Remove obsolete mistral variants
Browse files
inference-cache-config/mistral-variants.json
CHANGED
|
@@ -37,125 +37,13 @@
|
|
| 37 |
"auto_cast_type": "bf16"
|
| 38 |
}
|
| 39 |
],
|
| 40 |
-
"
|
| 41 |
-
{
|
| 42 |
-
"batch_size": 1,
|
| 43 |
-
"sequence_length": 4096,
|
| 44 |
-
"num_cores": 2,
|
| 45 |
-
"auto_cast_type": "bf16"
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"batch_size": 4,
|
| 49 |
-
"sequence_length": 4096,
|
| 50 |
-
"num_cores": 2,
|
| 51 |
-
"auto_cast_type": "bf16"
|
| 52 |
-
}
|
| 53 |
-
],
|
| 54 |
-
"BioMistral/BioMistral-7B": [
|
| 55 |
-
{
|
| 56 |
-
"batch_size": 1,
|
| 57 |
-
"sequence_length": 4096,
|
| 58 |
-
"num_cores": 2,
|
| 59 |
-
"auto_cast_type": "bf16"
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"batch_size": 4,
|
| 63 |
-
"sequence_length": 4096,
|
| 64 |
-
"num_cores": 2,
|
| 65 |
-
"auto_cast_type": "bf16"
|
| 66 |
-
}
|
| 67 |
-
],
|
| 68 |
-
"NousResearch/Hermes-2-Pro-Mistral-7B": [
|
| 69 |
-
{
|
| 70 |
-
"batch_size": 1,
|
| 71 |
-
"sequence_length": 4096,
|
| 72 |
-
"num_cores": 2,
|
| 73 |
-
"auto_cast_type": "bf16"
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"batch_size": 4,
|
| 77 |
-
"sequence_length": 4096,
|
| 78 |
-
"num_cores": 2,
|
| 79 |
-
"auto_cast_type": "bf16"
|
| 80 |
-
}
|
| 81 |
-
],
|
| 82 |
-
"NousResearch/Nous-Hermes-2-Mistral-7B-DPO": [
|
| 83 |
-
{
|
| 84 |
-
"batch_size": 1,
|
| 85 |
-
"sequence_length": 4096,
|
| 86 |
-
"num_cores": 2,
|
| 87 |
-
"auto_cast_type": "bf16"
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"batch_size": 4,
|
| 91 |
-
"sequence_length": 4096,
|
| 92 |
-
"num_cores": 2,
|
| 93 |
-
"auto_cast_type": "bf16"
|
| 94 |
-
}
|
| 95 |
-
],
|
| 96 |
-
"ibm/merlinite-7b": [
|
| 97 |
-
{
|
| 98 |
-
"batch_size": 1,
|
| 99 |
-
"sequence_length": 4096,
|
| 100 |
-
"num_cores": 2,
|
| 101 |
-
"auto_cast_type": "bf16"
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"batch_size": 4,
|
| 105 |
-
"sequence_length": 4096,
|
| 106 |
-
"num_cores": 2,
|
| 107 |
-
"auto_cast_type": "bf16"
|
| 108 |
-
}
|
| 109 |
-
],
|
| 110 |
-
"mlabonne/AlphaMonarch-7B": [
|
| 111 |
{
|
| 112 |
"batch_size": 1,
|
| 113 |
"sequence_length": 4096,
|
| 114 |
"num_cores": 2,
|
| 115 |
"auto_cast_type": "fp16"
|
| 116 |
},
|
| 117 |
-
{
|
| 118 |
-
"batch_size": 4,
|
| 119 |
-
"sequence_length": 4096,
|
| 120 |
-
"num_cores": 2,
|
| 121 |
-
"auto_cast_type": "fp16"
|
| 122 |
-
}
|
| 123 |
-
],
|
| 124 |
-
"teknium/OpenHermes-2.5-Mistral-7B": [
|
| 125 |
-
{
|
| 126 |
-
"batch_size": 1,
|
| 127 |
-
"sequence_length": 4096,
|
| 128 |
-
"num_cores": 2,
|
| 129 |
-
"auto_cast_type": "bf16"
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"batch_size": 4,
|
| 133 |
-
"sequence_length": 4096,
|
| 134 |
-
"num_cores": 2,
|
| 135 |
-
"auto_cast_type": "bf16"
|
| 136 |
-
}
|
| 137 |
-
],
|
| 138 |
-
"TencentARC/Mistral_Pro_8B_v0.1": [
|
| 139 |
-
{
|
| 140 |
-
"batch_size": 1,
|
| 141 |
-
"sequence_length": 4096,
|
| 142 |
-
"num_cores": 2,
|
| 143 |
-
"auto_cast_type": "bf16"
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"batch_size": 4,
|
| 147 |
-
"sequence_length": 4096,
|
| 148 |
-
"num_cores": 2,
|
| 149 |
-
"auto_cast_type": "bf16"
|
| 150 |
-
}
|
| 151 |
-
],
|
| 152 |
-
"openchat/openchat-3.5-0106": [
|
| 153 |
-
{
|
| 154 |
-
"batch_size": 1,
|
| 155 |
-
"sequence_length": 4096,
|
| 156 |
-
"num_cores": 2,
|
| 157 |
-
"auto_cast_type": "bf16"
|
| 158 |
-
},
|
| 159 |
{
|
| 160 |
"batch_size": 4,
|
| 161 |
"sequence_length": 4096,
|
|
|
|
| 37 |
"auto_cast_type": "bf16"
|
| 38 |
}
|
| 39 |
],
|
| 40 |
+
"Intel/neural-chat-7b-v3-3": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
{
|
| 42 |
"batch_size": 1,
|
| 43 |
"sequence_length": 4096,
|
| 44 |
"num_cores": 2,
|
| 45 |
"auto_cast_type": "fp16"
|
| 46 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
{
|
| 48 |
"batch_size": 4,
|
| 49 |
"sequence_length": 4096,
|