Update README.md
Browse files
README.md
CHANGED
|
@@ -2,17 +2,21 @@
|
|
| 2 |
license: unknown
|
| 3 |
---
|
| 4 |
|
| 5 |
-
This is TinyLlama/TinyLlama-1.1B-Chat-v1.0 quantized with AutoGPTQ in 4-bit.
|
| 6 |
|
| 7 |
**Quantize config:**
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
"
|
| 11 |
-
"
|
| 12 |
-
"
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: unknown
|
| 3 |
---
|
| 4 |
|
| 5 |
+
This is TinyLlama/TinyLlama-1.1B-Chat-v1.0 quantized with AutoGPTQ in GPTQ 4-bit format.
|
| 6 |
|
| 7 |
**Quantize config:**
|
| 8 |
+
```
|
| 9 |
+
{
|
| 10 |
+
"bits": 4,
|
| 11 |
+
"group_size": 128,
|
| 12 |
+
"damp_percent": 0.01,
|
| 13 |
+
"desc_act": false,
|
| 14 |
+
"static_groups": false,
|
| 15 |
+
"sym": true,
|
| 16 |
+
"true_sequential": true,
|
| 17 |
+
"model_name_or_path": null,
|
| 18 |
+
"model_file_base_name": null,
|
| 19 |
+
"quant_method": "gptq",
|
| 20 |
+
"checkpoint_format": "gptq"
|
| 21 |
+
}
|
| 22 |
+
```
|