Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,47 @@
|
|
| 1 |
-
---
|
| 2 |
-
license:
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: llama3.2
|
| 3 |
+
datasets:
|
| 4 |
+
- mlabonne/orpo-dpo-mix-40k
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
base_model:
|
| 8 |
+
- meta-llama/Llama-3.2-1B
|
| 9 |
+
library_name: transformers
|
| 10 |
+
pipeline_tag: text-generation
|
| 11 |
+
model-index:
|
| 12 |
+
- name: week2-llama3-1B
|
| 13 |
+
results:
|
| 14 |
+
- task:
|
| 15 |
+
type: text-generation
|
| 16 |
+
dataset:
|
| 17 |
+
name: mlabonne/orpo-dpo-mix-40k
|
| 18 |
+
type: mlabonne/orpo-dpo-mix-40k
|
| 19 |
+
metrics:
|
| 20 |
+
- name: EQ-Bench (0-Shot)
|
| 21 |
+
type: EQ-Bench (0-Shot)
|
| 22 |
+
value: 1.5355
|
| 23 |
+
---
|
| 24 |
+
## Model Overview
|
| 25 |
+
This model is a fine-tuned variant of **Llama-3.2-1B**, leveraging **ORPO** (Optimized Regularization for Prompt Optimization) for enhanced performance. It has been fine-tuned using the **mlabonne/orpo-dpo-mix-40k** dataset as part of the *Finetuning Open Source LLMs Course - Week 2 Project*.
|
| 26 |
+
|
| 27 |
+
## Intended Use
|
| 28 |
+
This model is optimized for general-purpose language tasks, including text parsing, understanding contextual prompts, and enhanced interpretability in natural language processing applications.
|
| 29 |
+
|
| 30 |
+
## Evaluation Results
|
| 31 |
+
The model was evaluated on the following benchmarks, with the following performance metrics:
|
| 32 |
+
| Tasks |Version|Filter|n-shot| Metric | | Value | |Stderr|
|
| 33 |
+
|--------|------:|------|-----:|-----------------|---|------:|---|-----:|
|
| 34 |
+
|eq_bench| 2.1|none | 0|eqbench |↑ | 1.5355|± |0.9184|
|
| 35 |
+
| | |none | 0|percent_parseable|↑ |16.9591|± |2.8782|
|
| 36 |
+
|hellaswag| 1|none | 0|acc |↑ |0.4812|± |0.0050|
|
| 37 |
+
| | |none | 0|acc_norm |↑ |0.6467|± |0.0049|
|
| 38 |
+
|ifeval | 4|none | 0|inst_level_loose_acc |↑ |0.3984|± | N/A|
|
| 39 |
+
| | |none | 0|inst_level_strict_acc |↑ |0.2974|± | N/A|
|
| 40 |
+
| | |none | 0|prompt_level_loose_acc |↑ |0.2755|± |0.0193|
|
| 41 |
+
| | |none | 0|prompt_level_strict_acc|↑ |0.1848|± |0.0168|
|
| 42 |
+
|tinyMMLU | 0|none | 0|acc_norm |↑ |0.3995|± | N/A|
|
| 43 |
+
|
| 44 |
+
## Key Features
|
| 45 |
+
- **Model Size**: 1 Billion parameters
|
| 46 |
+
- **Fine-tuning Method**: ORPO
|
| 47 |
+
- **Dataset**: mlabonne/orpo-dpo-mix-40k
|