File size: 745 Bytes
87eb9b3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
---

# The Quantized BAAI FlagEmbedding (BGE-EN-ICL) Model

Original Base Embedding Model: `BAAI/bge-en-icl`.<br>
Link: [https://huggingface.co/BAAI/bge-en-icl](https://huggingface.co/BAAI/bge-en-icl)

## Quantization Configurations

```text
"quantization_config": {
    "bits": 4,
    "checkpoint_format": "gptq",
    "damp_percent": 0.01,
    "desc_act": true,
    "group_size": 128,
    "model_file_base_name": null,
    "model_name_or_path": null,
    "quant_method": "gptq",
    "static_groups": false,
    "sym": true,
    "true_sequential": true
  },
```

## Source Codes
Source Codes: [https://github.com/vkola-lab/medpodgpt/tree/main/quantization](https://github.com/vkola-lab/medpodgpt/tree/main/quantization).