Text Classification
Transformers
Safetensors
English
Latin
Greek
distilbert
text-embeddings-inference
Instructions to use sjhuskey/distilbert_multilingual_cased_greek_latin_classifier with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sjhuskey/distilbert_multilingual_cased_greek_latin_classifier with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="sjhuskey/distilbert_multilingual_cased_greek_latin_classifier")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("sjhuskey/distilbert_multilingual_cased_greek_latin_classifier") model = AutoModelForSequenceClassification.from_pretrained("sjhuskey/distilbert_multilingual_cased_greek_latin_classifier", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Samuel J. Huskey commited on
Commit ·
aa57e7c
1
Parent(s): 3171508
update readme with emissions info
Browse files
README.md
CHANGED
|
@@ -25,3 +25,35 @@ Most critical editions of ancient Greek texts bear a Latin version of the title
|
|
| 25 |
Consequently, metadata records tagged with the subject "Latin" from collections such as the [HathiTrust Digital Library](https://www.hathitrust.org/) inevitably include Greek works.
|
| 26 |
|
| 27 |
Since the Digital Latin Library is interested only in records of Latin works, we need a good way of winnowing out the Greek editions. This model does a good job of that.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
Consequently, metadata records tagged with the subject "Latin" from collections such as the [HathiTrust Digital Library](https://www.hathitrust.org/) inevitably include Greek works.
|
| 26 |
|
| 27 |
Since the Digital Latin Library is interested only in records of Latin works, we need a good way of winnowing out the Greek editions. This model does a good job of that.
|
| 28 |
+
|
| 29 |
+
## Emissions
|
| 30 |
+
|
| 31 |
+
Here is the `codecarbon` output from training on Google Colab with an A100 runtime:
|
| 32 |
+
|
| 33 |
+
```properties
|
| 34 |
+
timestamp: 2024-12-25T15:16:45
|
| 35 |
+
project_name: codecarbon
|
| 36 |
+
run_id: f658b237-20c1-45cf-a8ee-cdb5e8521351
|
| 37 |
+
experiment_id: 5b0fa12a-3dd7-45bb-9766-cc326314d9f1
|
| 38 |
+
duration (seconds): 591.3640720729998
|
| 39 |
+
emissions (kilograms of carbon): 0.0178715223377235
|
| 40 |
+
emissions_rate (kg/sec): 3.0220845637570942e-05
|
| 41 |
+
cpu_power (average in watts): 42.5
|
| 42 |
+
gpu_power (average in watts): 105.04535891243202
|
| 43 |
+
ram_power (average in watts): 31.30389261245728
|
| 44 |
+
cpu_energy (total watts): 0.0069742756409347
|
| 45 |
+
gpu_energy (total watts): 0.025850867347344
|
| 46 |
+
ram_energy (total watts): 0.0051361307572122
|
| 47 |
+
energy_consumed (total watts): 0.037961273745491
|
| 48 |
+
os: Linux-6.1.85+-x86_64-with-glibc2.35
|
| 49 |
+
python_version: 3.10.12
|
| 50 |
+
codecarbon_version: 2.8.2
|
| 51 |
+
cpu_count: 12
|
| 52 |
+
cpu_model: Intel(R) Xeon(R) CPU @ 2.20GHz
|
| 53 |
+
gpu_count: 1
|
| 54 |
+
gpu_model: 1 x NVIDIA A100-SXM4-40GB
|
| 55 |
+
ram_total_size: 83.47704696655273
|
| 56 |
+
tracking_mode: machine
|
| 57 |
+
on_cloud: N
|
| 58 |
+
pue: 1.0
|
| 59 |
+
```
|