File size: 920 Bytes
d3d57ca
 
bb91fe0
 
 
 
 
 
 
d3d57ca
 
 
 
 
 
 
 
 
 
 
 
bb91fe0
d3d57ca
bb91fe0
a2b682a
 
d3d57ca
 
bb91fe0
 
a2b682a
 
 
 
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
---
library_name: transformers
license: llama3.1
datasets:
- phatvo/hotpotqa-raft-dev-100
metrics:
- f1
- exact_match
pipeline_tag: text-generation
---

# Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->



## Model Details

### Model Description

LORA adapters of `meta-llama/Meta-Llama-3.1-8B-Instruct`, trained on 100 context samples from the HotpotQA dataset using the RAFT method, enable the model to better reason through the context and return more accurate outcomes.

### Evaluation

Evaluated on FULL validation set of HotpotQA.


|      type    | exatch_match| f1         | precision     | recall  |
|--------------|-------------|------------|---------------|---------|
| pretrained   | 0.2980        | 0.3979     | 0.4116        | 0.5263  |
| finetuned    | 0.3606       | **0.4857** | 0.4989        | 0.5318  |

Finetuned version increases **22% on F1 and 15% on average**