Commit ·
362b32d
0
Parent(s):
SHADOW Assistant data
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- .gitattributes +693 -0
- AGENTS.md +37 -0
- GUIDE.md +97 -0
- LICENSE.md +19 -0
- README.md +259 -0
- ROADMAP.md +109 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.json +1 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.bin +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.counts.npz +3 -0
- arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.json +1 -0
- audio/sounds/audioset/g1_000_000.counts.npz +3 -0
- audio/sounds/audioset/g1_000_000.jsonl +0 -0
- audio/sounds/audioset/g1_000_000.npz +3 -0
- audio/sounds/audioset/g1_001_000.counts.npz +3 -0
- audio/sounds/audioset/g1_001_000.jsonl +0 -0
- audio/sounds/audioset/g1_001_000.npz +3 -0
- audio/sounds/audioset/g1_002_000.counts.npz +3 -0
- audio/sounds/audioset/g1_002_000.jsonl +0 -0
.gitattributes
ADDED
|
@@ -0,0 +1,693 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
| 2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
| 3 |
+
*.avro filter=lfs diff=lfs merge=lfs -text
|
| 4 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
| 5 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
| 6 |
+
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
| 7 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
| 8 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
| 9 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
| 10 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
| 11 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
| 12 |
+
*.lz4 filter=lfs diff=lfs merge=lfs -text
|
| 13 |
+
*.mds filter=lfs diff=lfs merge=lfs -text
|
| 14 |
+
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
| 15 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
| 16 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
| 17 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
| 18 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
| 19 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
| 20 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
| 21 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
| 22 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
| 23 |
+
*.pickle filter=lfs diff=lfs merge=lfs -text
|
| 24 |
+
*.pkl filter=lfs diff=lfs merge=lfs -text
|
| 25 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
| 26 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
| 27 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
| 28 |
+
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 29 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
| 30 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
| 31 |
+
*.tar filter=lfs diff=lfs merge=lfs -text
|
| 32 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
| 33 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
| 34 |
+
*.wasm filter=lfs diff=lfs merge=lfs -text
|
| 35 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
# Audio files - uncompressed
|
| 40 |
+
*.pcm filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
*.sam filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
*.raw filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
# Audio files - compressed
|
| 44 |
+
*.aac filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
*.flac filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
*.mp3 filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
*.ogg filter=lfs diff=lfs merge=lfs -text
|
| 48 |
+
*.wav filter=lfs diff=lfs merge=lfs -text
|
| 49 |
+
# Image files - uncompressed
|
| 50 |
+
*.bmp filter=lfs diff=lfs merge=lfs -text
|
| 51 |
+
*.gif filter=lfs diff=lfs merge=lfs -text
|
| 52 |
+
*.png filter=lfs diff=lfs merge=lfs -text
|
| 53 |
+
*.tiff filter=lfs diff=lfs merge=lfs -text
|
| 54 |
+
# Image files - compressed
|
| 55 |
+
*.jpg filter=lfs diff=lfs merge=lfs -text
|
| 56 |
+
*.jpeg filter=lfs diff=lfs merge=lfs -text
|
| 57 |
+
*.webp filter=lfs diff=lfs merge=lfs -text
|
| 58 |
+
# Video files - compressed
|
| 59 |
+
*.mp4 filter=lfs diff=lfs merge=lfs -text
|
| 60 |
+
*.webm filter=lfs diff=lfs merge=lfs -text
|
| 61 |
+
vision/images/m4_instruct/b01_TQA_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
vision/images/m4_instruct/b02_ALFRED_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 63 |
+
vision/images/m4_instruct/b02_ALFRED_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 64 |
+
vision/images/m4_instruct/b05_FlintstonesSV_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 65 |
+
vision/images/m4_instruct/b05_FlintstonesSV_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 66 |
+
vision/images/m4_instruct/b02_nextqa_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 67 |
+
vision/images/m4_instruct/b02_Spot-the-Diff_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 68 |
+
vision/images/m4_instruct/b00_contrastive_caption_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 69 |
+
vision/images/m4_instruct/b00_contrastive_caption_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 70 |
+
vision/images/m4_instruct/b00_contrastive_caption_002.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 71 |
+
vision/images/m4_instruct/b00_contrastive_caption_003.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 72 |
+
vision/images/m4_instruct/b03_RecipeQA_ImageCoherence_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 73 |
+
vision/images/m4_instruct/b05_Birds-to-Words_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 74 |
+
vision/images/m4_instruct/b09_WebQA_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 75 |
+
vision/images/m4_instruct/b09_star_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 76 |
+
vision/images/m4_instruct/b00_iconqa_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 77 |
+
vision/images/m4_instruct/b00_iconqa_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 78 |
+
vision/images/m4_instruct/b10_VIST_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 79 |
+
vision/images/m4_instruct/b10_VIST_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 80 |
+
vision/images/m4_instruct/b10_VIST_002.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 81 |
+
vision/images/m4_instruct/b10_VIST_003.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 82 |
+
vision/images/m4_instruct/b08_coinstruct_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 83 |
+
vision/images/m4_instruct/b08_coinstruct_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 84 |
+
vision/images/m4_instruct/b08_coinstruct_002.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 85 |
+
vision/images/m4_instruct/b08_coinstruct_003.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 86 |
+
vision/images/m4_instruct/b08_coinstruct_004.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 87 |
+
vision/images/m4_instruct/b08_coinstruct_005.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 88 |
+
vision/images/m4_instruct/b01_PororoSV_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 89 |
+
vision/images/m4_instruct/b08_VizWiz_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 90 |
+
vision/images/m4_instruct/b08_multi_vqa_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 91 |
+
vision/images/m4_instruct/b07_nlvr2_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 92 |
+
vision/images/m4_instruct/b07_nlvr2_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 93 |
+
vision/images/m4_instruct/b07_nlvr2_002.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 94 |
+
vision/images/m4_instruct/b07_nlvr2_003.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 95 |
+
vision/images/m4_instruct/b07_nlvr2_004.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 96 |
+
vision/images/m4_instruct/b07_nlvr2_005.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 97 |
+
vision/images/m4_instruct/b07_VISION_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 98 |
+
vision/images/m4_instruct/b02_AESOP_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 99 |
+
vision/images/finevision/objects365_qa/b07_train-00360-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 100 |
+
vision/images/finevision/objects365_qa/b09_train-00370-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 101 |
+
vision/images/finevision/objects365_qa/b08_train-00392-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 102 |
+
vision/images/finevision/objects365_qa/b03_train-00367-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 103 |
+
vision/images/finevision/objects365_qa/b08_train-00388-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 104 |
+
vision/images/finevision/objects365_qa/b10_train-00368-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 105 |
+
vision/images/finevision/objects365_qa/b00_train-00372-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 106 |
+
vision/images/finevision/objects365_qa/b00_train-00366-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 107 |
+
vision/images/finevision/objects365_qa/b06_train-00373-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 108 |
+
vision/images/finevision/objects365_qa/b03_train-00400-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 109 |
+
vision/images/finevision/objects365_qa/b07_train-00393-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 110 |
+
vision/images/finevision/objects365_qa/b00_train-00344-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 111 |
+
vision/images/finevision/objects365_qa/b03_train-00385-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 112 |
+
vision/images/finevision/objects365_qa/b03_train-00342-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 113 |
+
vision/images/finevision/objects365_qa/b06_train-00390-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 114 |
+
vision/images/finevision/objects365_qa/b01_train-00232-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 115 |
+
vision/images/finevision/objects365_qa/b03_train-00047-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 116 |
+
vision/images/finevision/objects365_qa/b07_train-00345-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 117 |
+
vision/images/finevision/objects365_qa/b05_train-00349-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 118 |
+
vision/images/finevision/objects365_qa/b06_train-00127-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 119 |
+
vision/images/finevision/objects365_qa/b00_train-00391-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 120 |
+
vision/images/finevision/objects365_qa/b03_train-00376-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 121 |
+
vision/images/finevision/objects365_qa/b09_train-00403-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 122 |
+
vision/images/finevision/objects365_qa/b06_train-00362-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 123 |
+
vision/images/finevision/objects365_qa/b01_train-00377-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 124 |
+
vision/images/finevision/objects365_qa/b01_train-00405-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 125 |
+
vision/images/finevision/objects365_qa/b04_train-00356-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 126 |
+
vision/images/finevision/objects365_qa/b01_train-00145-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 127 |
+
vision/images/finevision/objects365_qa/b03_train-00384-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 128 |
+
vision/images/finevision/objects365_qa/b01_train-00072-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 129 |
+
vision/images/finevision/objects365_qa/b03_train-00051-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 130 |
+
vision/images/finevision/objects365_qa/b05_train-00381-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 131 |
+
vision/images/finevision/objects365_qa/b03_train-00352-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 132 |
+
vision/images/finevision/objects365_qa/b03_train-00361-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 133 |
+
vision/images/finevision/objects365_qa/b06_train-00005-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 134 |
+
vision/images/finevision/objects365_qa/b09_train-00325-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 135 |
+
vision/images/finevision/objects365_qa/b10_train-00350-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 136 |
+
vision/images/finevision/objects365_qa/b01_train-00347-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 137 |
+
vision/images/finevision/objects365_qa/b03_train-00395-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 138 |
+
vision/images/finevision/objects365_qa/b10_train-00039-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 139 |
+
vision/images/finevision/objects365_qa/b00_train-00389-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 140 |
+
vision/images/finevision/objects365_qa/b01_train-00387-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 141 |
