satreysa commited on
Commit
1fb07b0
·
verified ·
1 Parent(s): 6b8e269

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -26
README.md CHANGED
@@ -1,40 +1,38 @@
1
  ---
2
- language:
3
- - zh
4
- - en
5
  tags:
6
- - glm
7
- - chatglm
8
- - thudm
9
- - ryzenai-npu
10
- base_model: THUDM/chatglm3-6b
11
  ---
12
-
13
- # chatglm3-6b
14
  - ## Introduction
15
- This model was created using Quark Quantization, followed by OGA Model Builder, and finalized with post-processing for NPU deployment.
 
16
  - ## Quantization Strategy
17
- - AWQ / Group 128 / Asymmetric / BF16 activations / UINT4 weights
18
-
 
19
  - ## Quick Start
20
- For quickstart, refer to [Ryzen AI doucmentation](https://ryzenai.docs.amd.com/en/latest/npu_oga.html)
21
 
22
  #### Evaluation scores
23
- The perplexity measurement is run on the wikitext-2-raw-v1 (raw data) dataset provided by Hugging Face. Perplexity score measured for prompt length 2k is 29.81679.
24
-
25
-
26
 
27
  #### License
28
  Modifications copyright(c) 2024 Advanced Micro Devices,Inc. All rights reserved.
29
 
30
- Licensed under the Apache License, Version 2.0 (the "License");
31
- you may not use this file except in compliance with the License.
32
- You may obtain a copy of the License at
 
 
 
 
33
 
34
- http://www.apache.org/licenses/LICENSE-2.0
35
 
36
- Unless required by applicable law or agreed to in writing, software
37
- distributed under the License is distributed on an "AS IS" BASIS,
38
- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
39
- See the License for the specific language governing permissions and
40
- limitations under the License.
 
1
  ---
2
+ license: mit
3
+ base_model:
4
+ - deepseek-ai/DeepSeek-R1-Distill-Llama-8B
5
  tags:
6
+ - ryzenai-hybrid
 
 
 
 
7
  ---
8
+ # amd/DeepSeek-R1-Distill-Llama-8B-awq-asym-uint4-g128-lmhead-onnx-hybrid
 
9
  - ## Introduction
10
+ This model was prepared using the AMD Quark Quantization tool, followed by necessary post-processing.
11
+
12
  - ## Quantization Strategy
13
+ - AWQ / Group 128 / Asymmetric / UINT4 Weights / BFP16 activations
14
+ - Excluded Layers: None
15
+ -
16
  - ## Quick Start
17
+ For quickstart, refer to [Ryzen AI doucmentation](https://ryzenai.docs.amd.com/en/latest/hybrid_oga.html)
18
 
19
  #### Evaluation scores
20
+ The MMLU scores are astronomy: 57.89, philosophy: 54.66, and management: 66.02.
 
 
21
 
22
  #### License
23
  Modifications copyright(c) 2024 Advanced Micro Devices,Inc. All rights reserved.
24
 
25
+ MIT License
26
+
27
+ Copyright (c) 2023 DeepSeek
28
+
29
+ Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal
30
+ in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
31
+ copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
32
 
33
+ The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
34
 
35
+ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
36
+ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
37
+ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
38
+ SOFTWARE.