Skip to content

Commit 6e4bb36

Browse files
Added default in4 configs for (#1194)
1 parent 755a833 commit 6e4bb36

File tree

1 file changed

+18
-0
lines changed

1 file changed

+18
-0
lines changed

optimum/intel/openvino/configuration.py

+18
Original file line numberDiff line numberDiff line change
@@ -89,6 +89,15 @@ class OVQuantizationMethod(str, Enum):
8989
"pansophic/rocket-3B": {"bits": 4, "sym": True, "group_size": 128, "ratio": 0.8},
9090
"THUDM/chatglm2-6b": {"bits": 4, "sym": True, "group_size": 128, "ratio": 0.72},
9191
"Qwen/Qwen-7B-Chat": {"bits": 4, "sym": True, "group_size": 128, "ratio": 0.6},
92+
"Qwen/Qwen2.5-7B-Instruct": {
93+
"bits": 4,
94+
"sym": False,
95+
"group_size": 128,
96+
"ratio": 1.0,
97+
"dataset": "wikitext2",
98+
"quant_method": OVQuantizationMethod.AWQ,
99+
"scale_estimation": True,
100+
},
92101
"openlm-research/open_llama_3b": {"bits": 4, "sym": False, "group_size": 64, "all_layers": True},
93102
"openlm-research/open_llama_3b_v2": {
94103
"bits": 4,
@@ -210,6 +219,15 @@ class OVQuantizationMethod(str, Enum):
210219
"quant_method": OVQuantizationMethod.AWQ,
211220
"scale_estimation": True,
212221
},
222+
"microsoft/Phi-4-mini-instruct": {
223+
"bits": 4,
224+
"sym": False,
225+
"group_size": 64,
226+
"ratio": 1.0,
227+
"dataset": "wikitext2",
228+
"quant_method": OVQuantizationMethod.AWQ,
229+
"scale_estimation": True,
230+
},
213231
"deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B": {
214232
"bits": 4,
215233
"sym": False,

0 commit comments

Comments
 (0)