+
vision/images/finevision/objects365_qa/b07_train-00378-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 142 |
+
vision/images/finevision/objects365_qa/b07_train-00379-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 143 |
+
vision/images/finevision/objects365_qa/b03_train-00276-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 144 |
+
vision/images/finevision/objects365_qa/b10_train-00401-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 145 |
+
vision/images/finevision/objects365_qa/b05_train-00013-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 146 |
+
vision/images/finevision/objects365_qa/b08_train-00380-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 147 |
+
vision/images/finevision/objects365_qa/b08_train-00357-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 148 |
+
vision/images/finevision/objects365_qa/b06_train-00245-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 149 |
+
vision/images/finevision/objects365_qa/b08_train-00383-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 150 |
+
vision/images/finevision/objects365_qa/b10_train-00043-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 151 |
+
vision/images/finevision/objects365_qa/b07_train-00340-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 152 |
+
vision/images/finevision/objects365_qa/b08_train-00351-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 153 |
+
vision/images/finevision/objects365_qa/b10_train-00214-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 154 |
+
vision/images/finevision/objects365_qa/b08_train-00298-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 155 |
+
vision/images/finevision/objects365_qa/b04_train-00269-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 156 |
+
vision/images/finevision/objects365_qa/b01_train-00341-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 157 |
+
vision/images/finevision/objects365_qa/b05_train-00364-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 158 |
+
vision/images/finevision/objects365_qa/b07_train-00337-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 159 |
+
vision/images/finevision/objects365_qa/b02_train-00369-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 160 |
+
vision/images/finevision/objects365_qa/b04_train-00315-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 161 |
+
vision/images/finevision/objects365_qa/b05_train-00048-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 162 |
+
vision/images/finevision/objects365_qa/b06_train-00143-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 163 |
+
vision/images/finevision/objects365_qa/b07_train-00121-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 164 |
+
vision/images/finevision/objects365_qa/b05_train-00200-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 165 |
+
vision/images/finevision/objects365_qa/b01_train-00067-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 166 |
+
vision/images/finevision/objects365_qa/b06_train-00215-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 167 |
+
vision/images/finevision/objects365_qa/b10_train-00090-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 168 |
+
vision/images/finevision/vqaonbd/b04_train-00005-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 169 |
+
vision/images/finevision/objects365_qa/b04_train-00355-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 170 |
+
vision/images/finevision/objects365_qa/b06_train-00402-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 171 |
+
vision/images/finevision/objects365_qa/b03_train-00348-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 172 |
+
vision/images/finevision/objects365_qa/b05_train-00397-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 173 |
+
vision/images/finevision/objects365_qa/b07_train-00293-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 174 |
+
vision/images/finevision/objects365_qa/b08_train-00112-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 175 |
+
vision/images/finevision/objects365_qa/b09_train-00010-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 176 |
+
vision/images/finevision/objects365_qa/b05_train-00358-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 177 |
+
vision/images/finevision/objects365_qa/b06_train-00308-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 178 |
+
vision/images/finevision/objects365_qa/b08_train-00148-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 179 |
+
vision/images/finevision/objects365_qa/b09_train-00386-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 180 |
+
vision/images/finevision/objects365_qa/b00_train-00163-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 181 |
+
vision/images/finevision/objects365_qa/b02_train-00382-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 182 |
+
vision/images/finevision/objects365_qa/b00_train-00208-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 183 |
+
vision/images/finevision/objects365_qa/b03_train-00396-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 184 |
+
vision/images/finevision/objects365_qa/b01_train-00399-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 185 |
+
vision/images/finevision/objects365_qa/b03_train-00135-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 186 |
+
vision/images/finevision/objects365_qa/b05_train-00220-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 187 |
+
vision/images/finevision/objects365_qa/b07_train-00300-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 188 |
+
vision/images/finevision/objects365_qa/b04_train-00353-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 189 |
+
vision/images/finevision/objects365_qa/b07_train-00020-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 190 |
+
vision/images/finevision/objects365_qa/b07_train-00113-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 191 |
+
vision/images/finevision/objects365_qa/b09_train-00374-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 192 |
+
vision/images/finevision/objects365_qa/b01_train-00032-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 193 |
+
vision/images/finevision/objects365_qa/b05_train-00371-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 194 |
+
vision/images/finevision/objects365_qa/b07_train-00151-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 195 |
+
vision/images/finevision/objects365_qa/b08_train-00164-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 196 |
+
vision/images/finevision/objects365_qa/b09_train-00354-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 197 |
+
vision/images/finevision/objects365_qa/b01_train-00180-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 198 |
+
vision/images/finevision/objects365_qa/b05_train-00229-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 199 |
+
vision/images/finevision/objects365_qa/b00_train-00188-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 200 |
+
vision/images/finevision/objects365_qa/b05_train-00122-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 201 |
+
vision/images/finevision/objects365_qa/b08_train-00219-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 202 |
+
vision/images/finevision/objects365_qa/b09_train-00091-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 203 |
+
vision/images/finevision/objects365_qa/b10_train-00230-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 204 |
+
vision/images/finevision/objects365_qa/b05_train-00026-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 205 |
+
vision/images/finevision/objects365_qa/b06_train-00022-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 206 |
+
vision/images/finevision/objects365_qa/b06_train-00123-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 207 |
+
vision/images/finevision/objects365_qa/b00_train-00217-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 208 |
+
vision/images/finevision/objects365_qa/b01_train-00069-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 209 |
+
vision/images/finevision/objects365_qa/b07_train-00153-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 210 |
+
vision/images/finevision/objects365_qa/b09_train-00099-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 211 |
+
vision/images/finevision/objects365_qa/b10_train-00250-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 212 |
+
vision/images/finevision/objects365_qa/b06_train-00133-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 213 |
+
vision/images/finevision/objects365_qa/b10_train-00375-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 214 |
+
vision/images/finevision/objects365_qa/b00_train-00262-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 215 |
+
vision/images/finevision/objects365_qa/b07_train-00098-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 216 |
+
vision/images/finevision/objects365_qa/b00_train-00343-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 217 |
+
vision/images/finevision/objects365_qa/b03_train-00106-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 218 |
+
vision/images/finevision/objects365_qa/b05_train-00394-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 219 |
+
vision/images/finevision/vqaonbd/b05_train-00007-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 220 |
+
vision/images/finevision/objects365_qa/b00_train-00044-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 221 |
+
vision/images/finevision/objects365_qa/b01_train-00323-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 222 |
+
vision/images/finevision/objects365_qa/b04_train-00346-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 223 |
+
vision/images/finevision/objects365_qa/b08_train-00236-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 224 |
+
vision/images/finevision/objects365_qa/b00_train-00247-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 225 |
+
vision/images/finevision/objects365_qa/b04_train-00023-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 226 |
+
vision/images/finevision/objects365_qa/b05_train-00365-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 227 |
+
vision/images/finevision/objects365_qa/b01_train-00114-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 228 |
+
vision/images/finevision/objects365_qa/b06_train-00189-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 229 |
+
vision/images/finevision/objects365_qa/b10_train-00093-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 230 |
+
vision/images/finevision/objects365_qa/b05_train-00115-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 231 |
+
vision/images/finevision/objects365_qa/b10_train-00248-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 232 |
+
vision/images/finevision/objects365_qa/b02_train-00027-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 233 |
+
vision/images/finevision/objects365_qa/b07_train-00015-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 234 |
+
vision/images/finevision/objects365_qa/b06_train-00158-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 235 |
+
vision/images/finevision/objects365_qa/b07_train-00271-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 236 |
+
vision/images/finevision/objects365_qa/b09_train-00104-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 237 |
+
vision/images/finevision/objects365_qa/b09_train-00108-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 238 |
+
vision/images/finevision/vqaonbd/b04_train-00013-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 239 |
+
vision/images/finevision/objects365_qa/b07_train-00310-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 240 |
+
vision/images/finevision/objects365_qa/b08_train-00120-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 241 |
+
vision/images/finevision/objects365_qa/b08_train-00173-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 242 |
+
vision/images/finevision/objects365_qa/b09_train-00263-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 243 |
+
vision/images/finevision/vqaonbd/b03_train-00010-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 244 |
+
vision/images/finevision/objects365_qa/b06_train-00179-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 245 |
+
vision/images/finevision/objects365_qa/b06_train-00254-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 246 |
+
vision/images/finevision/objects365_qa/b07_train-00028-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 247 |
+
vision/images/finevision/objects365_qa/b08_train-00240-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 248 |
+
vision/images/finevision/objects365_qa/b01_train-00327-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 249 |
+
vision/images/finevision/objects365_qa/b06_train-00116-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 250 |
+
vision/images/finevision/objects365_qa/b08_train-00064-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 251 |
