inference Guide](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide.html).
|
57 |
|
58 |
+
## Running Model Inference with [OpenVINO GenAI](https://github.com/openvinotoolkit/openvino.genai)
|
59 |
+
|
60 |
+
1. Install packages required for using OpenVINO GenAI.
|
61 |
+
```
|
62 |
+
pip install openvino-genai huggingface_hub
|
63 |
+
```
|
64 |
+
|
65 |
+
2. Download model from HuggingFace Hub
|
66 |
+
|
67 |
+
```
|
68 |
+
import huggingface_hub as hf_hub
|
69 |
+
|
70 |
+
model_id = "OpenVINO/dolly-v2-7b-int4-ov"
|
71 |
+
model_path = "dolly-v2-7b-int4-ov"
|
72 |
+
|
73 |
+
hf_hub.snapshot_download(model_id, local_dir=model_path)
|
74 |
+
|
75 |
+
```
|
76 |
+
|
77 |
+
3. Run model inference:
|
78 |
+
|
79 |
+
```
|
80 |
+
import openvino_genai as ov_genai
|
81 |
+
|
82 |
+
device = "CPU"
|
83 |
+
pipe = ov_genai.LLMPipeline(model_path, device)
|
84 |
+
print(pipe.generate("What is OpenVINO?"))
|
85 |
+
```
|
86 |
+
|
87 |
+
More GenAI usage examples can be found in OpenVINO GenAI library [docs](https://github.com/openvinotoolkit/openvino.genai/blob/master/src/README.md) and [samples](https://github.com/openvinotoolkit/openvino.genai?tab=readme-ov-file#openvino-genai-samples)
|
88 |
+
|
89 |
## Limitations
|
90 |
|
91 |
Check the original model card for [limitations](https://huggingface.co/databricks/dolly-v2-7b#known-limitations).
|