change to gptq format
Browse filesSigned-off-by: wenhuach <wenhuach87@gmail.com>
- config.json +2 -3
config.json
CHANGED
|
@@ -44,7 +44,6 @@
|
|
| 44 |
"batch_size": 4,
|
| 45 |
"bits": 4,
|
| 46 |
"data_type": "int",
|
| 47 |
-
"dataset": "NeelNanda/pile-10k",
|
| 48 |
"enable_minmax_tuning": true,
|
| 49 |
"enable_norm_bias_tuning": false,
|
| 50 |
"enable_quanted_input": true,
|
|
@@ -55,7 +54,7 @@
|
|
| 55 |
"lr": 0.005,
|
| 56 |
"minmax_lr": 0.005,
|
| 57 |
"nsamples": 512,
|
| 58 |
-
"quant_method": "
|
| 59 |
"scale_dtype": "torch.float16",
|
| 60 |
"seqlen": 2048,
|
| 61 |
"sym": true,
|
|
@@ -83,4 +82,4 @@
|
|
| 83 |
"use_cache": true,
|
| 84 |
"v_head_dim": 128,
|
| 85 |
"vocab_size": 129280
|
| 86 |
-
}
|
|
|
|
| 44 |
"batch_size": 4,
|
| 45 |
"bits": 4,
|
| 46 |
"data_type": "int",
|
|
|
|
| 47 |
"enable_minmax_tuning": true,
|
| 48 |
"enable_norm_bias_tuning": false,
|
| 49 |
"enable_quanted_input": true,
|
|
|
|
| 54 |
"lr": 0.005,
|
| 55 |
"minmax_lr": 0.005,
|
| 56 |
"nsamples": 512,
|
| 57 |
+
"quant_method": "gptq",
|
| 58 |
"scale_dtype": "torch.float16",
|
| 59 |
"seqlen": 2048,
|
| 60 |
"sym": true,
|
|
|
|
| 82 |
"use_cache": true,
|
| 83 |
"v_head_dim": 128,
|
| 84 |
"vocab_size": 129280
|
| 85 |
+
}
|