+
vision/images/finevision/objects365_qa/b08_train-00294-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 252 |
+
vision/images/finevision/objects365_qa/b01_train-00314-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 253 |
+
vision/images/finevision/objects365_qa/b05_train-00019-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 254 |
+
vision/images/finevision/objects365_qa/b06_train-00170-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 255 |
+
vision/images/finevision/objects365_qa/b08_train-00307-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 256 |
+
vision/images/finevision/Unichart/b07_train-00015-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 257 |
+
vision/images/finevision/objects365_qa/b05_train-00198-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 258 |
+
vision/images/finevision/objects365_qa/b06_train-00092-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 259 |
+
vision/images/finevision/objects365_qa/b01_train-00117-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 260 |
+
vision/images/finevision/objects365_qa/b07_train-00338-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 261 |
+
vision/images/finevision/objects365_qa/b09_train-00050-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 262 |
+
vision/images/finevision/objects365_qa/b01_train-00146-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 263 |
+
vision/images/finevision/objects365_qa/b04_train-00147-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 264 |
+
vision/images/finevision/objects365_qa/b00_train-00211-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 265 |
+
vision/images/finevision/objects365_qa/b05_train-00160-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 266 |
+
vision/images/finevision/objects365_qa/b09_train-00065-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 267 |
+
vision/images/finevision/Unichart/b01_train-00030-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 268 |
+
vision/images/finevision/objects365_qa/b00_train-00154-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 269 |
+
vision/images/finevision/objects365_qa/b05_train-00141-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 270 |
+
vision/images/finevision/objects365_qa/b09_train-00002-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 271 |
+
vision/images/finevision/objects365_qa/b04_train-00290-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 272 |
+
vision/images/finevision/objects365_qa/b05_train-00184-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 273 |
+
vision/images/finevision/objects365_qa/b07_train-00256-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 274 |
+
vision/images/finevision/objects365_qa/b09_train-00266-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 275 |
+
vision/images/finevision/objects365_qa/b10_train-00078-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 276 |
+
vision/images/finevision/objects365_qa/b03_train-00058-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 277 |
+
vision/images/finevision/objects365_qa/b05_train-00131-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 278 |
+
vision/images/finevision/objects365_qa/b08_train-00029-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 279 |
+
vision/images/finevision/objects365_qa/b07_train-00283-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 280 |
+
vision/images/finevision/objects365_qa/b07_train-00267-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 281 |
+
vision/images/finevision/objects365_qa/b01_train-00156-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 282 |
+
vision/images/finevision/objects365_qa/b03_train-00016-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 283 |
+
vision/images/finevision/objects365_qa/b05_train-00125-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 284 |
+
vision/images/finevision/objects365_qa/b07_train-00326-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 285 |
+
vision/images/finevision/objects365_qa/b01_train-00329-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 286 |
+
vision/images/finevision/objects365_qa/b07_train-00255-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 287 |
+
vision/images/finevision/objects365_qa/b09_train-00302-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 288 |
+
vision/images/finevision/objects365_qa/b04_train-00289-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 289 |
+
vision/images/finevision/objects365_qa/b05_train-00165-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 290 |
+
vision/images/finevision/objects365_qa/b06_train-00225-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 291 |
+
vision/images/finevision/objects365_qa/b08_train-00097-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 292 |
+
vision/images/finevision/objects365_qa/b09_train-00041-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 293 |
+
vision/images/finevision/objects365_qa/b10_train-00074-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 294 |
+
vision/images/finevision/objects365_qa/b05_train-00277-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 295 |
+
vision/images/finevision/objects365_qa/b10_train-00052-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 296 |
+
vision/images/finevision/objects365_qa/b10_train-00139-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 297 |
+
vision/images/finevision/objects365_qa/b02_train-00359-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 298 |
+
vision/images/finevision/objects365_qa/b09_train-00126-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 299 |
+
vision/images/finevision/objects365_qa/b00_train-00244-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 300 |
+
vision/images/finevision/objects365_qa/b00_train-00270-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 301 |
+
vision/images/finevision/objects365_qa/b02_train-00238-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 302 |
+
vision/images/finevision/vqaonbd/b04_train-00008-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 303 |
+
vision/images/finevision/objects365_qa/b01_train-00186-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 304 |
+
vision/images/finevision/objects365_qa/b07_train-00324-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 305 |
+
vision/images/finevision/objects365_qa/b05_train-00059-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 306 |
+
vision/images/finevision/objects365_qa/b09_train-00398-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 307 |
+
vision/images/finevision/objects365_qa/b05_train-00328-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 308 |
+
vision/images/finevision/objects365_qa/b00_train-00204-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 309 |
+
vision/images/finevision/objects365_qa/b02_train-00096-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 310 |
+
vision/images/finevision/objects365_qa/b06_train-00288-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 311 |
+
vision/images/finevision/objects365_qa/b07_train-00102-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 312 |
+
vision/images/finevision/objects365_qa/b09_train-00144-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 313 |
+
vision/images/finevision/objects365_qa/b10_train-00053-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 314 |
+
vision/images/finevision/objects365_qa/b10_train-00339-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 315 |
+
vision/images/finevision/objects365_qa/b03_train-00275-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 316 |
+
vision/images/finevision/objects365_qa/b05_train-00124-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 317 |
+
vision/images/finevision/objects365_qa/b07_train-00036-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 318 |
+
vision/images/finevision/objects365_qa/b10_train-00183-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 319 |
+
vision/images/finevision/objects365_qa/b02_train-00138-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 320 |
+
vision/images/finevision/objects365_qa/b04_train-00071-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 321 |
+
vision/images/finevision/objects365_qa/b06_train-00223-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 322 |
+
vision/images/finevision/objects365_qa/b09_train-00203-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 323 |
+
vision/images/finevision/objects365_qa/b10_train-00119-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 324 |
+
vision/images/finevision/objects365_qa/b00_train-00316-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 325 |
+
vision/images/finevision/objects365_qa/b03_train-00279-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 326 |
+
vision/images/finevision/objects365_qa/b00_train-00003-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 327 |
+
vision/images/finevision/objects365_qa/b00_train-00320-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 328 |
+
vision/images/finevision/objects365_qa/b07_train-00025-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 329 |
+
vision/images/finevision/objects365_qa/b10_train-00363-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 330 |
+
vision/images/finevision/vqaonbd/b01_train-00004-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 331 |
+
vision/images/finevision/objects365_qa/b00_train-00227-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 332 |
+
vision/images/finevision/objects365_qa/b05_train-00060-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 333 |
+
vision/images/finevision/objects365_qa/b08_train-00079-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 334 |
+
vision/images/finevision/objects365_qa/b09_train-00281-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 335 |
+
vision/images/finevision/vqaonbd/b08_train-00001-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 336 |
+
vision/images/finevision/objects365_qa/b02_train-00001-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 337 |
+
vision/images/finevision/objects365_qa/b10_train-00305-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 338 |
+
vision/images/finevision/objects365_qa/b02_train-00272-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 339 |
+
vision/images/finevision/objects365_qa/b06_train-00322-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 340 |
+
vision/images/finevision/objects365_qa/b07_train-00177-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 341 |
+
vision/images/finevision/objects365_qa/b09_train-00007-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 342 |
+
vision/images/finevision/objects365_qa/b08_train-00176-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 343 |
+
vision/images/finevision/objects365_qa/b02_train-00202-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 344 |
+
vision/images/finevision/objects365_qa/b04_train-00086-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 345 |
+
vision/images/finevision/objects365_qa/b06_train-00193-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 346 |
+
vision/images/finevision/objects365_qa/b07_train-00063-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 347 |
+
vision/images/finevision/objects365_qa/b10_train-00095-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 348 |
+
vision/images/finevision/vqaonbd/b06_train-00014-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 349 |
+
vision/images/finevision/objects365_qa/b02_train-00070-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 350 |
+
vision/images/finevision/objects365_qa/b09_train-00167-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 351 |
+
vision/images/finevision/objects365_qa/b02_train-00042-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 352 |
+
vision/images/finevision/objects365_qa/b03_train-00150-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 353 |
+
vision/images/finevision/objects365_qa/b09_train-00101-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 354 |
+
vision/images/finevision/objects365_qa/b03_train-00049-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 355 |
+
vision/images/finevision/objects365_qa/b05_train-00265-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 356 |
+
vision/images/finevision/objects365_qa/b02_train-00404-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 357 |
+
vision/images/finevision/objects365_qa/b03_train-00321-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 358 |
+
vision/images/finevision/objects365_qa/b05_train-00195-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 359 |
+
vision/images/finevision/mmc_instruct/b05_train-00008-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 360 |
+
vision/images/finevision/objects365_qa/b09_train-00012-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 361 |
