Update README.md
Browse files
README.md
CHANGED
@@ -68,6 +68,7 @@ This repository has the [source code](https://github.com/Nkluge-correa/TeenyTiny
|
|
68 |
- [Tokenizers](https://github.com/huggingface/tokenizers)
|
69 |
- [Sentencepiece](https://github.com/google/sentencepiece)
|
70 |
- [Accelerate](https://github.com/huggingface/accelerate)
|
|
|
71 |
- [Codecarbon](https://github.com/mlco2/codecarbon)
|
72 |
- [AutoAWQ](https://github.com/casper-hansen/AutoAWQ)
|
73 |
|
@@ -185,9 +186,7 @@ for i, completion in enumerate(completions):
|
|
185 |
| 1,000,000 | 2.50 | 12.27 | 96.19 kWh | 34.18 KgCO2eq |
|
186 |
| 1,100,000 | 2.48 | 11.96 | 106.06 kWh | 37.70 KgCO2eq |
|
187 |
| 1,200,000 | 2.46 | 11.77 | 115.69 kWh | 41.11 KgCO2eq |
|
188 |
-
|
189 |
-
- Note: Each evaluation consumed around 0.26 kWh of energy (~ 0.09 KgCO2eq), totaling 3.12 kWh (~ 1,11
|
190 |
-
KgCO2eq).
|
191 |
|
192 |
## Benchmarks
|
193 |
|
@@ -202,8 +201,8 @@ Evaluations on benchmarks were performed using the [Language Model Evaluation Ha
|
|
202 |
| [OPT-350m](https://huggingface.co/facebook/opt-350m) | 31.78 | 23.55* | 36.73* | 26.02* | 40.83* |
|
203 |
| [TeenyTinyLlama-160m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-160m) | 31.16 | 26.15 | 29.29 | 28.11 | 41.12 |
|
204 |
| [Pythia-160m](https://huggingface.co/EleutherAI/pythia-160m-deduped) | 31.16 | 24.06* | 31.39* | 24.86* | 44.34* |
|
205 |
-
| [OPT-125m](https://huggingface.co/facebook/opt-125m) | 30.80 | 22.87
|
206 |
-
| [
|
207 |
| [Gpt2-small](https://huggingface.co/gpt2) | 29.97 | 21.48* | 31.60* | 25.79* | 40.65* |
|
208 |
| [Multilingual GPT](https://huggingface.co/ai-forever/mGPT) | 29.45 | 24.79 | 26.37* | 25.17* | 41.50 |
|
209 |
|
@@ -215,7 +214,6 @@ Evaluations on benchmarks were performed using the [Language Model Evaluation Ha
|
|
215 |
| [Bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) | 91.64 | 92.22 | 93.07 | 91.28 | 87.45 | 94.19 |
|
216 |
| [TeenyTinyLlama-460m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-460m) | 91.19 | 91.64 | 91.18 | 92.28 | 86.43 | 94.42 |
|
217 |
| [TeenyTinyLlama-160m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-160m) | 90.33 | 91.14 | 90.00 | 90.71 | 85.78 | 94.05 |
|
218 |
-
| [Gpt2-small-portuguese](https://huggingface.co/pierreguillou/gpt2-small-portuguese) | 89.13 | 91.60 | 86.46 | 87.42 | 86.11 | 94.07 |
|
219 |
|
220 |
## Cite as 🤗
|
221 |
|
|
|
68 |
- [Tokenizers](https://github.com/huggingface/tokenizers)
|
69 |
- [Sentencepiece](https://github.com/google/sentencepiece)
|
70 |
- [Accelerate](https://github.com/huggingface/accelerate)
|
71 |
+
- [FlashAttention](https://github.com/Dao-AILab/flash-attention)
|
72 |
- [Codecarbon](https://github.com/mlco2/codecarbon)
|
73 |
- [AutoAWQ](https://github.com/casper-hansen/AutoAWQ)
|
74 |
|
|
|
186 |
| 1,000,000 | 2.50 | 12.27 | 96.19 kWh | 34.18 KgCO2eq |
|
187 |
| 1,100,000 | 2.48 | 11.96 | 106.06 kWh | 37.70 KgCO2eq |
|
188 |
| 1,200,000 | 2.46 | 11.77 | 115.69 kWh | 41.11 KgCO2eq |
|
189 |
+
| 1,300,000 | 2.50 | 12.20 | 125.49 kWh | 44,59 KgCO2eq |
|
|
|
|
|
190 |
|
191 |
## Benchmarks
|
192 |
|
|
|
201 |
| [OPT-350m](https://huggingface.co/facebook/opt-350m) | 31.78 | 23.55* | 36.73* | 26.02* | 40.83* |
|
202 |
| [TeenyTinyLlama-160m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-160m) | 31.16 | 26.15 | 29.29 | 28.11 | 41.12 |
|
203 |
| [Pythia-160m](https://huggingface.co/EleutherAI/pythia-160m-deduped) | 31.16 | 24.06* | 31.39* | 24.86* | 44.34* |
|
204 |
+
| [OPT-125m](https://huggingface.co/facebook/opt-125m) | 30.80 | 22.87* | 31.47* | 26.02* | 42.87* |
|
205 |
+
| [GPortuguese-2](https://huggingface.co/pierreguillou/gpt2-small-portuguese) | 30.22 | 22.48 | 29.62 | 27.36 | 41.44 |
|
206 |
| [Gpt2-small](https://huggingface.co/gpt2) | 29.97 | 21.48* | 31.60* | 25.79* | 40.65* |
|
207 |
| [Multilingual GPT](https://huggingface.co/ai-forever/mGPT) | 29.45 | 24.79 | 26.37* | 25.17* | 41.50 |
|
208 |
|
|
|
214 |
| [Bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) | 91.64 | 92.22 | 93.07 | 91.28 | 87.45 | 94.19 |
|
215 |
| [TeenyTinyLlama-460m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-460m) | 91.19 | 91.64 | 91.18 | 92.28 | 86.43 | 94.42 |
|
216 |
| [TeenyTinyLlama-160m](https://huggingface.co/nicholasKluge/TeenyTinyLlama-160m) | 90.33 | 91.14 | 90.00 | 90.71 | 85.78 | 94.05 |
|
|
|
217 |
|
218 |
## Cite as 🤗
|
219 |
|