Bram Vanroy
commited on
Commit
•
351f9fe
1
Parent(s):
b1e6575
add correct evals:
Browse files- evals/models.json +53 -45
evals/models.json
CHANGED
|
@@ -1,122 +1,130 @@
|
|
| 1 |
{
|
| 2 |
"gpt-neo-1.3b-dutch": {
|
| 3 |
"compute_dtype": "bfloat16",
|
|
|
|
| 4 |
"model_name": "yhavinga/gpt-neo-1.3B-dutch",
|
| 5 |
-
"num_parameters": 1315575808,
|
| 6 |
-
"quantization": "8-bit",
|
| 7 |
"model_type": "pretrained",
|
| 8 |
-
"
|
|
|
|
| 9 |
},
|
| 10 |
"gpt-neo-125m-dutch": {
|
| 11 |
"compute_dtype": "bfloat16",
|
|
|
|
| 12 |
"model_name": "yhavinga/gpt-neo-125M-dutch",
|
| 13 |
-
"num_parameters": 125198592,
|
| 14 |
-
"quantization": "8-bit",
|
| 15 |
"model_type": "pretrained",
|
| 16 |
-
"
|
|
|
|
| 17 |
},
|
| 18 |
"gpt2-large-dutch": {
|
| 19 |
"compute_dtype": "bfloat16",
|
|
|
|
| 20 |
"model_name": "yhavinga/gpt2-large-dutch",
|
| 21 |
-
"num_parameters": 774030080,
|
| 22 |
-
"quantization": "8-bit",
|
| 23 |
"model_type": "pretrained",
|
| 24 |
-
"
|
|
|
|
| 25 |
},
|
| 26 |
"gpt2-medium-dutch": {
|
| 27 |
"compute_dtype": "bfloat16",
|
|
|
|
| 28 |
"model_name": "yhavinga/gpt2-medium-dutch",
|
| 29 |
-
"num_parameters": 354823168,
|
| 30 |
-
"quantization": "8-bit",
|
| 31 |
"model_type": "pretrained",
|
| 32 |
-
"
|
|
|
|
| 33 |
},
|
| 34 |
"llama-2-13b-chat-dutch": {
|
| 35 |
"compute_dtype": "bfloat16",
|
|
|
|
| 36 |
"model_name": "BramVanroy/Llama-2-13b-chat-dutch",
|
| 37 |
-
"num_parameters": 13015864320,
|
| 38 |
-
"quantization": "8-bit",
|
| 39 |
"model_type": "instruction-tuned",
|
| 40 |
-
"
|
|
|
|
| 41 |
},
|
| 42 |
"llama-2-13b-chat-hf": {
|
| 43 |
"compute_dtype": "bfloat16",
|
|
|
|
| 44 |
"model_name": "meta-llama/Llama-2-13b-chat-hf",
|
| 45 |
-
"num_parameters": 13015864320,
|
| 46 |
-
"quantization": "8-bit",
|
| 47 |
"model_type": "instruction-tuned",
|
| 48 |
-
"
|
|
|
|
| 49 |
},
|
| 50 |
"llama-2-13b-hf": {
|
| 51 |
"compute_dtype": "bfloat16",
|
|
|
|
| 52 |
"model_name": "meta-llama/Llama-2-13b-hf",
|
| 53 |
-
"num_parameters": 13015864320,
|
| 54 |
-
"quantization": "8-bit",
|
| 55 |
"model_type": "pretrained",
|
| 56 |
-
"
|
|
|
|
| 57 |
},
|
| 58 |
"llama-2-7b-chat-hf": {
|
| 59 |
"compute_dtype": "bfloat16",
|
|
|
|
| 60 |
"model_name": "meta-llama/Llama-2-7b-chat-hf",
|
| 61 |
-
"num_parameters": 6738415616,
|
| 62 |
-
"quantization": "8-bit",
|
| 63 |
"model_type": "instruction-tuned",
|
| 64 |
-
"
|
|
|
|
| 65 |
},
|
| 66 |
"llama-2-7b-hf": {
|
| 67 |
"compute_dtype": "bfloat16",
|
|
|
|
| 68 |
"model_name": "meta-llama/Llama-2-7b-hf",
|
| 69 |
-
"num_parameters": 6738415616,
|
| 70 |
-
"quantization": "8-bit",
|
| 71 |
"model_type": "pretrained",
|
| 72 |
-
"
|
|
|
|
| 73 |
},
|
| 74 |
"llama2-13b-ft-mc4_nl_cleaned_tiny": {
|
| 75 |
"compute_dtype": "bfloat16",
|
|
|
|
| 76 |
"model_name": "BramVanroy/llama2-13b-ft-mc4_nl_cleaned_tiny",
|
| 77 |
-
"num_parameters": 13015864320,
|
| 78 |
-
"quantization": "8-bit",
|
| 79 |
"model_type": "fine-tuned",
|
| 80 |
-
"
|
|
|
|
| 81 |
},
|
| 82 |
"mistral-7b-v0.1": {
|
| 83 |
"compute_dtype": "bfloat16",
|
|
|
|
| 84 |
"model_name": "mistralai/Mistral-7B-v0.1",
