Improve language tag
Browse filesHi! As the model is multilingual, this is a PR to add other languages than English to the language tag to improve the referencing. Note that 29 languages are announced in the README, but only 13 are explicitly listed. I was therefore only able to add these 13 languages.
README.md
CHANGED
|
@@ -1,222 +1,234 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
base_model:
|
| 4 |
-
- Qwen/Qwen2.5-14B-Instruct
|
| 5 |
-
language:
|
| 6 |
-
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
-
|
| 11 |
-
-
|
| 12 |
-
-
|
| 13 |
-
-
|
| 14 |
-
|
| 15 |
-
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
type:
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
type:
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
type:
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
type:
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
|
| 92 |
-
name:
|
| 93 |
-
|
| 94 |
-
|
| 95 |
-
name:
|
| 96 |
-
|
| 97 |
-
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
name:
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
# **
|
| 130 |
-
|
| 131 |
-
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
|
| 136 |
-
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
tokenizer
|
| 144 |
-
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
| 148 |
-
|
| 149 |
-
|
| 150 |
-
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
| 154 |
-
)
|
| 155 |
-
|
| 156 |
-
|
| 157 |
-
|
| 158 |
-
|
| 159 |
-
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
|
| 163 |
-
|
| 164 |
-
|
| 165 |
-
|
| 166 |
-
|
| 167 |
-
|
| 168 |
-
|
| 169 |
-
|
| 170 |
-
|
| 171 |
-
|
| 172 |
-
|
| 173 |
-
|
| 174 |
-
|
| 175 |
-
|
| 176 |
-
|
| 177 |
-
|
| 178 |
-
|
| 179 |
-
|
| 180 |
-
|
| 181 |
-
|
| 182 |
-
|
| 183 |
-
|
| 184 |
-
|
| 185 |
-
|
| 186 |
-
|
| 187 |
-
|
| 188 |
-
|
| 189 |
-
|
| 190 |
-
|
| 191 |
-
|
| 192 |
-
|
| 193 |
-
|
| 194 |
-
|
| 195 |
-
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
|
| 199 |
-
|
| 200 |
-
|
| 201 |
-
|
| 202 |
-
|
| 203 |
-
|
| 204 |
-
|
| 205 |
-
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
|
| 209 |
-
|
| 210 |
-
|
| 211 |
-
|
| 212 |
-
|
| 213 |
-
|
| 214 |
-
|
| 215 |
-
|
| 216 |
-
|
| 217 |
-
|
| 218 |
-
|
| 219 |
-
|
| 220 |
-
|
| 221 |
-
|
| 222 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model:
|
| 4 |
+
- Qwen/Qwen2.5-14B-Instruct
|
| 5 |
+
language:
|
| 6 |
+
- zho
|
| 7 |
+
- eng
|
| 8 |
+
- fra
|
| 9 |
+
- spa
|
| 10 |
+
- por
|
| 11 |
+
- deu
|
| 12 |
+
- ita
|
| 13 |
+
- rus
|
| 14 |
+
- jpn
|
| 15 |
+
- kor
|
| 16 |
+
- vie
|
| 17 |
+
- tha
|
| 18 |
+
- ara
|
| 19 |
+
pipeline_tag: text-generation
|
| 20 |
+
library_name: transformers
|
| 21 |
+
tags:
|
| 22 |
+
- qwen2.5
|
| 23 |
+
- Cot
|
| 24 |
+
- elite
|
| 25 |
+
- calcium
|
| 26 |
+
model-index:
|
| 27 |
+
- name: Calcium-Opus-14B-Elite3
|
| 28 |
+
results:
|
| 29 |
+
- task:
|
| 30 |
+
type: text-generation
|
| 31 |
+
name: Text Generation
|
| 32 |
+
dataset:
|
| 33 |
+
name: IFEval (0-Shot)
|
| 34 |
+
type: wis-k/instruction-following-eval
|
| 35 |
+
split: train
|
| 36 |
+
args:
|
| 37 |
+
num_few_shot: 0
|
| 38 |
+
metrics:
|
| 39 |
+
- type: inst_level_strict_acc and prompt_level_strict_acc
|
| 40 |
+
value: 54.28
|
| 41 |
+
name: averaged accuracy
|
| 42 |
+
source:
|
| 43 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 44 |
+
name: Open LLM Leaderboard
|
| 45 |
+
- task:
|
| 46 |
+
type: text-generation
|
| 47 |
+
name: Text Generation
|
| 48 |
+
dataset:
|
| 49 |
+
name: BBH (3-Shot)
|
| 50 |
+
type: SaylorTwift/bbh
|
| 51 |
+
split: test
|
| 52 |
+
args:
|
| 53 |
+
num_few_shot: 3
|
| 54 |
+
metrics:
|
| 55 |
+
- type: acc_norm
|
| 56 |
+
value: 47.07
|
| 57 |
+
name: normalized accuracy
|
| 58 |
+
source:
|
| 59 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 60 |
+
name: Open LLM Leaderboard
|
| 61 |
+
- task:
|
| 62 |
+
type: text-generation
|
| 63 |
+
name: Text Generation
|
| 64 |
+
dataset:
|
| 65 |
+
name: MATH Lvl 5 (4-Shot)
|
| 66 |
+
type: lighteval/MATH-Hard
|
| 67 |
+
split: test
|
| 68 |
+
args:
|
| 69 |
+
num_few_shot: 4
|
| 70 |
