jondurbin commited on
Commit
9fac89f
·
verified ·
1 Parent(s): 979380c

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. README.md +2 -2
  2. special_tokens_map.json +1 -1
  3. tokenizer_config.json +1 -9
README.md CHANGED
@@ -20,7 +20,7 @@ tags:
20
  - meta
21
  - pytorch
22
  - llama
23
- - llama-4
24
  extra_gated_prompt: >-
25
  **LLAMA 4 COMMUNITY LICENSE AGREEMENT**
26
 
@@ -269,7 +269,7 @@ In this section, we report the results for Llama 4 relative to our previous mode
269
  | Image Understanding | ChartQA | 0 | relaxed\_accuracy | | | 88.8 | 90.0 |
270
  | | DocVQA (test) | 0 | anls | | | 94.4 | 94.4 |
271
  | Coding | LiveCodeBench (10/01/2024-02/01/2025) | 0 | pass@1 | 33.3 | 27.7 | 32.8 | 43.4 |
272
- | Reasoning & Knowledge | MMLU Pro | 0 | macro\_avg/em | 68.9 | 73.4 | 74.3 | 80.5 |
273
  | | GPQA Diamond | 0 | accuracy | 50.5 | 49.0 | 57.2 | 69.8 |
274
  | Multilingual | MGSM | 0 | average/em | 91.1 | 91.6 | 90.6 | 92.3 |
275
  | Long context | MTOB (half book) eng-\>kgv/kgv-\>eng | \- | chrF | Context window is 128K | | 42.2/36.6 | 54.0/46.4 |
 
20
  - meta
21
  - pytorch
22
  - llama
23
+ - llama4
24
  extra_gated_prompt: >-
25
  **LLAMA 4 COMMUNITY LICENSE AGREEMENT**
26
 
 
269
  | Image Understanding | ChartQA | 0 | relaxed\_accuracy | | | 88.8 | 90.0 |
270
  | | DocVQA (test) | 0 | anls | | | 94.4 | 94.4 |
271
  | Coding | LiveCodeBench (10/01/2024-02/01/2025) | 0 | pass@1 | 33.3 | 27.7 | 32.8 | 43.4 |
272
+ | Reasoning & Knowledge | MMLU Pro | 0 | macro\_avg/acc | 68.9 | 73.4 | 74.3 | 80.5 |
273
  | | GPQA Diamond | 0 | accuracy | 50.5 | 49.0 | 57.2 | 69.8 |
274
  | Multilingual | MGSM | 0 | average/em | 91.1 | 91.6 | 90.6 | 92.3 |
275
  | Long context | MTOB (half book) eng-\>kgv/kgv-\>eng | \- | chrF | Context window is 128K | | 42.2/36.6 | 54.0/46.4 |
special_tokens_map.json CHANGED
@@ -14,7 +14,7 @@
14
  "single_word": false
15
  },
16
  "pad_token": {
17
- "content": "<|finetune_right_pad_id|>",
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
@@ -9071,14 +9071,6 @@
9071
  "rstrip": false,
9072
  "single_word": false,
9073
  "special": true
9074
- },
9075
- "201134": {
9076
- "content": "<|finetune_right_pad_id|>",
9077
- "lstrip": false,
9078
- "normalized": false,
9079
- "rstrip": false,
9080
- "single_word": false,
9081
- "special": true
9082
  }
9083
  },
9084
  "bos_token": "<|begin_of_text|>",
@@ -9091,7 +9083,7 @@
9091
  "attention_mask"
9092
  ],
9093
  "model_max_length": 1048576,
9094
- "pad_token": "<|finetune_right_pad_id|>",
9095
  "processor_class": "Llama4Processor",
9096
  "tokenizer_class": "PreTrainedTokenizer"
9097
  }
 
9071
  "rstrip": false,
9072
  "single_word": false,
9073
  "special": true
 
 
 
 
 
 
 
 
9074
  }
9075
  },
9076
  "bos_token": "<|begin_of_text|>",
 
9083
  "attention_mask"
9084
  ],
9085
  "model_max_length": 1048576,
9086
+ "pad_token": "<|finetune_right_pad|>",
9087
  "processor_class": "Llama4Processor",
9088
  "tokenizer_class": "PreTrainedTokenizer"
9089
  }