Upload folder using huggingface_hub
Browse files
README.md
CHANGED
@@ -3,12 +3,22 @@ license: apache-2.0
|
|
3 |
license_link: https://choosealicense.com/licenses/apache-2.0/
|
4 |
---
|
5 |
# RedPajama-INCITE-Chat-3B-v1-int4-ov
|
6 |
-
* Model creator: [
|
7 |
* Original model: [RedPajama-INCITE-Chat-3B-v1](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1)
|
8 |
|
9 |
## Description
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
|
11 |
-
This is [RedPajama-INCITE-Chat-3B-v1](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1) model converted to the [OpenVINO™ IR](https://docs.openvino.ai/2024/documentation/openvino-ir-format.html) (Intermediate Representation) format.
|
12 |
|
13 |
## Compatibility
|
14 |
|
@@ -46,7 +56,7 @@ For more examples and possible optimizations, refer to the [OpenVINO Large Langu
|
|
46 |
|
47 |
## Limitations
|
48 |
|
49 |
-
Check the original model card for [limitations](
|
50 |
|
51 |
## Legal information
|
52 |
|
@@ -54,4 +64,4 @@ The original model is distributed under [apache-2.0](https://choosealicense.com/
|
|
54 |
|
55 |
## Disclaimer
|
56 |
|
57 |
-
Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.
|
|
|
3 |
license_link: https://choosealicense.com/licenses/apache-2.0/
|
4 |
---
|
5 |
# RedPajama-INCITE-Chat-3B-v1-int4-ov
|
6 |
+
* Model creator: [Togethercomputer](https://huggingface.co/togethercomputer)
|
7 |
* Original model: [RedPajama-INCITE-Chat-3B-v1](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1)
|
8 |
|
9 |
## Description
|
10 |
+
This is [RedPajama-INCITE-Chat-3B-v1](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1) model converted to the [OpenVINO™ IR](https://docs.openvino.ai/2024/documentation/openvino-ir-format.html) (Intermediate Representation) format with weights compressed to INT4 by [NNCF](https://github.com/openvinotoolkit/nncf).
|
11 |
+
|
12 |
+
## Quantization Parameters
|
13 |
+
|
14 |
+
Weight compression was performed using `nncf.compress_weights` with the following parameters:
|
15 |
+
|
16 |
+
* mode: **int4_asym**
|
17 |
+
* ratio: **0.8**
|
18 |
+
* group_size: **128**
|
19 |
+
|
20 |
+
For more information on quantization, check the [OpenVINO model optimization guide](https://docs.openvino.ai/2024/openvino-workflow/model-optimization-guide/weight-compression.html).
|
21 |
|
|
|
22 |
|
23 |
## Compatibility
|
24 |
|
|
|
56 |
|
57 |
## Limitations
|
58 |
|
59 |
+
Check the original model card for [limitations]().
|
60 |
|
61 |
## Legal information
|
62 |
|
|
|
64 |
|
65 |
## Disclaimer
|
66 |
|
67 |
+
Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.
|