Upload folder using huggingface_hub
Browse files- config.json +4 -3
- special_tokens_map.json +1 -1
- tokenizer_config.json +1 -1
config.json
CHANGED
|
@@ -18,7 +18,8 @@
|
|
| 18 |
"llm_int8_skip_modules": [
|
| 19 |
"vision_model",
|
| 20 |
"multi_modal_projector",
|
| 21 |
-
"lm_head"
|
|
|
|
| 22 |
],
|
| 23 |
"llm_int8_threshold": 6.0,
|
| 24 |
"load_in_4bit": true,
|
|
@@ -158,8 +159,8 @@
|
|
| 158 |
"pad_token_id": 200018,
|
| 159 |
"rms_norm_eps": 1e-05,
|
| 160 |
"rope_scaling": {
|
| 161 |
-
"factor":
|
| 162 |
-
"high_freq_factor":
|
| 163 |
"low_freq_factor": 1.0,
|
| 164 |
"original_max_position_embeddings": 8192,
|
| 165 |
"rope_type": "llama3"
|
|
|
|
| 18 |
"llm_int8_skip_modules": [
|
| 19 |
"vision_model",
|
| 20 |
"multi_modal_projector",
|
| 21 |
+
"lm_head",
|
| 22 |
+
"vision_model"
|
| 23 |
],
|
| 24 |
"llm_int8_threshold": 6.0,
|
| 25 |
"load_in_4bit": true,
|
|
|
|
| 159 |
"pad_token_id": 200018,
|
| 160 |
"rms_norm_eps": 1e-05,
|
| 161 |
"rope_scaling": {
|
| 162 |
+
"factor": 16.0,
|
| 163 |
+
"high_freq_factor": 1.0,
|
| 164 |
"low_freq_factor": 1.0,
|
| 165 |
"original_max_position_embeddings": 8192,
|
| 166 |
"rope_type": "llama3"
|
special_tokens_map.json
CHANGED
|
@@ -14,7 +14,7 @@
|
|
| 14 |
"single_word": false
|
| 15 |
},
|
| 16 |
"pad_token": {
|
| 17 |
-
"content": "<|
|
| 18 |
"lstrip": false,
|
| 19 |
"normalized": false,
|
| 20 |
"rstrip": false,
|
|
|
|
| 14 |
"single_word": false
|
| 15 |
},
|
| 16 |
"pad_token": {
|
| 17 |
+
"content": "<|finetune_right_pad|>",
|
| 18 |
"lstrip": false,
|
| 19 |
"normalized": false,
|
| 20 |
"rstrip": false,
|
tokenizer_config.json
CHANGED
|
@@ -9091,7 +9091,7 @@
|
|
| 9091 |
"attention_mask"
|
| 9092 |
],
|
| 9093 |
"model_max_length": 10485760,
|
| 9094 |
-
"pad_token": "<|
|
| 9095 |
"processor_class": "Llama4Processor",
|
| 9096 |
"tokenizer_class": "PreTrainedTokenizer"
|
| 9097 |
}
|
|
|
|
| 9091 |
"attention_mask"
|
| 9092 |
],
|
| 9093 |
"model_max_length": 10485760,
|
| 9094 |
+
"pad_token": "<|finetune_right_pad|>",
|
| 9095 |
"processor_class": "Llama4Processor",
|
| 9096 |
"tokenizer_class": "PreTrainedTokenizer"
|
| 9097 |
}
|