+
metrics:
|
| 71 |
+
- type: exact_match
|
| 72 |
+
value: 29.38
|
| 73 |
+
name: exact match
|
| 74 |
+
source:
|
| 75 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 76 |
+
name: Open LLM Leaderboard
|
| 77 |
+
- task:
|
| 78 |
+
type: text-generation
|
| 79 |
+
name: Text Generation
|
| 80 |
+
dataset:
|
| 81 |
+
name: GPQA (0-shot)
|
| 82 |
+
type: Idavidrein/gpqa
|
| 83 |
+
split: train
|
| 84 |
+
args:
|
| 85 |
+
num_few_shot: 0
|
| 86 |
+
metrics:
|
| 87 |
+
- type: acc_norm
|
| 88 |
+
value: 16.11
|
| 89 |
+
name: acc_norm
|
| 90 |
+
source:
|
| 91 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 92 |
+
name: Open LLM Leaderboard
|
| 93 |
+
- task:
|
| 94 |
+
type: text-generation
|
| 95 |
+
name: Text Generation
|
| 96 |
+
dataset:
|
| 97 |
+
name: MuSR (0-shot)
|
| 98 |
+
type: TAUR-Lab/MuSR
|
| 99 |
+
args:
|
| 100 |
+
num_few_shot: 0
|
| 101 |
+
metrics:
|
| 102 |
+
- type: acc_norm
|
| 103 |
+
value: 20.13
|
| 104 |
+
name: acc_norm
|
| 105 |
+
source:
|
| 106 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 107 |
+
name: Open LLM Leaderboard
|
| 108 |
+
- task:
|
| 109 |
+
type: text-generation
|
| 110 |
+
name: Text Generation
|
| 111 |
+
dataset:
|
| 112 |
+
name: MMLU-PRO (5-shot)
|
| 113 |
+
type: TIGER-Lab/MMLU-Pro
|
| 114 |
+
config: main
|
| 115 |
+
split: test
|
| 116 |
+
args:
|
| 117 |
+
num_few_shot: 5
|
| 118 |
+
metrics:
|
| 119 |
+
- type: acc
|
| 120 |
+
value: 48.17
|
| 121 |
+
name: accuracy
|
| 122 |
+
source:
|
| 123 |
+
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/?search=prithivMLmods%2FCalcium-Opus-14B-Elite3
|
| 124 |
+
name: Open LLM Leaderboard
|
| 125 |
+
---
|
| 126 |
+
|
| 127 |
+
|
| 128 |
+
|
| 129 |
+
# **Calcium-Opus-14B-Elite3**
|
| 130 |
+
|
| 131 |
+
Calcium-Opus-14B-Elite3 is based on the Qwen 2.5 14B modality architecture, designed to enhance the reasoning capabilities of 14B-parameter models. These models have proven effective in context understanding, reasoning, and mathematical problem-solving. It has been fine-tuned using a long chain-of-thought reasoning model and specialized datasets, with a focus on chain-of-thought (CoT) reasoning for problem-solving. This model is optimized for tasks requiring logical reasoning, detailed explanations, and multi-step problem-solving, making it ideal for applications such as instruction-following, text generation, and complex reasoning tasks.
|
| 132 |
+
|
| 133 |
+
Key improvements include:
|
| 134 |
+
|
| 135 |
+
1. **Enhanced Knowledge and Expertise**: The model demonstrates significantly more knowledge and greatly improved capabilities in coding and mathematics, thanks to specialized expert models in these domains.
|
| 136 |
+
2. **Improved Instruction Following**: It shows significant advancements in following instructions, generating long texts (over 8K tokens), understanding structured data (e.g., tables), and producing structured outputs, especially in JSON format.
|
| 137 |
+
3. **Better Adaptability**: The model is more resilient to diverse system prompts, enabling enhanced role-playing implementations and condition-setting for chatbots.
|
| 138 |
+
4. **Long-Context Support**: It offers long-context support of up to 128K tokens and can generate up to 8K tokens in a single output.
|
| 139 |
+
5. **Multilingual Proficiency**: The model supports over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.
|
| 140 |
+
|
| 141 |
+
# **Quickstart with transformers**
|
| 142 |
+
|
| 143 |
+
Here provides a code snippet with `apply_chat_template` to show you how to load the tokenizer and model and how to generate contents.
|
| 144 |
+
|
| 145 |
+
```python
|
| 146 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 147 |
+
|
| 148 |
+
model_name = "prithivMLmods/Calcium-Opus-14B-Elite3"
|
| 149 |
+
|
| 150 |
+
model = AutoModelForCausalLM.from_pretrained(
|
| 151 |
+
model_name,
|
| 152 |
+
torch_dtype="auto",
|
| 153 |
+
device_map="auto"
|
| 154 |
+
)
|
| 155 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
| 156 |
+
|
| 157 |
+
prompt = "Give me a short introduction to large language model."