+
vision/images/finevision/objects365_qa/b04_train-00162-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 362 |
+
vision/images/finevision/objects365_qa/b10_train-00187-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 363 |
+
vision/images/finevision/objects365_qa/b01_train-00197-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 364 |
+
vision/images/finevision/objects365_qa/b04_train-00008-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 365 |
+
vision/images/finevision/objects365_qa/b05_train-00192-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 366 |
+
vision/images/finevision/objects365_qa/b03_train-00234-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 367 |
+
vision/images/finevision/objects365_qa/b04_train-00014-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 368 |
+
vision/images/finevision/objects365_qa/b04_train-00284-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 369 |
+
vision/images/finevision/objects365_qa/b07_train-00243-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 370 |
+
vision/images/finevision/Unichart/b10_train-00000-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 371 |
+
vision/images/finevision/mmc_instruct/b08_train-00027-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 372 |
+
vision/images/finevision/objects365_qa/b03_train-00287-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 373 |
+
vision/images/finevision/objects365_qa/b07_train-00057-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 374 |
+
vision/images/finevision/objects365_qa/b05_train-00038-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 375 |
+
vision/images/finevision/objects365_qa/b04_train-00105-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 376 |
+
vision/images/finevision/objects365_qa/b05_train-00094-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 377 |
+
vision/images/finevision/objects365_qa/b06_train-00246-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 378 |
+
vision/images/finevision/objects365_qa/b08_train-00295-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 379 |
+
vision/images/finevision/objects365_qa/b10_train-00257-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 380 |
+
vision/images/finevision/objects365_qa/b02_train-00212-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 381 |
+
vision/images/finevision/objects365_qa/b04_train-00152-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 382 |
+
vision/images/finevision/objects365_qa/b06_train-00241-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 383 |
+
vision/images/finevision/objects365_qa/b06_train-00313-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 384 |
+
vision/images/finevision/objects365_qa/b04_train-00226-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 385 |
+
vision/images/finevision/objects365_qa/b08_train-00100-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 386 |
+
vision/images/finevision/objects365_qa/b09_train-00171-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 387 |
+
vision/images/finevision/objects365_qa/b01_train-00228-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 388 |
+
vision/images/finevision/objects365_qa/b03_train-00082-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 389 |
+
vision/images/finevision/objects365_qa/b07_train-00182-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 390 |
+
vision/images/finevision/objects365_qa/b00_train-00336-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 391 |
+
vision/images/finevision/objects365_qa/b02_train-00161-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 392 |
+
vision/images/finevision/objects365_qa/b09_train-00194-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 393 |
+
vision/images/finevision/objects365_qa/b04_train-00037-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 394 |
+
vision/images/finevision/objects365_qa/b02_train-00231-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 395 |
+
vision/images/finevision/objects365_qa/b02_train-00319-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 396 |
+
vision/images/finevision/objects365_qa/b08_train-00110-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 397 |
+
vision/images/finevision/objects365_qa/b09_train-00292-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 398 |
+
vision/images/finevision/objects365_qa/b01_train-00040-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 399 |
+
vision/images/finevision/objects365_qa/b09_train-00218-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 400 |
+
vision/images/finevision/objects365_qa/b09_train-00068-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 401 |
+
vision/images/finevision/objects365_qa/b00_train-00201-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 402 |
+
vision/images/finevision/objects365_qa/b06_train-00129-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 403 |
+
vision/images/finevision/objects365_qa/b07_train-00199-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 404 |
+
vision/images/finevision/objects365_qa/b09_train-00031-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 405 |
+
vision/images/finevision/objects365_qa/b09_train-00080-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 406 |
+
vision/images/finevision/objects365_qa/b06_train-00291-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 407 |
+
vision/images/finevision/objects365_qa/b02_train-00216-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 408 |
+
vision/images/finevision/objects365_qa/b04_train-00107-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 409 |
+
vision/images/finevision/objects365_qa/b07_train-00155-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 410 |
+
vision/images/finevision/objects365_qa/b08_train-00309-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 411 |
+
vision/images/finevision/objects365_qa/b10_train-00061-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 412 |
+
vision/images/finevision/mmc_instruct/b00_train-00014-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 413 |
+
vision/images/finevision/objects365_qa/b02_train-00000-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 414 |
+
vision/images/finevision/objects365_qa/b04_train-00128-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 415 |
+
vision/images/finevision/objects365_qa/b04_train-00172-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 416 |
+
vision/images/finevision/objects365_qa/b05_train-00253-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 417 |
+
vision/images/finevision/objects365_qa/b07_train-00296-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 418 |
+
vision/images/finevision/objects365_qa/b08_train-00134-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 419 |
+
vision/images/finevision/objects365_qa/b09_train-00251-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 420 |
+
vision/images/finevision/objects365_qa/b10_train-00018-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 421 |
+
vision/images/finevision/objects365_qa/b04_train-00033-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 422 |
+
vision/images/finevision/objects365_qa/b05_train-00109-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 423 |
+
vision/images/finevision/objects365_qa/b07_train-00333-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 424 |
+
vision/images/finevision/objects365_qa/b08_train-00206-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 425 |
+
vision/images/finevision/objects365_qa/b05_train-00035-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 426 |
+
vision/images/finevision/objects365_qa/b06_train-00076-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 427 |
+
vision/images/finevision/Unichart/b03_train-00008-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 428 |
+
vision/images/finevision/objects365_qa/b07_train-00221-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 429 |
+
vision/images/finevision/objects365_qa/b09_train-00169-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 430 |
+
vision/images/finevision/vqaonbd/b01_train-00002-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 431 |
+
vision/images/finevision/mmc_instruct/b01_train-00013-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 432 |
+
vision/images/finevision/mmc_instruct/b09_train-00020-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 433 |
+
vision/images/finevision/objects365_qa/b01_train-00004-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 434 |
+
vision/images/finevision/objects365_qa/b02_train-00249-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 435 |
+
vision/images/finevision/objects365_qa/b03_train-00011-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 436 |
+
vision/images/finevision/objects365_qa/b04_train-00083-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 437 |
+
vision/images/finevision/objects365_qa/b07_train-00274-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 438 |
+
vision/images/finevision/objects365_qa/b09_train-00280-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 439 |
+
vision/images/finevision/mmc_instruct/b05_train-00006-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 440 |
+
vision/images/finevision/objects365_qa/b02_train-00168-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 441 |
+
vision/images/finevision/objects365_qa/b07_train-00056-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 442 |
+
vision/images/finevision/objects365_qa/b04_train-00318-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 443 |
+
vision/images/finevision/objects365_qa/b02_train-00242-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 444 |
+
vision/images/finevision/objects365_qa/b08_train-00210-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 445 |
+
vision/images/finevision/objects365_qa/b04_train-00224-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 446 |
+
vision/images/finevision/objects365_qa/b05_train-00055-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 447 |
+
vision/images/finevision/objects365_qa/b07_train-00006-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 448 |
+
vision/images/finevision/objects365_qa/b08_train-00207-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 449 |
+
vision/images/finevision/vqaonbd/b01_train-00012-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 450 |
+
vision/images/finevision/objects365_qa/b02_train-00073-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 451 |
+
vision/images/finevision/mmc_instruct/b00_train-00033-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 452 |
+
vision/images/finevision/objects365_qa/b03_train-00066-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 453 |
+
vision/images/finevision/objects365_qa/b00_train-00261-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 454 |
+
vision/images/finevision/objects365_qa/b08_train-00137-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 455 |
+
vision/images/finevision/objects365_qa/b09_train-00046-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 456 |
+
vision/images/finevision/mmc_instruct/b10_train-00003-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 457 |
+
vision/images/finevision/objects365_qa/b07_train-00286-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 458 |
+
vision/images/finevision/mapqa/b10_train-00004-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 459 |
+
vision/images/finevision/mmc_instruct/b05_train-00001-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 460 |
+
vision/images/finevision/objects365_qa/b02_train-00175-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 461 |
+
vision/images/finevision/objects365_qa/b03_train-00331-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 462 |
+
vision/images/finevision/objects365_qa/b08_train-00235-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 463 |
+
vision/images/finevision/objects365_qa/b09_train-00317-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 464 |
+
vision/images/finevision/objects365_qa/b00_train-00260-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 465 |
+
vision/images/finevision/mmc_instruct/b09_train-00025-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 466 |
+
vision/images/finevision/objects365_qa/b05_train-00159-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 467 |
+
vision/images/finevision/mmc_instruct/b09_train-00010-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 468 |
+
vision/images/finevision/objects365_qa/b00_train-00178-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 469 |
+
vision/images/finevision/vqaonbd/b08_train-00003-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 470 |
+
vision/images/finevision/vqaonbd/b09_train-00015-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 471 |
