mgoin commited on
Commit
c29ffa7
1 Parent(s): 6c8cf14

Updated compression_config to quantization_config

Browse files
Files changed (1) hide show
  1. config.json +41 -41
config.json CHANGED
@@ -10,46 +10,6 @@
10
  "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
11
  },
12
  "bos_token_id": 1,
13
- "compression_config": {
14
- "config_groups": {
15
- "group_0": {
16
- "input_activations": {
17
- "block_structure": null,
18
- "dynamic": false,
19
- "group_size": null,
20
- "num_bits": 8,
21
- "observer": "minmax",
22
- "observer_kwargs": {},
23
- "strategy": "tensor",
24
- "symmetric": true,
25
- "type": "float"
26
- },
27
- "output_activations": null,
28
- "targets": [
29
- "Linear"
30
- ],
31
- "weights": {
32
- "block_structure": null,
33
- "dynamic": false,
34
- "group_size": null,
35
- "num_bits": 8,
36
- "observer": "minmax",
37
- "observer_kwargs": {},
38
- "strategy": "tensor",
39
- "symmetric": true,
40
- "type": "float"
41
- }
42
- }
43
- },
44
- "format": "naive-quantized",
45
- "global_compression_ratio": 1.246153835878366,
46
- "ignore": [
47
- "lm_head"
48
- ],
49
- "kv_cache_scheme": null,
50
- "quant_method": "compressed-tensors",
51
- "quantization_status": "frozen"
52
- },
53
  "embd_pdrop": 0.0,
54
  "eos_token_id": 32000,
55
  "hidden_act": "silu",
@@ -174,5 +134,45 @@
174
  "torch_dtype": "float16",
175
  "transformers_version": "4.44.0.dev0",
176
  "use_cache": true,
177
- "vocab_size": 32064
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
178
  }
 
10
  "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
11
  },
12
  "bos_token_id": 1,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
13
  "embd_pdrop": 0.0,
14
  "eos_token_id": 32000,
15
  "hidden_act": "silu",
 
134
  "torch_dtype": "float16",
135
  "transformers_version": "4.44.0.dev0",
136
  "use_cache": true,
137
+ "vocab_size": 32064,
138
+ "quantization_config": {
139
+ "config_groups": {
140
+ "group_0": {
141
+ "input_activations": {
142
+ "block_structure": null,
143
+ "dynamic": false,
144
+ "group_size": null,
145
+ "num_bits": 8,
146
+ "observer": "minmax",
147
+ "observer_kwargs": {},
148
+ "strategy": "tensor",
149
+ "symmetric": true,
150
+ "type": "float"
151
+ },
152
+ "output_activations": null,
153
+ "targets": [
154
+ "Linear"
155
+ ],
156
+ "weights": {
157
+ "block_structure": null,
158
+ "dynamic": false,
159
+ "group_size": null,
160
+ "num_bits": 8,
161
+ "observer": "minmax",
162
+ "observer_kwargs": {},
163
+ "strategy": "tensor",
164
+ "symmetric": true,
165
+ "type": "float"
166
+ }
167
+ }
168
+ },
169
+ "format": "naive-quantized",
170
+ "global_compression_ratio": 1.246153835878366,
171
+ "ignore": [
172
+ "lm_head"
173
+ ],
174
+ "kv_cache_scheme": null,
175
+ "quant_method": "compressed-tensors",
176
+ "quantization_status": "frozen"
177
+ }
178
  }