Update config.json
Browse files- config.json +26 -29
config.json
CHANGED
@@ -26,41 +26,38 @@
|
|
26 |
"use_cache": true,
|
27 |
"vocab_size": 32000,
|
28 |
"quantization_config": {
|
29 |
-
"
|
30 |
-
"
|
31 |
-
"
|
32 |
-
|
33 |
-
|
34 |
-
"
|
35 |
-
|
36 |
-
|
37 |
-
"
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
"type": "int"
|
47 |
-
}
|
48 |
}
|
49 |
-
}
|
50 |
-
"format": "pack-quantized",
|
51 |
-
"global_compression_ratio": 2.2202386697247802,
|
52 |
-
"ignore": [
|
53 |
-
"lm_head"
|
54 |
-
],
|
55 |
-
"quant_method": "sparseml",
|
56 |
-
"quantization_status": "frozen"
|
57 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
"sparsity_config": {
|
59 |
"format": "dense",
|
60 |
"global_sparsity": 13.035292091297688,
|
61 |
"registry_requires_subclass": false,
|
62 |
"sparsity_structure": "unstructured"
|
63 |
-
}
|
64 |
-
"quant_method": "compressed-tensors"
|
65 |
}
|
66 |
}
|
|
|
26 |
"use_cache": true,
|
27 |
"vocab_size": 32000,
|
28 |
"quantization_config": {
|
29 |
+
"config_groups": {
|
30 |
+
"group_0": {
|
31 |
+
"input_activations": null,
|
32 |
+
"output_activations": null,
|
33 |
+
"targets": [
|
34 |
+
"Linear"
|
35 |
+
],
|
36 |
+
"weights": {
|
37 |
+
"block_structure": null,
|
38 |
+
"dynamic": false,
|
39 |
+
"group_size": 128,
|
40 |
+
"num_bits": 4,
|
41 |
+
"observer": "minmax",
|
42 |
+
"observer_kwargs": {},
|
43 |
+
"strategy": "group",
|
44 |
+
"symmetric": true,
|
45 |
+
"type": "int"
|
|
|
|
|
46 |
}
|
47 |
+
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
},
|
49 |
+
"quant_method": "sparseml"
|
50 |
+
"format": "pack-quantized",
|
51 |
+
"quantization_status": "frozen"
|
52 |
+
"global_compression_ratio": 2.2202386697247802,
|
53 |
+
"ignore": [
|
54 |
+
"lm_head"
|
55 |
+
],
|
56 |
"sparsity_config": {
|
57 |
"format": "dense",
|
58 |
"global_sparsity": 13.035292091297688,
|
59 |
"registry_requires_subclass": false,
|
60 |
"sparsity_structure": "unstructured"
|
61 |
+
}
|
|
|
62 |
}
|
63 |
}
|