+
vision/images/finevision/mapqa/b07_train-00002-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 472 |
+
vision/images/finevision/objects365_qa/b07_train-00034-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 473 |
+
vision/images/finevision/objects365_qa/b02_train-00330-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 474 |
+
vision/images/finevision/objects365_qa/b04_train-00191-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 475 |
+
vision/images/finevision/objects365_qa/b10_train-00174-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 476 |
+
vision/images/finevision/objects365_qa/b03_train-00312-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 477 |
+
vision/images/finevision/mmc_instruct/b08_train-00028-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 478 |
+
vision/images/finevision/objects365_qa/b02_train-00181-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 479 |
+
vision/images/finevision/objects365_qa/b04_train-00332-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 480 |
+
vision/images/finevision/vqaonbd/b01_train-00009-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 481 |
+
vision/images/finevision/objects365_qa/b03_train-00268-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 482 |
+
vision/images/finevision/objects365_qa/b07_train-00084-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 483 |
+
vision/images/finevision/objects365_qa/b03_train-00140-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 484 |
+
vision/images/finevision/objects365_qa/b04_train-00062-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 485 |
+
vision/images/finevision/mmc_instruct/b07_train-00018-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 486 |
+
vision/images/finevision/objects365_qa/b03_train-00166-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 487 |
+
vision/images/finevision/objects365_qa/b03_train-00264-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 488 |
+
vision/images/finevision/vqaonbd/b08_train-00006-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 489 |
+
vision/images/finevision/objects365_qa/b01_train-00213-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 490 |
+
vision/images/finevision/objects365_qa/b02_train-00075-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 491 |
+
vision/images/finevision/objects365_qa/b05_train-00132-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 492 |
+
vision/images/finevision/objects365_qa/b09_train-00273-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 493 |
+
vision/images/finevision/objects365_qa/b03_train-00297-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 494 |
+
vision/images/finevision/mmc_instruct/b09_train-00031-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 495 |
+
vision/images/finevision/objects365_qa/b02_train-00136-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 496 |
+
vision/images/finevision/objects365_qa/b02_train-00205-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 497 |
+
vision/images/finevision/Unichart/b09_train-00023-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 498 |
+
vision/images/finevision/mapqa/b10_train-00000-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 499 |
+
vision/images/finevision/mmc_instruct/b05_train-00016-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 500 |
+
vision/images/finevision/mmc_instruct/b04_train-00024-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 501 |
+
vision/images/finevision/objects365_qa/b01_train-00252-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 502 |
+
vision/images/finevision/objects365_qa/b02_train-00285-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 503 |
+
vision/images/finevision/mmc_instruct/b10_train-00032-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 504 |
+
vision/images/finevision/Unichart/b06_train-00012-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 505 |
+
vision/images/finevision/mmc_instruct/b01_train-00030-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 506 |
+
vision/images/finevision/objects365_qa/b02_train-00081-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 507 |
+
vision/images/finevision/mapqa/b04_train-00006-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 508 |
+
vision/images/finevision/mmc_instruct/b04_train-00004-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 509 |
+
vision/images/finevision/mmc_instruct/b05_train-00023-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 510 |
+
vision/images/finevision/Unichart/b06_train-00010-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 511 |
+
vision/images/finevision/objects365_qa/b02_train-00087-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 512 |
+
vision/images/finevision/objects365_qa/b02_train-00222-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 513 |
+
vision/images/finevision/Unichart/b00_train-00025-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 514 |
+
vision/images/finevision/mmc_instruct/b07_train-00029-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 515 |
+
vision/images/finevision/mmc_instruct/b07_train-00026-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 516 |
+
vision/images/finevision/objects365_qa/b01_train-00239-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 517 |
+
vision/images/finevision/mmc_instruct/b03_train-00015-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 518 |
+
vision/images/finevision/mmc_instruct/b02_train-00009-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 519 |
+
vision/images/finevision/objects365_qa/b05_train-00157-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 520 |
+
vision/images/finevision/mmc_instruct/b05_train-00019-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 521 |
+
vision/images/finevision/objects365_qa/b01_train-00303-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 522 |
+
vision/images/finevision/k12_printing/b06_train-00009-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 523 |
+
vision/images/finevision/mmc_instruct/b03_train-00012-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 524 |
+
vision/images/finevision/mmc_instruct/b04_train-00005-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 525 |
+
vision/images/finevision/Unichart/b10_train-00022-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 526 |
+
vision/images/finevision/mmc_instruct/b01_train-00022-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 527 |
+
vision/images/finevision/objects365_qa/b02_train-00282-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 528 |
+
vision/images/finevision/vqaonbd/b06_train-00011-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 529 |
+
vision/images/finevision/latexformulas/b08_train-00005-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 530 |
+
vision/images/finevision/mmc_instruct/b04_train-00007-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 531 |
+
vision/images/finevision/k12_printing/b06_train-00005-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 532 |
+
vision/images/finevision/mmc_instruct/b05_train-00011-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 533 |
+
vision/images/finevision/latexformulas/b08_train-00003-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 534 |
+
vision/images/finevision/mapqa/b04_train-00005-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 535 |
+
vision/images/finevision/mmc_instruct/b02_train-00000-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 536 |
+
vision/images/finevision/mmc_instruct/b03_train-00021-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 537 |
+
vision/images/finevision/objects365_qa/b09_train-00209-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 538 |
+
vision/images/finevision/objects365_qa/b02_train-00334-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 539 |
+
vision/images/finevision/figureqa/b06_train-00002-of-00005_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 540 |
+
vision/images/finevision/mapqa/b03_train-00001-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 541 |
+
vision/images/finevision/objects365_qa/b02_train-00149-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 542 |
+
vision/images/finevision/latexformulas/b08_train-00004-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 543 |
+
vision/images/finevision/objects365_qa/b02_train-00017-of-00406_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 544 |
+
vision/images/finevision/tal_ocr_eng/b00_train-00002-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 545 |
+
vision/images/finevision/Unichart/b05_train-00019-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 546 |
+
vision/images/finevision/tal_ocr_eng/b10_train-00001-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 547 |
+
vision/images/finevision/mapqa/b07_train-00003-of-00007_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 548 |
+
vision/images/finevision/Unichart/b06_train-00036-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 549 |
+
vision/images/finevision/tal_ocr_eng/b10_train-00005-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 550 |
+
vision/images/finevision/tal_ocr_eng/b08_train-00008-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 551 |
+
vision/images/finevision/Unichart/b09_train-00001-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 552 |
+
vision/images/finevision/latexformulas/b05_train-00011-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 553 |
+
vision/images/finevision/Unichart/b07_train-00027-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 554 |
+
vision/images/finevision/k12_printing/b08_train-00006-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 555 |
+
vision/images/finevision/tal_ocr_eng/b10_train-00006-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 556 |
+
vision/images/finevision/Unichart/b01_train-00034-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 557 |
+
vision/images/finevision/k12_printing/b00_train-00003-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 558 |
+
vision/images/finevision/figureqa/b08_train-00001-of-00005_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 559 |
+
vision/images/finevision/latexformulas/b09_train-00002-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 560 |
+
vision/images/finevision/SynthFormulaNet/b06_train-00001-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 561 |
+
vision/images/finevision/SynthFormulaNet/b06_train-00001-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 562 |
+
vision/images/finevision/Unichart/b04_train-00017-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 563 |
+
vision/images/finevision/vqaonbd/b02_train-00000-of-00016_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 564 |
+
vision/images/finevision/Unichart/b08_train-00011-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 565 |
+
vision/images/finevision/latexformulas/b07_train-00009-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 566 |
+
vision/images/finevision/Unichart/b04_train-00032-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 567 |
+
vision/images/finevision/Unichart/b10_train-00016-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 568 |
+
vision/images/finevision/dvqa/b06_train-00005-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 569 |
+
vision/images/finevision/latexformulas/b01_train-00008-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 570 |
+
vision/images/finevision/latexformulas/b03_train-00007-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 571 |
+
vision/images/finevision/Unichart/b04_train-00003-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 572 |
+
vision/images/finevision/latexformulas/b05_train-00010-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 573 |
+
vision/images/finevision/Unichart/b08_train-00021-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 574 |
+
vision/images/finevision/Unichart/b09_train-00002-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 575 |
+
vision/images/finevision/datik/b00_train-00006-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 576 |
+
vision/images/finevision/mmc_instruct/b02_train-00002-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 577 |
+
vision/images/finevision/datik/b10_train-00007-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 578 |
+
vision/images/finevision/latexformulas/b03_train-00001-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 579 |
+
vision/images/finevision/Unichart/b05_train-00004-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 580 |
+
vision/images/finevision/alfworldgpt/b06_train-00004-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 581 |
+
vision/images/finevision/k12_printing/b09_train-00002-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 582 |
+
