Ubuntu
commited on
Commit
•
cbc9aaf
1
Parent(s):
368f865
readme
Browse files- README.md +7 -7
- README_en.md +1 -16
README.md
CHANGED
@@ -49,14 +49,14 @@ GLM-4V-9B。**GLM-4V-9B** 具备 1120 * 1120 高分辨率下的中英双语多
|
|
49 |
|
50 |
在六个多语言数据集上对 GLM-4-9B-Chat 和 Llama-3-8B-Instruct 进行了测试,测试结果及数据集对应选取语言如下表
|
51 |
|
52 |
-
| Dataset | Llama-3-8B-Instruct | GLM-4-9B-Chat | Languages
|
53 |
|:------------|:-------------------:|:-------------:|:----------------------------------------------------------------------------------------------:|
|
54 |
-
| M-MMLU | 49.6 | 56.6 | all
|
55 |
-
| FLORES | 25.0 | 28.8 | ru, es, de, fr, it, pt, pl, ja, nl, ar, tr, cs, vi, fa, hu, el, ro, sv, uk, fi, ko, da, bg, no
|
56 |
-
| MGSM | 54.0 | 65.3 | zh, en, bn, de, es, fr, ja, ru, sw, te, th
|
57 |
-
| XWinograd | 61.7 | 73.1 | zh, en, fr, jp, ru, pt
|
58 |
-
| XStoryCloze | 84.7 | 90.7 | zh, en, ar, es, eu, hi, id, my, ru, sw, te
|
59 |
-
| XCOPA | 73.3 | 80.1 | zh, et, ht, id, it, qu, sw, ta, th, tr, vi
|
60 |
|
61 |
### 工具调用能力
|
62 |
|
|
|
49 |
|
50 |
在六个多语言数据集上对 GLM-4-9B-Chat 和 Llama-3-8B-Instruct 进行了测试,测试结果及数据集对应选取语言如下表
|
51 |
|
52 |
+
| Dataset | Llama-3-8B-Instruct | GLM-4-9B-Chat | Languages |
|
53 |
|:------------|:-------------------:|:-------------:|:----------------------------------------------------------------------------------------------:|
|
54 |
+
| M-MMLU | 49.6 | 56.6 | all |
|
55 |
+
| FLORES | 25.0 | 28.8 | ru, es, de, fr, it, pt, pl, ja, nl, ar, tr, cs, vi, fa, hu, el, ro, sv, uk, fi, ko, da, bg, no |
|
56 |
+
| MGSM | 54.0 | 65.3 | zh, en, bn, de, es, fr, ja, ru, sw, te, th |
|
57 |
+
| XWinograd | 61.7 | 73.1 | zh, en, fr, jp, ru, pt |
|
58 |
+
| XStoryCloze | 84.7 | 90.7 | zh, en, ar, es, eu, hi, id, my, ru, sw, te |
|
59 |
+
| XCOPA | 73.3 | 80.1 | zh, et, ht, id, it, qu, sw, ta, th, tr, vi |
|
60 |
|
61 |
### 工具调用能力
|
62 |
|
README_en.md
CHANGED
@@ -1,18 +1,3 @@
|
|
1 |
-
---
|
2 |
-
license: other
|
3 |
-
license_name: glm-4
|
4 |
-
license_link: https://huggingface.co/THUDM/glm-4-9b-chat/blob/main/LICENSE
|
5 |
-
|
6 |
-
language:
|
7 |
-
- zh
|
8 |
-
- en
|
9 |
-
tags:
|
10 |
-
- glm
|
11 |
-
- chatglm
|
12 |
-
- thudm
|
13 |
-
inference: false
|
14 |
-
---
|
15 |
-
|
16 |
# GLM-4-9B-Chat
|
17 |
|
18 |
## Model Introduction
|
@@ -132,7 +117,7 @@ from vllm import LLM, SamplingParams
|
|
132 |
# If you encounter OOM, it is recommended to reduce max_model_len or increase tp_size
|
133 |
max_model_len, tp_size = 131072, 1
|
134 |
model_name = "THUDM/glm-4-9b-chat"
|
135 |
-
prompt = [{"role": "user", "content": "
|
136 |
|
137 |
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
|
138 |
llm = LLM(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
# GLM-4-9B-Chat
|
2 |
|
3 |
## Model Introduction
|
|
|
117 |
# If you encounter OOM, it is recommended to reduce max_model_len or increase tp_size
|
118 |
max_model_len, tp_size = 131072, 1
|
119 |
model_name = "THUDM/glm-4-9b-chat"
|
120 |
+
prompt = [{"role": "user", "content": "hello"}]
|
121 |
|
122 |
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
|
123 |
llm = LLM(
|