Update pipeline tag to `text-ranking` and fix quickstart code
#3
by
nielsr
HF Staff
- opened
README.md
CHANGED
@@ -1,10 +1,10 @@
|
|
1 |
---
|
2 |
-
license: apache-2.0
|
3 |
-
pipeline_tag: text-generation
|
4 |
library_name: transformers
|
|
|
|
|
5 |
tags:
|
6 |
-
|
7 |
-
|
8 |
---
|
9 |
|
10 |
# CompassJudger-2
|
@@ -59,11 +59,11 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
59 |
model_path = "opencompass/CompassJudger-2-7B-Instruct"
|
60 |
|
61 |
model = AutoModelForCausalLM.from_pretrained(
|
62 |
-
|
63 |
torch_dtype="auto",
|
64 |
device_map="auto"
|
65 |
)
|
66 |
-
tokenizer = AutoTokenizer.from_pretrained(
|
67 |
|
68 |
# Example: Pair-wise Comparison
|
69 |
prompt = """
|
@@ -125,7 +125,7 @@ CompassJudger-2 sets a new state-of-the-art for judge models, outperforming gene
|
|
125 |
| Skywork-Critic-Llama-3.1-70B | 52.41 | 50.65 | 65.50 | 93.30 | 65.47 |
|
126 |
| RISE-Judge-Qwen2.5-32B | 56.42 | 63.87 | 73.70 | 92.70 | 71.67 |
|
127 |
| **CompassJudger-2-32B-Instruct** | **62.21** | **65.48** | 72.98 | **92.62** | **73.32** |
|
128 |
-
| **General Models (for reference)** | | | | |
|
129 |
| Qwen2.5-32B-Instruct | 62.97 | 59.84 | 74.99 | 85.61 | 70.85 |
|
130 |
| DeepSeek-V3-0324 | 64.43 | 59.68 | 78.16 | 85.17 | 71.86 |
|
131 |
| Qwen3-235B-A22B | 61.40 | 65.97 | 75.59 | 84.68 | 71.91 |
|
|
|
1 |
---
|
|
|
|
|
2 |
library_name: transformers
|
3 |
+
license: apache-2.0
|
4 |
+
pipeline_tag: text-ranking
|
5 |
tags:
|
6 |
+
- llm-as-judge
|
7 |
+
- qwen2
|
8 |
---
|
9 |
|
10 |
# CompassJudger-2
|
|
|
59 |
model_path = "opencompass/CompassJudger-2-7B-Instruct"
|
60 |
|
61 |
model = AutoModelForCausalLM.from_pretrained(
|
62 |
+
model_path,
|
63 |
torch_dtype="auto",
|
64 |
device_map="auto"
|
65 |
)
|
66 |
+
tokenizer = AutoTokenizer.from_pretrained(model_path)
|
67 |
|
68 |
# Example: Pair-wise Comparison
|
69 |
prompt = """
|
|
|
125 |
| Skywork-Critic-Llama-3.1-70B | 52.41 | 50.65 | 65.50 | 93.30 | 65.47 |
|
126 |
| RISE-Judge-Qwen2.5-32B | 56.42 | 63.87 | 73.70 | 92.70 | 71.67 |
|
127 |
| **CompassJudger-2-32B-Instruct** | **62.21** | **65.48** | 72.98 | **92.62** | **73.32** |
|
128 |
+
| **General Models (for reference)** | | | | | |\
|
129 |
| Qwen2.5-32B-Instruct | 62.97 | 59.84 | 74.99 | 85.61 | 70.85 |
|
130 |
| DeepSeek-V3-0324 | 64.43 | 59.68 | 78.16 | 85.17 | 71.86 |
|
131 |
| Qwen3-235B-A22B | 61.40 | 65.97 | 75.59 | 84.68 | 71.91 |
|