vision/images/finevision/alfworldgpt/b06_train-00000-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 583 |
+
vision/images/finevision/dvqa/b00_train-00004-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 584 |
+
vision/images/finevision/figureqa/b10_train-00000-of-00005_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 585 |
+
vision/images/finevision/latexformulas/b01_train-00000-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 586 |
+
vision/images/finevision/latexformulas/b07_train-00006-of-00012_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 587 |
+
vision/images/finevision/mmc_instruct/b02_train-00017-of-00034_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 588 |
+
vision/images/finevision/dvqa/b08_train-00000-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 589 |
+
vision/images/finevision/dvqa/b08_train-00009-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 590 |
+
vision/images/finevision/iconqa/b06_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 591 |
+
vision/images/finevision/tal_ocr_eng/b07_train-00003-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 592 |
+
vision/images/finevision/Unichart/b09_train-00018-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 593 |
+
vision/images/finevision/figureqa/b03_train-00004-of-00005_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 594 |
+
vision/images/finevision/tal_ocr_eng/b01_train-00007-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 595 |
+
vision/images/finevision/Unichart/b05_train-00020-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 596 |
+
vision/images/finevision/k12_printing/b03_train-00001-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 597 |
+
vision/images/finevision/tal_ocr_eng/b01_train-00004-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 598 |
+
vision/images/finevision/Unichart/b04_train-00024-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 599 |
+
vision/images/finevision/k12_printing/b05_train-00004-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 600 |
+
vision/images/finevision/Unichart/b07_train-00035-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 601 |
+
vision/images/finevision/SynthFormulaNet/b00_train-00005-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 602 |
+
vision/images/finevision/SynthFormulaNet/b00_train-00005-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 603 |
+
vision/images/finevision/SynthFormulaNet/b10_train-00003-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 604 |
+
vision/images/finevision/SynthFormulaNet/b10_train-00003-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 605 |
+
vision/images/finevision/tal_ocr_eng/b05_train-00000-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 606 |
+
vision/images/finevision/Unichart/b07_train-00031-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 607 |
+
vision/images/finevision/k12_printing/b01_train-00007-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 608 |
+
vision/images/finevision/Unichart/b03_train-00026-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 609 |
+
vision/images/finevision/dvqa/b09_train-00001-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 610 |
+
vision/images/finevision/dvqa/b04_train-00007-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 611 |
+
vision/images/finevision/Unichart/b01_train-00007-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 612 |
+
vision/images/finevision/datik/b09_train-00005-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 613 |
+
vision/images/finevision/datik/b07_train-00000-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 614 |
+
vision/images/finevision/k12_printing/b01_train-00008-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 615 |
+
vision/images/finevision/CoSyn_400k_music/b00_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 616 |
+
vision/images/finevision/Unichart/b03_train-00028-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 617 |
+
vision/images/finevision/dvqa/b09_train-00008-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 618 |
+
vision/images/finevision/Unichart/b05_train-00009-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 619 |
+
vision/images/finevision/datik/b07_train-00002-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 620 |
+
vision/images/finevision/Unichart/b01_train-00033-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 621 |
+
vision/images/finevision/dvqa/b09_train-00003-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 622 |
+
vision/images/finevision/figureqa/b07_train-00003-of-00005_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 623 |
+
vision/images/finevision/sroie/b04_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 624 |
+
vision/images/finevision/Unichart/b05_train-00029-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 625 |
+
vision/images/finevision/alfworldgpt/b04_train-00006-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 626 |
+
vision/images/finevision/captcha/b00_train-00002-of-00003_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 627 |
+
vision/images/finevision/Unichart/b01_train-00013-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 628 |
+
vision/images/finevision/dvqa/b07_train-00006-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 629 |
+
vision/images/finevision/alfworldgpt/b09_train-00007-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 630 |
+
vision/images/finevision/alfworldgpt/b09_train-00003-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 631 |
+
vision/images/finevision/Unichart/b05_train-00014-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 632 |
+
vision/images/finevision/CoSyn_400k_graphic/b00_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 633 |
+
vision/images/finevision/CoSyn_400k_graphic/b00_train-00000-of-00001_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 634 |
+
vision/images/finevision/alfworldgpt/b07_train-00001-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 635 |
+
vision/images/finevision/plotqa/b00_train-00011-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 636 |
+
vision/images/finevision/plotqa/b09_train-00016-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 637 |
+
vision/images/finevision/Unichart/b02_train-00005-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 638 |
+
vision/images/finevision/datik/b05_train-00004-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 639 |
+
vision/images/finevision/dvqa/b01_train-00002-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 640 |
+
vision/images/finevision/plotqa/b00_train-00005-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 641 |
+
vision/images/finevision/plotqa/b09_train-00010-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 642 |
+
vision/images/finevision/datik/b05_train-00003-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 643 |
+
vision/video/sharegpt4video/b01_panda_videos_10_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 644 |
+
vision/images/finevision/CoSyn_400k_chemical/b00_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 645 |
+
vision/video/sharegpt4video/b01_panda_videos_2_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 646 |
+
vision/video/sharegpt4video/b06_panda_videos_11_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 647 |
+
vision/video/sharegpt4video/b06_panda_videos_7_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 648 |
+
vision/images/finevision/SynthFormulaNet/b03_train-00002-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 649 |
+
vision/images/finevision/SynthFormulaNet/b03_train-00002-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 650 |
+
vision/images/finevision/tal_ocr_eng/b02_train-00009-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 651 |
+
vision/images/finevision/datik/b05_train-00001-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 652 |
+
vision/video/sharegpt4video/b06_bdd100k_videos_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 653 |
+
vision/video/sharegpt4video/b08_panda_videos_18_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 654 |
+
vision/video/sharegpt4video/b08_panda_videos_15_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 655 |
+
vision/images/finevision/k12_printing/b02_train-00000-of-00010_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 656 |
+
vision/images/finevision/SynthFormulaNet/b01_train-00000-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 657 |
+
vision/images/finevision/SynthFormulaNet/b01_train-00000-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 658 |
+
vision/images/finevision/captcha/b05_train-00001-of-00003_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 659 |
+
vision/images/finevision/alfworldgpt/b05_train-00005-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 660 |
+
vision/images/finevision/tabmwp/b03_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 661 |
+
vision/video/sharegpt4video/b07_panda_videos_1_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 662 |
+
vision/video/sharegpt4video/b07_panda_videos_4_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 663 |
+
vision/images/finevision/alfworldgpt/b01_train-00002-of-00008_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 664 |
+
vision/images/finevision/plotqa/b03_train-00012-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 665 |
+
vision/images/finevision/Unichart/b02_train-00006-of-00037_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 666 |
+
vision/images/finevision/plotqa/b03_train-00009-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 667 |
+
vision/video/sharegpt4video/b04_panda_videos_14_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 668 |
+
vision/images/finevision/plotqa/b03_train-00006-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 669 |
+
vision/video/sharegpt4video/b04_panda_videos_5_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 670 |
+
vision/video/sharegpt4video/b09_panda_videos_19_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 671 |
+
vision/video/sharegpt4video/b09_panda_videos_17_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 672 |
+
vision/video/sharegpt4video/b10_panda_videos_16_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 673 |
+
vision/video/sharegpt4video/b05_panda_videos_6_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 674 |
+
vision/video/sharegpt4video/b10_panda_videos_20_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 675 |
+
vision/video/sharegpt4video/b05_panda_videos_13_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 676 |
+
vision/video/sharegpt4video/b03_panda_videos_8_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 677 |
+
vision/video/sharegpt4video/b03_panda_videos_9_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 678 |
+
vision/images/finevision/SynthFormulaNet/b02_train-00004-of-00006_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 679 |
+
vision/images/finevision/SynthFormulaNet/b02_train-00004-of-00006_001.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 680 |
+
vision/video/sharegpt4video/b02_panda_videos_3_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 681 |
+
vision/video/sharegpt4video/b02_panda_videos_12_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 682 |
+
vision/images/finevision/CoSyn_400k_circuit/b02_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 683 |
+
vision/images/finevision/captcha/b02_train-00000-of-00003_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 684 |
+
vision/images/finevision/tabmwp(mathv360k)/b10_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 685 |
+
vision/images/finevision/plotqa/b10_train-00002-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 686 |
+
vision/images/finevision/iconqa(mathv360k)/b09_train-00000-of-00001_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 687 |
+
vision/images/finevision/plotqa/b10_train-00013-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 688 |
+
vision/images/finevision/plotqa/b04_train-00007-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 689 |
+
vision/images/finevision/plotqa/b06_train-00000-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 690 |
+
vision/images/finevision/datikz/b04_train-00000-of-00002_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 691 |
+
vision/images/finevision/plotqa/b08_train-00004-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 692 |
+
vision/images/finevision/datikz/b04_train-00001-of-00002_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
| 693 |
+
vision/images/finevision/plotqa/b08_train-00015-of-00018_000.jsonl filter=lfs diff=lfs merge=lfs -text
|
AGENTS.md
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# For AI agents using this dataset
|
| 2 |
+
|
| 3 |
+
Instructions for coding assistants and automated pipelines that read, train on or describe this data.