|
| 85 |
-
"num_parameters": 7241732096,
|
| 86 |
-
"quantization": "8-bit",
|
| 87 |
"model_type": "pretrained",
|
| 88 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
},
|
| 90 |
"neural-chat-7b-v3-1": {
|
| 91 |
"compute_dtype": "bfloat16",
|
|
|
|
| 92 |
"model_name": "Intel/neural-chat-7b-v3-1",
|
| 93 |
-
"num_parameters": 7241732096,
|
| 94 |
-
"quantization": "8-bit",
|
| 95 |
"model_type": "RL-tuned",
|
| 96 |
-
"
|
|
|
|
| 97 |
},
|
| 98 |
"orca-2-13b": {
|
| 99 |
"compute_dtype": "bfloat16",
|
|
|
|
| 100 |
"model_name": "microsoft/Orca-2-13b",
|
| 101 |
-
"num_parameters": 13015895040,
|
| 102 |
-
"quantization": "8-bit",
|
| 103 |
"model_type": "fine-tuned",
|
| 104 |
-
"
|
|
|
|
| 105 |
},
|
| 106 |
"orca-2-7b": {
|
| 107 |
"compute_dtype": "bfloat16",
|
|
|
|
| 108 |
"model_name": "microsoft/Orca-2-7b",
|
| 109 |
-
"num_parameters": 6738440192,
|
| 110 |
-
"quantization": "8-bit",
|
| 111 |
"model_type": "fine-tuned",
|
| 112 |
-
"
|
|
|
|
| 113 |
},
|
| 114 |
"zephyr-7b-beta": {
|
| 115 |
"compute_dtype": "bfloat16",
|
|
|
|
| 116 |
"model_name": "HuggingFaceH4/zephyr-7b-beta",
|
| 117 |
-
"num_parameters": 7241732096,
|
| 118 |
-
"quantization": "8-bit",
|
| 119 |
"model_type": "RL-tuned",
|
| 120 |
-
"
|
|
|
|
| 121 |
}
|
| 122 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"gpt-neo-1.3b-dutch": {
|
| 3 |
"compute_dtype": "bfloat16",
|
| 4 |
+
"dutch_coverage": "pretrained",
|
| 5 |
"model_name": "yhavinga/gpt-neo-1.3B-dutch",
|
|
|
|
|
|
|
| 6 |
"model_type": "pretrained",
|
| 7 |
+
"num_parameters": 1315575808,
|
| 8 |
+
"quantization": "8-bit"
|
| 9 |
},
|
| 10 |
"gpt-neo-125m-dutch": {
|
| 11 |
"compute_dtype": "bfloat16",
|
| 12 |
+
"dutch_coverage": "pretrained",
|
| 13 |
"model_name": "yhavinga/gpt-neo-125M-dutch",
|
|
|
|
|
|
|
| 14 |
"model_type": "pretrained",
|
| 15 |
+
"num_parameters": 125198592,
|
| 16 |
+
"quantization": "8-bit"
|
| 17 |
},
|
| 18 |
"gpt2-large-dutch": {
|
| 19 |
"compute_dtype": "bfloat16",
|
| 20 |
+
"dutch_coverage": "pretrained",
|
| 21 |
"model_name": "yhavinga/gpt2-large-dutch",
|
|
|
|
|
|
|
| 22 |
"model_type": "pretrained",
|
| 23 |
+
"num_parameters": 774030080,
|
| 24 |
+
"quantization": "8-bit"
|
| 25 |
},
|
| 26 |
"gpt2-medium-dutch": {
|
| 27 |
"compute_dtype": "bfloat16",
|
| 28 |
+
"dutch_coverage": "pretrained",
|
| 29 |
"model_name": "yhavinga/gpt2-medium-dutch",
|
|
|
|
|
|
|
| 30 |
"model_type": "pretrained",
|
| 31 |
+
"num_parameters": 354823168,
|
| 32 |
+
"quantization": "8-bit"
|
| 33 |
},
|
| 34 |
"llama-2-13b-chat-dutch": {
|
| 35 |
"compute_dtype": "bfloat16",
|
| 36 |
+
"dutch_coverage": "fine-tuned",
|
| 37 |
"model_name": "BramVanroy/Llama-2-13b-chat-dutch",
|
|
|
|
|
|
|
| 38 |
"model_type": "instruction-tuned",
|
| 39 |
+
"num_parameters": 13015864320,
|
| 40 |
+
"quantization": "8-bit"
|
| 41 |
},
|
| 42 |
"llama-2-13b-chat-hf": {
|
| 43 |
"compute_dtype": "bfloat16",
|
| 44 |
+
"dutch_coverage": "none",
|
| 45 |
"model_name": "meta-llama/Llama-2-13b-chat-hf",
|
|
|
|
|
|
|
| 46 |
"model_type": "instruction-tuned",
|
| 47 |
+
