File size: 7,590 Bytes
0ed0716 658350b 0ed0716 658350b 0ed0716 b396464 9978766 0ed0716 658350b ed7d96e 658350b |
1 2 3 4 5 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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 |
---
language:
- hi
- en
license: apache-2.0
base_model: teknium/OpenHermes-2.5
model-index:
- name: open-aditi-hi-v4
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 60.15
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 81.84
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 61.32
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 44.89
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 79.95
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 57.24
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=manishiitg/open-aditi-hi-v4
name: Open LLM Leaderboard
---
Model trained on Hindi and English data.
This model also includes dataset https://huggingface.co/datasets/sarvamai/samvaad-hi-v1
Check latest evals at https://github.com/manishiitg/IndicEval
Try it out: https://colab.research.google.com/drive/1A_hbsq1vrCeAh3dEMvtwxxNxcNZ1BUyW?usp=sharing
For sample responose on different prompts checkout: https://github.com/manishiitg/hi-llm-eval
#### Language Hi
| Model | xlsum-hi | truthfulqa-hi | indic-arc-easy | mmlu_hi | indicqa | flores | indicheadline | indicxparaphrase | hellaswag-indic | indicwikibio | boolq-hi | implicit_hate | indic-arc-challenge | indicsentiment |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| open-aditi-hi-v2 | 0.4213 | 0.6934 | 0.4979 | 0.3253 | 0.0795 | 43.6822 | 0.4565 | 0.6838 | 0.2404 | 0.4846 | 0.8541 | 11.5021 | 0.4462 | 0.9729 |
| open-aditi-hi-v3 | 0.4490 | 0.5369 | 0.5480 | 0.1351 | 0.0058 | 48.2859 | 0.4682 | 0.8846 | 0.4891 | 0.5034 | 0.5401 | 8.8315 | 0.4633 | 0.9519 |
| open-aditi-hi-v4 | 0.4046 | 0.7671 | 0.4529 | 0.2124 | 0.0026 | 47.8500 | 0.1980 | 0.7737 | 0.3595 | 0.4894 | 0.7015 | 5.9709 | 0.3857 | 0.9699 |
| OpenHermes-2.5-Mistral-7B | 0.1774 | 0.3234 | 0.3523 | 0.2769 | 0.2721 | 30.3465 | 0.1996 | 0.8766 | 0.2485 | 0.3332 | 0.5979 | 0.2068 | 0.3396 | 0.9048 |
| OpenHermes-2.5-Mistral-7B-AWQ | 0.1894 | 0.3428 | 0.3291 | 0.2750 | 0.3116 | 29.3681 | 0.2062 | 0.8536 | 0.2479 | 0.3067 | 0.5272 | 6.0594 | 0.3157 | 0.9218 |
| open-aditi-hi-v1 | 0.4212 | 0.4230 | 0.3889 | 0.1398 | 0.1306 | 40.2376 | 0.4248 | 0.5939 | 0.0848 | 0.4104 | 0.3758 | 8.6105 | 0.3558 | 0.8798 |
| Airavata | 0.4650 | 0.0466 | 0.1128 | 0.1336 | 0.0155 | 58.5260 | 0.4346 | 0.6419 | 0.0550 | 0.0637 | 0.0128 | 6.3612 | 0.0836 | 0.0992 |
#### Language En
| Model | boolq | truthfulqa | arc-easy-exact | mmlu | hellaswag | xlsum | arc-challenge |
| --- | --- | --- | --- | --- | --- | --- | --- |
| open-aditi-hi-v4 | 0.3905 | 0.3378 | 0.8460 | 0.5725 | 0.7603 | 0.4384 | 0.7491 |
| OpenHermes-2.5-Mistral-7B | 0.4061 | 0.2081 | 0.8687 | 0.5991 | 0.7999 | 0.4328 | 0.7790 |
| OpenHermes-2.5-Mistral-7B-AWQ | 0.4199 | 0.1897 | 0.8569 | 0.5816 | 0.7826 | 0.4317 | 0.7611 |
| open-aditi-hi-v3 | 0.3749 | 0.3097 | 0.8384 | 0.5478 | 0.7645 | 0.4352 | 0.7415 |
| open-aditi-hi-v2 | 0.3982 | 0.2999 | 0.8388 | 0.5544 | 0.4738 | 0.4349 | 0.7235 |
| open-aditi-hi-v1 | 0.0434 | 0.3317 | 0.7588 | 0.2597 | 0.3509 | 0.4288 | 0.6271 |
| Airavata | 0.5086 | 0.3574 | 0.6772 | 0.1165 | 0.1799 | 0.4393 | 0.1630 |
Task: flores Metric: chrf
Task: implicit_hate Metric: chrf
Task: indicsentiment Metric: accuracy
Task: indicxparaphrase Metric: accuracy
Task: boolq-hi Metric: accuracy
Task: truthfulqa-hi Metric: accuracy
Task: indic-arc-easy Metric: accuracy
Task: indicwikibio Metric: bleurt
Task: hellaswag-indic Metric: accuracy
Task: indicheadline Metric: bleurt
Task: xlsum-hi Metric: bleurt
Task: indic-arc-challenge Metric: accuracy
Task: mmlu_hi Metric: average_acc
Task: indicqa Metric: accuracy
Task: arc-easy-exact Metric: accuracy
Task: hellaswag Metric: accuracy
Task: arc-challenge Metric: accuracy
Task: mmlu Metric: average_acc
Task: boolq Metric: accuracy
Task: xlsum Metric: bleurt
Task: truthfulqa Metric: accuracy
Model evaluation on OpenLLM LeaderBoard
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_manishiitg__open-aditi-hi-v4)
| Metric |Value|
|---------------------------------|----:|
|Avg. |64.23|
|AI2 Reasoning Challenge (25-Shot)|60.15|
|HellaSwag (10-Shot) |81.84|
|MMLU (5-Shot) |61.32|
|TruthfulQA (0-shot) |44.89|
|Winogrande (5-shot) |79.95|
|GSM8k (5-shot) |57.24|
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_manishiitg__open-aditi-hi-v4)
| Metric |Value|
|---------------------------------|----:|
|Avg. |64.23|
|AI2 Reasoning Challenge (25-Shot)|60.15|
|HellaSwag (10-Shot) |81.84|
|MMLU (5-Shot) |61.32|
|TruthfulQA (0-shot) |44.89|
|Winogrande (5-shot) |79.95|
|GSM8k (5-shot) |57.24|
|