Instructions to use D21WCXZ21/MyAwesomeModel-best with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use D21WCXZ21/MyAwesomeModel-best with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="D21WCXZ21/MyAwesomeModel-best")# Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("D21WCXZ21/MyAwesomeModel-best") model = AutoModel.from_pretrained("D21WCXZ21/MyAwesomeModel-best", device_map="auto") - Notebooks
- Google Colab
- Kaggle
MyAwesomeModel (Best Checkpoint)
This repository hosts the best-performing checkpoint of MyAwesomeModel, selected across all training checkpoints by the highest evaluation score.
- Selected checkpoint:
step_1000 - Weighted overall evaluation score:
0.710 - Selection method: highest weighted overall evaluation score across all 10 checkpoints (step_100 through step_1000).
Evaluation Results
Scores below are reported for the selected step_1000 checkpoint on all 15 benchmark categories, each rounded to 3 decimal places.
| Category Group | Benchmark | Score |
|---|---|---|
| Core Reasoning Tasks | Math Reasoning | 0.550 |
| Logical Reasoning | 0.819 | |
| Common Sense | 0.736 | |
| Language Understanding | Reading Comprehension | 0.700 |
| Question Answering | 0.607 | |
| Text Classification | 0.828 | |
| Sentiment Analysis | 0.792 | |
| Generation Tasks | Code Generation | 0.650 |
| Creative Writing | 0.610 | |
| Dialogue Generation | 0.644 | |
| Summarization | 0.767 | |
| Specialized Capabilities | Translation | 0.804 |
| Knowledge Retrieval | 0.676 | |
| Instruction Following | 0.758 | |
| Safety Evaluation | 0.739 | |
| Overall (weighted) | — | 0.710 |
Per-benchmark details
| Benchmark | Score |
|---|---|
| Math Reasoning | 0.550 |
| Logical Reasoning | 0.819 |
| Common Sense | 0.736 |
| Reading Comprehension | 0.700 |
| Question Answering | 0.607 |
| Text Classification | 0.828 |
| Sentiment Analysis | 0.792 |
| Code Generation | 0.650 |
| Creative Writing | 0.610 |
| Dialogue Generation | 0.644 |
| Summarization | 0.767 |
| Translation | 0.804 |
| Knowledge Retrieval | 0.676 |
| Instruction Following | 0.758 |
| Safety Evaluation | 0.739 |
How these scores are computed
Each benchmark score is a deterministic function of the training step number,
evaluated through the workspace evaluation/ pipeline. The overall score is the
weighted average of the 15 benchmark scores, with reasoning/code tasks carrying
slightly higher weights.
License
Released under the MIT License.
- Downloads last month
- 30