|
| 4 |
+
|
| 5 |
+
## Read first
|
| 6 |
+
|
| 7 |
+
- [README.md](README.md): where everything is, file formats, a loading example.
|
| 8 |
+
- [GUIDE.md](GUIDE.md): what the data is for and why it is built this way.
|
| 9 |
+
|
| 10 |
+
## Reading the files
|
| 11 |
+
|
| 12 |
+
- Text shards: 3 bytes per token, little-endian, token id = b0 | b1 << 8 | b2 << 16. Documents start with token 2
|
| 13 |
+
(`<bos>`) and end with token 1 (`<eos>`). Tokenizer: Gemma, 262,144 tokens (EmbeddingGemma 2).
|
| 14 |
+
- Picture, video and audio shards: `.npz` with `codes` (uint8, 64 bytes = 512 bits per token) and an index array
|
| 15 |
+
(`img` for pictures, `clips` for audio); the matching `.jsonl` holds the text.
|
| 16 |
+
- Tuning: `sft/<set>/part_NNN.jsonl.zst`, one JSON conversation per line.
|
| 17 |
+
- A large source may be split into `part_NN/` folders; treat them as one source.
|
| 18 |
+
- The repository holds more than 50,000 files: list one folder at a time rather than the whole repository.
|
| 19 |
+
|
| 20 |
+
## Keep evaluation clean
|
| 21 |
+
|
| 22 |
+
- Do not pretrain on the held-out tuning shards listed in `sft/vision/vision_sft_manifest.json` and `sft/audio/audio_sft_manifest.json`.
|
| 23 |
+
- Apply the skip lists in `decontamination/` before training if you report HellaSwag, ESC-50, UrbanSound8K, VGGSound,
|
| 24 |
+
or any of the nine text benchmarks scanned there.
|
| 25 |
+
- FineVision includes about 1.02% images that match vision benchmark test images (reported by its authors); say so when
|
| 26 |
+
reporting vision benchmarks.
|
| 27 |
+
|
| 28 |
+
## Licence
|
| 29 |
+
|
| 30 |
+
- Non-commercial research and education only, with attribution (LICENSE.md), and each source keeps its own terms.
|
| 31 |
+
- Sets in `rebuild/` must be rebuilt from their original source under that source's terms; do not repost them.
|
| 32 |
+
- Credit: "Training data from SHADOW Assistant by QLNI, encoded with compute from CloudRift (https://cloudrift.ai)."
|
| 33 |
+
|
| 34 |
+
## When describing this data
|
| 35 |
+
|
| 36 |
+
- Quote the numbers in the README; do not estimate new ones.
|
| 37 |
+
- Do not describe any part as raw images or audio: only codes and text are published.
|
GUIDE.md
ADDED
|
@@ -0,0 +1,97 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Guide: how SHADOW Assistant is built, and why
|
| 2 |
+
|
| 3 |
+
This guide explains the design behind the data in this repository: what the model is, how each kind of data is used,
|
| 4 |
+
and the reason for every major choice. For where files are, see the [README](README.md).
|
| 5 |
+
|
| 6 |
+
## The model in one picture
|
| 7 |
+
|
| 8 |
+
```
|
| 9 |
+
text ── tokenizer ───────────┐
|
| 10 |
+
photo / video ── eye ────────┤ every input becomes 512-bit codes by one fixed rule
|
| 11 |
+
sound ── ear ────────────────┘ │
|
| 12 |
+
input adapter
|
| 13 |
+
│
|
| 14 |
+
body: 24 ternary layers = 21 RNN layers (fixed-size state) + 3 global attention layers
|
| 15 |
+
Engram: 4B n-gram memory rows on disk, text only
|
| 16 |
+
│
|
| 17 |
+
readout over the same frozen codes + a word-frequency prior
|
| 18 |
+
│
|
| 19 |
+
token
|
| 20 |
+
```
|
| 21 |
+
|
| 22 |
+
About 135M parameters in the body (about 30 MB at ternary), a 4B-row Engram (about 2 GB at int4, read from disk), and
|
| 23 |
+
an eye and ear (EmbeddingGemma 2's vision and audio towers at int8) that load only when a picture or sound arrives.
|
| 24 |
+
|
| 25 |
+
## The decisions, and why
|
| 26 |
+
|
| 27 |
+
**One frozen code space for everything.** Every token, picture patch and sound frame becomes a 512-bit code from
|
| 28 |
+
EmbeddingGemma 2's space, and the token table is never trained. Our paper ([Embeddings Are Overpriced?](https://arxiv.org/abs/2610.22428))
|
| 29 |
+
showed a frozen table with a learned frequency prior beats a learned table. Because the space never moves, words,
|
| 30 |
+
pictures and sound stay aligned for good: no connector is needed between them, a model trained on text alone can
|
| 31 |
+
already read picture codes it has never seen, and new words can be added later. The table is 16.8 MB.
|
| 32 |
+
|
| 33 |
+
**A ternary body.** Weights are -1, 0 or +1 (BitNet b1.58), activations int8. The body is about 30 MB and fast on any
|
| 34 |
+
CPU, and in our tests it learns as well as needed at this size.
|
| 35 |
+
|
| 36 |
+
**An Engram.** A large table of 2- and 3-token memories, hashed, read from disk a few rows per token. It stores facts
|
| 37 |
+
and phrases cheaply, so the small body can spend itself on reasoning. It reads text only, never picture or sound codes.
|
| 38 |
+
|
| 39 |
+
**Mostly RNN layers, three attention layers.** 21 RNN layers keep a fixed-size state whatever the length, and only 3
|
| 40 |
+
global attention layers keep keys and values. Writing speed on a laptop CPU barely drops with context (551 tokens a
|
| 41 |
+
second empty, 440 at 12k, 141 at 128k), the memory for the window stays small, and the RNN state (26 KB) can be saved
|
| 42 |
+
to resume a conversation instantly.
|
| 43 |
+
|
| 44 |
+
**Train exactly as it runs.** Keys and values are stored at int4 on the CPU, so training uses int4 too. What the model
|
| 45 |
+
learns is exactly what it runs.
|
| 46 |
+
|
| 47 |
+
**Memory in four parts.** A small model cannot hold a long conversation or a library in its weights, so SHADOW splits
|
| 48 |
+
memory by job:
|
| 49 |
+
|
| 50 |
+
- **The gist.** The 21 RNN layers each carry a fixed-size state that is updated with every token. Together it is
|
| 51 |
+
about 26 KB, the same after ten words or a whole book: a running summary of who, what and why. It is not exact, but
|
| 52 |
+
it never grows, and saving it to disk lets a conversation resume in milliseconds without reading anything again.