|
| 158 |
+
messages = [
|
| 159 |
+
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
|
| 160 |
+
{"role": "user", "content": prompt}
|
| 161 |
+
]
|
| 162 |
+
text = tokenizer.apply_chat_template(
|
| 163 |
+
messages,
|
| 164 |
+
tokenize=False,
|
| 165 |
+
add_generation_prompt=True
|
| 166 |
+
)
|
| 167 |
+
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
| 168 |
+
|
| 169 |
+
generated_ids = model.generate(
|
| 170 |
+
**model_inputs,
|
| 171 |
+
max_new_tokens=512
|
| 172 |
+
)
|
| 173 |
+
generated_ids = [
|
| 174 |
+
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
| 175 |
+
]
|
| 176 |
+
|
| 177 |
+
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
| 178 |
+
```
|
| 179 |
+
|
| 180 |
+
# **Intended Use**
|
| 181 |
+
|
| 182 |
+
1. **Reasoning and Context Understanding**:\
|
| 183 |
+
Designed to assist with complex reasoning tasks, contextual understanding, and solving problems requiring logical deduction and critical thinking.
|
| 184 |
+
|
| 185 |
+
2. **Mathematical Problem-Solving**:\
|
| 186 |
+
Specialized for performing advanced mathematical reasoning and calculations, making it suitable for educational, scientific, and research-oriented applications.
|
| 187 |
+
|
| 188 |
+
3. **Code Generation and Debugging**:\
|
| 189 |
+
Offers robust support for coding tasks, including writing, debugging, and optimizing code in various programming languages, ideal for developers and software engineers.
|
| 190 |
+
|
| 191 |
+
4. **Structured Data Analysis**:\
|
| 192 |
+
Excels in processing and analyzing structured data, such as tables and JSON, and generating structured outputs, which is useful for data analysts and automation workflows.
|
| 193 |
+
|
| 194 |
+
5. **Multilingual Applications**:\
|
| 195 |
+
Supports over 29 languages, making it versatile for global applications like multilingual chatbots, content generation, and translations.
|
| 196 |
+
|
| 197 |
+
6. **Extended Content Generation**:\
|
| 198 |
+
Capable of generating long-form content (over 8K tokens), useful for writing reports, articles, and creating detailed instructional guides.
|
| 199 |
+
|
| 200 |
+
# **Limitations**
|
| 201 |
+
|
| 202 |
+
1. **Hardware Requirements**:\
|
| 203 |
+
Due to its 20B parameter size and support for long-context inputs, running the model requires significant computational resources, including high-memory GPUs or TPUs.
|
| 204 |
+
|
| 205 |
+
2. **Potential Bias in Multilingual Outputs**:\
|
| 206 |
+
While it supports 29 languages, the quality and accuracy of outputs may vary depending on the language, especially for less-resourced languages.
|
| 207 |
+
|
| 208 |
+
3. **Inconsistent Outputs for Creative Tasks**:\
|
| 209 |
+
The model may occasionally produce inconsistent or repetitive results in creative writing, storytelling, or highly subjective tasks.
|
| 210 |
+
|
| 211 |
+
4. **Limited Real-World Awareness**:\
|
| 212 |
+
It lacks real-time knowledge of current events beyond its training cutoff, which may limit its ability to respond accurately to the latest information.
|
| 213 |
+
|
| 214 |
+
5. **Error Propagation in Long-Text Outputs**:\
|
| 215 |
+
In generating long texts, minor errors in early outputs can sometimes propagate, reducing the overall coherence and accuracy of the response.
|
| 216 |
+
|
| 217 |
+
6. **Dependency on High-Quality Prompts**:\
|
| 218 |
+
Performance may depend on the quality and specificity of the input prompt, requiring users to carefully design queries for optimal results.
|
| 219 |
+
|
| 220 |
+
|
| 221 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
|
| 222 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/prithivMLmods__Calcium-Opus-14B-Elite3-details)!
|
| 223 |
+
Summarized results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/contents/viewer/default/train?q=prithivMLmods%2FCalcium-Opus-14B-Elite3&sort[column]=Average%20%E2%AC%86%EF%B8%8F&sort[direction]=desc)!
|
| 224 |
+
|
| 225 |
+
| Metric |Value (%)|
|
| 226 |
+
|-------------------|--------:|
|
| 227 |
+
|**Average** | 35.86|
|
| 228 |
+
|IFEval (0-Shot) | 54.28|
|
| 229 |
+
|BBH (3-Shot) | 47.07|
|
| 230 |
+
|MATH Lvl 5 (4-Shot)| 29.38|
|
| 231 |
+
|GPQA (0-shot) | 16.11|
|
| 232 |
+
|MuSR (0-shot) | 20.13|
|
| 233 |
+
|MMLU-PRO (5-shot) | 48.17|
|
| 234 |
+
|