mgoin commited on
Commit
b39503f
·
verified ·
1 Parent(s): fbf83b4

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +33 -1
config.json CHANGED
@@ -22,5 +22,37 @@
22
  "torch_dtype": "bfloat16",
23
  "transformers_version": "4.52.4",
24
  "use_cache": true,
25
- "vocab_size": 131072
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
26
  }
 
22
  "torch_dtype": "bfloat16",
23
  "transformers_version": "4.52.4",
24
  "use_cache": true,
25
+ "vocab_size": 131072,
26
+ "quantization_config": {
27
+ "config_groups": {
28
+ "group_0": {
29
+ "input_activations": {
30
+ "dynamic": true,
31
+ "num_bits": 8,
32
+ "observer": null,
33
+ "strategy": "token",
34
+ "symmetric": true,
35
+ "type": "float"
36
+ },
37
+ "targets": [
38
+ "Linear"
39
+ ],
40
+ "weights": {
41
+ "dynamic": false,
42
+ "num_bits": 8,
43
+ "observer": "minmax",
44
+ "strategy": "tensor",
45
+ "symmetric": true,
46
+ "type": "float"
47
+ }
48
+ }
49
+ },
50
+ "format": "float-quantized",
51
+ "ignore": [
52
+ "lm_head",
53
+ "output"
54
+ ],
55
+ "quant_method": "compressed-tensors",
56
+ "quantization_status": "compressed"
57
+ }
58
  }