"num_parameters": 13015864320,
|
| 48 |
+
"quantization": "8-bit"
|
| 49 |
},
|
| 50 |
"llama-2-13b-hf": {
|
| 51 |
"compute_dtype": "bfloat16",
|
| 52 |
+
"dutch_coverage": "none",
|
| 53 |
"model_name": "meta-llama/Llama-2-13b-hf",
|
|
|
|
|
|
|
| 54 |
"model_type": "pretrained",
|
| 55 |
+
"num_parameters": 13015864320,
|
| 56 |
+
"quantization": "8-bit"
|
| 57 |
},
|
| 58 |
"llama-2-7b-chat-hf": {
|
| 59 |
"compute_dtype": "bfloat16",
|
| 60 |
+
"dutch_coverage": "none",
|
| 61 |
"model_name": "meta-llama/Llama-2-7b-chat-hf",
|
|
|
|
|
|
|
| 62 |
"model_type": "instruction-tuned",
|
| 63 |
+
"num_parameters": 6738415616,
|
| 64 |
+
"quantization": "8-bit"
|
| 65 |
},
|
| 66 |
"llama-2-7b-hf": {
|
| 67 |
"compute_dtype": "bfloat16",
|
| 68 |
+
"dutch_coverage": "none",
|
| 69 |
"model_name": "meta-llama/Llama-2-7b-hf",
|
|
|
|
|
|
|
| 70 |
"model_type": "pretrained",
|
| 71 |
+
"num_parameters": 6738415616,
|
| 72 |
+
"quantization": "8-bit"
|
| 73 |
},
|
| 74 |
"llama2-13b-ft-mc4_nl_cleaned_tiny": {
|
| 75 |
"compute_dtype": "bfloat16",
|
| 76 |
+
"dutch_coverage": "fine-tuned",
|
| 77 |
"model_name": "BramVanroy/llama2-13b-ft-mc4_nl_cleaned_tiny",
|
|
|
|
|
|
|
| 78 |
"model_type": "fine-tuned",
|
| 79 |
+
"num_parameters": 13015864320,
|
| 80 |
+
"quantization": "8-bit"
|
| 81 |
},
|
| 82 |
"mistral-7b-v0.1": {
|
| 83 |
"compute_dtype": "bfloat16",
|
| 84 |
+
"dutch_coverage": "none",
|
| 85 |
"model_name": "mistralai/Mistral-7B-v0.1",
|
|
|
|
|
|
|
| 86 |
"model_type": "pretrained",
|
| 87 |
+
"num_parameters": 7241732096,
|
| 88 |
+
"quantization": "8-bit"
|
| 89 |
+
},
|
| 90 |
+
"mixtral-8x7b-v0.1": {
|
| 91 |
+
"compute_dtype": "auto",
|
| 92 |
+
"dutch_coverage": "not-given",
|
| 93 |
+
"model_name": "mistralai/Mixtral-8x7B-v0.1",
|
| 94 |
+
"model_type": "not-given",
|
| 95 |
+
"num_parameters": 46702792704,
|
| 96 |
+
"quantization": null
|
| 97 |
},
|
| 98 |
"neural-chat-7b-v3-1": {
|
| 99 |
"compute_dtype": "bfloat16",
|
| 100 |
+
"dutch_coverage": "none",
|
| 101 |
"model_name": "Intel/neural-chat-7b-v3-1",
|
|
|
|
|
|
|
| 102 |
"model_type": "RL-tuned",
|
| 103 |
+
"num_parameters": 7241732096,
|
| 104 |
+
"quantization": "8-bit"
|
| 105 |
},
|
| 106 |
"orca-2-13b": {
|
| 107 |
"compute_dtype": "bfloat16",
|
| 108 |
+
"dutch_coverage": "none",
|
| 109 |
"model_name": "microsoft/Orca-2-13b",
|
|
|
|
|
|
|
| 110 |
"model_type": "fine-tuned",
|
| 111 |
+
"num_parameters": 13015895040,
|
| 112 |
+
"quantization": "8-bit"
|
| 113 |
},
|
| 114 |
"orca-2-7b": {
|
| 115 |
"compute_dtype": "bfloat16",
|
| 116 |
+
"dutch_coverage": "none",
|
| 117 |
"model_name": "microsoft/Orca-2-7b",
|
|
|
|
|
|
|
| 118 |
"model_type": "fine-tuned",
|
| 119 |
+
"num_parameters": 6738440192,
|
| 120 |
+
"quantization": "8-bit"
|
| 121 |
},
|
| 122 |
"zephyr-7b-beta": {
|
| 123 |
"compute_dtype": "bfloat16",
|
| 124 |
+
"dutch_coverage": "none",
|
| 125 |
"model_name": "HuggingFaceH4/zephyr-7b-beta",
|
|
|
|
|
|
|
| 126 |
"model_type": "RL-tuned",
|
| 127 |
+
"num_parameters": 7241732096,
|
| 128 |
+
"quantization": "8-bit"
|
| 129 |
}
|
| 130 |
}
|