Upload folder using huggingface_hub
Browse files- mergekit_moe_config.yml +4 -65
- model-00005-of-00005.safetensors +1 -1
- special_tokens_map.json +2 -1
- tokenizer_config.json +1 -0
mergekit_moe_config.yml
CHANGED
|
@@ -5,71 +5,10 @@ dtype: bfloat16
|
|
| 5 |
experts_per_token: 2
|
| 6 |
experts:
|
| 7 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 8 |
-
positive_prompts:
|
| 9 |
-
- "What are the different"
|
| 10 |
-
- "what are the distinct"
|
| 11 |
-
- "Give me the unique"
|
| 12 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 13 |
-
positive_prompts:
|
| 14 |
-
- "When"
|
| 15 |
-
- "when"
|
| 16 |
-
- "Where"
|
| 17 |
-
- "where"
|
| 18 |
-
- "Which"
|
| 19 |
-
- "which"
|
| 20 |
-
- "Who"
|
| 21 |
-
- "who"
|
| 22 |
-
- "What"
|
| 23 |
-
- "what"
|
| 24 |
-
- "Whom"
|
| 25 |
-
- "whom"
|
| 26 |
-
- "Whose"
|
| 27 |
-
- "whose"
|
| 28 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 29 |
-
positive_prompts:
|
| 30 |
-
- "Larger"
|
| 31 |
-
- "larger"
|
| 32 |
-
- "Smaller"
|
| 33 |
-
- "smaller"
|
| 34 |
-
- "Bigger"
|
| 35 |
-
- "bigger"
|
| 36 |
-
- "Smallest"
|
| 37 |
-
- "smallest"
|
| 38 |
-
- "Largest"
|
| 39 |
-
- "largest"
|
| 40 |
-
- "Biggest"
|
| 41 |
-
- "biggest"
|
| 42 |
-
- "Most"
|
| 43 |
-
- "most"
|
| 44 |
-
- "Least"
|
| 45 |
-
- "least"
|
| 46 |
-
- "More"
|
| 47 |
-
- "more"
|
| 48 |
-
- "Less"
|
| 49 |
-
- "less"
|
| 50 |
-
- "Number"
|
| 51 |
-
- "number"
|
| 52 |
-
- "Numbers"
|
| 53 |
-
- "numbers"
|
| 54 |
-
- "Quantity"
|
| 55 |
-
- "quantity"
|
| 56 |
-
- "At least"
|
| 57 |
-
- "at least"
|
| 58 |
-
- "At most"
|
| 59 |
-
- "at most"
|
| 60 |
-
- "Greater"
|
| 61 |
-
- "greater"
|
| 62 |
-
- "Fewer"
|
| 63 |
-
- "fewer"
|
| 64 |
-
- "Than"
|
| 65 |
-
- "than"
|
| 66 |
-
- "Equal"
|
| 67 |
-
- "equal"
|
| 68 |
-
- "Same"
|
| 69 |
-
- "same"
|
| 70 |
-
- "Equal to"
|
| 71 |
-
- "equal to"
|
| 72 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 73 |
-
positive_prompts:
|
| 74 |
-
- "that also"
|
| 75 |
-
- "who have the same"
|
|
|
|
| 5 |
experts_per_token: 2
|
| 6 |
experts:
|
| 7 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 8 |
+
positive_prompts: []
|
|
|
|
|
|
|
|
|
|
| 9 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 10 |
+
positive_prompts: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 12 |
+
positive_prompts: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
- source_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 14 |
+
positive_prompts: []
|
|
|
|
|
|
model-00005-of-00005.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9967007856
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8e6de51562fbb8dbda2f705a6af7200cc1272963705af0ed90dd17d36c217e0
|
| 3 |
size 9967007856
|
special_tokens_map.json
CHANGED
|
@@ -12,5 +12,6 @@
|
|
| 12 |
"normalized": false,
|
| 13 |
"rstrip": false,
|
| 14 |
"single_word": false
|
| 15 |
-
}
|
|
|
|
| 16 |
}
|
|
|
|
| 12 |
"normalized": false,
|
| 13 |
"rstrip": false,
|
| 14 |
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|begin_of_text|>"
|
| 17 |
}
|
tokenizer_config.json
CHANGED
|
@@ -2058,5 +2058,6 @@
|
|
| 2058 |
"attention_mask"
|
| 2059 |
],
|
| 2060 |
"model_max_length": 1000000000000000019884624838656,
|
|
|
|
| 2061 |
"tokenizer_class": "PreTrainedTokenizerFast"
|
| 2062 |
}
|
|
|
|
| 2058 |
"attention_mask"
|
| 2059 |
],
|
| 2060 |
"model_max_length": 1000000000000000019884624838656,
|
| 2061 |
+
"pad_token": "<|begin_of_text|>",
|
| 2062 |
"tokenizer_class": "PreTrainedTokenizerFast"
|
| 2063 |
}
|