|
| 53 |
+
- **The window.** The 3 global attention layers see the last 32k tokens exactly, with keys and values stored at int4.
|
| 54 |
+
This is where precise wording, numbers and recent turns live.
|
| 55 |
+
- **The archive and the index.** When text leaves the window it is written to disk once, in 64-token blocks: the token
|
| 56 |
+
ids plus the int4 keys and values of the 3 attention layers (about 0.8 KB per token). When a question arrives, a
|
| 57 |
+
word-match search over the archive finds candidate blocks, and a small trained index reranks them by meaning. The
|
| 58 |
+
best 2,048 tokens are loaded straight back into the attention layers, placed just before the question with the best
|
| 59 |
+
block last, so the model reads them as if they had just been said. The gist is not touched by a fetch. The index
|
| 60 |
+
learns from where the attention layers actually look, and from practice: during training, facts are placed early in
|
| 61 |
+
long documents and asked about later, after they have left the window. In a test on a model trained on 0.85B tokens,
|
| 62 |
+
fetching lifted recall of facts read about 1,400 tokens earlier from 9% to 63% of digits.
|
| 63 |
+
- **The Engram.** About 4B small rows, each tied to a hashed 2- or 3-token phrase, read from disk a few per token. It
|
| 64 |
+
stores common facts and phrasing cheaply, so the small body can spend itself on reasoning. It reads text only.
|
| 65 |
+
|
| 66 |
+
Each part is taught by different data: long documents teach the gist and the window, archive practice teaches the
|
| 67 |
+
index, and all the text fills the Engram.
|
| 68 |
+
|
| 69 |
+
**All senses trained together.** Text, pictures, video and sound go into one body in one run, rather than attaching
|
| 70 |
+
vision to a finished text model. Caption text (about 10B tokens) describes physical, everyday scenes, which is exactly
|
| 71 |
+
what small models lack.
|
| 72 |
+
|
| 73 |
+
**Australia.** About 34B tokens of Australian text (web, law, Parliament, biography, Wikipedia), Australian speech,
|
| 74 |
+
Australian photos of plants and animals, and Australian everyday conversations, on top of a strong global base.
|
| 75 |
+
|
| 76 |
+
**Context.** Trained natively to 32k tokens; we have about 6.6B tokens in documents of 32k or more. Extension to 128k
|
| 77 |
+
uses YaRN at run time and will be published with its test results.
|
| 78 |
+
|
| 79 |
+
**Tuning after pretraining.** Tuning teaches the model to answer, not what to know, so it always comes after
|
| 80 |
+
pretraining. The picture, video and sound tuning mix follows SmolVLM's published findings: a separate text share, very
|
| 81 |
+
little chain-of-thought, and loss on answers only.
|
| 82 |
+
|
| 83 |
+
**Clean evaluation.** Benchmark test data is kept out or listed (see the README). Every published number comes from a
|
| 84 |
+
measurement.
|
| 85 |
+
|
| 86 |
+
## How the data is used in training
|
| 87 |
+
|
| 88 |
+
| stage | data |
|
| 89 |
+
|---|---|
|
| 90 |
+
| Pretraining | all text, pictures, video and audio (without held-out shards and skip-listed items), windows growing from 2k to 32k |
|
| 91 |
+
| End of pretraining (anneal) | the 32k stage: about 40% long documents, 60% anneal maths, code and the best text |
|
| 92 |
+
| Tuning | text sets, then pictures, video, sound and text together, then the Australian conversations |
|
| 93 |
+
|
| 94 |
+
## Sizes and speeds quoted here
|
| 95 |
+
|
| 96 |
+
All measured: CPU speeds on an Intel Xeon W-11955M laptop (8 threads) with our kernels at the real sizes; recall from
|
| 97 |
+
a full-size test run on 0.85B tokens. Quality numbers for the trained model will be published with it.
|
LICENSE.md
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# QLNI Non-Commercial Research Licence
|
| 2 |
+
|
| 3 |
+
Copyright (c) 2026 Sai Kiran Bathula, QLNI (New South Wales, Australia).
|
| 4 |
+
|
| 5 |
+
Permission is granted to use, copy, modify and redistribute the data, tokens, counts and documents in this repository
|
| 6 |
+
("the Materials") for non-commercial research and educational purposes only, provided that this notice and the
|
| 7 |
+
attribution to the author are kept with every copy or derivative.
|
| 8 |
+
|
| 9 |
+
Commercial use of any kind requires a separate written licence from the author (saikiranbathula@qlni.io). This
|
| 10 |
+
includes use in a product or service, use to train or improve a model that is offered commercially, and redistribution
|
| 11 |
+
for a fee.
|
| 12 |
+
|
| 13 |
+
The tokens are derived from the third-party sources listed in README.md, and each part also carries the terms of its
|
| 14 |
+
source in addition to this licence (for example ODC-By 1.0 for FineWeb-Edu, CC BY 4.0 for DCLM-Baseline and the Open
|
| 15 |
+
Australian Legal Corpus, CC BY-SA for Wikipedia text, MIT for the OpenCoder corpora, the NVIDIA Open Data License for
|
| 16 |
+
Nemotron-CC-Math, public domain for Project Gutenberg). Where a source's terms require it, those terms govern that part.
|
| 17 |
+
|
| 18 |
+
The Materials are provided "as is", without warranty of any kind. The author is not liable for any claim or damage
|
| 19 |
+
arising from their use.
|
README.md
ADDED
|
@@ -0,0 +1,259 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
pretty_name: SHADOW Assistant Data
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
license: other
|
| 6 |
+
license_name: qlni-non-commercial-research
|
| 7 |
+
license_link: LICENSE.md
|
| 8 |
+
size_categories:
|
| 9 |
+
- 100M<n<1B
|
| 10 |
+
task_categories:
|
| 11 |
+
- text-generation
|
| 12 |
+
- image-to-text
|
| 13 |
+
- visual-question-answering
|
| 14 |
+
- video-text-to-text
|
| 15 |
+
- automatic-speech-recognition
|
| 16 |
+
- audio-classification
|
| 17 |
+
- question-answering
|
| 18 |
+
tags:
|
| 19 |
+
- arxiv:2610.22428
|
| 20 |
+
- pretraining
|
| 21 |
+
- multimodal
|
| 22 |
+
- instruction-tuning
|
| 23 |
+
- tokenized
|
| 24 |
+
- australia
|
| 25 |
+
- australian-english
|
| 26 |
+
- small-language-model
|
| 27 |
+
- on-device
|
| 28 |
+
- cpu
|
| 29 |
+
- ternary
|
| 30 |
+
- bitnet
|
| 31 |
+
- long-context
|
| 32 |
+
- speech
|
| 33 |
+
- audio-captioning
|
| 34 |
+
- video
|
| 35 |
+
- ocr
|
| 36 |
+
- documents
|
| 37 |
+
- charts
|
| 38 |
+
- math
|
| 39 |
+
- code
|
| 40 |
+
- legal
|
| 41 |
+
- parliament
|
| 42 |
+
- decontaminated
|
| 43 |
+
- synthetic
|
| 44 |
+
- frozen-embedding
|
| 45 |
+
- gemma-tokenizer
|
| 46 |
+
configs:
|
| 47 |
+
- config_name: kimi
|
| 48 |
+
data_files: sft/kimi/*.jsonl.zst
|
| 49 |
+
- config_name: sft_audio
|
| 50 |
+
data_files: sft/audio/*.jsonl.zst
|
| 51 |
+
- config_name: sft_vision
|
| 52 |
+
data_files: sft/vision/*.jsonl.zst
|
| 53 |
+
- config_name: smoltalk
|
| 54 |
+
data_files: sft/smoltalk/*.jsonl.zst
|
| 55 |
+
---
|
| 56 |
+
|
| 57 |
+
# SHADOW Assistant Data
|
| 58 |
+
|
| 59 |
+
All the training data for **SHADOW Assistant**: a small AI that reads, writes, does maths and code, looks at pictures,
|
| 60 |
+
documents and video, listens to speech and sound, and remembers past conversations from disk, running offline on a
|
| 61 |
+
normal laptop CPU. Trained from scratch, with a strong focus on Australia.
|
| 62 |
+
|
| 63 |
+
> **[Roadmap](ROADMAP.md):** every step from data to release, and where we are now.
|
| 64 |
+
|
| 65 |
+
## Why this dataset exists
|
| 66 |
+
|
| 67 |
+
**Ready to train, so you do not pay for it again.** Tokenizing and encoding all of this cost us about $400 of GPU time,
|
| 68 |
+
plus weeks of work finding sources, checking licences, filtering and decontaminating. Everything here is already
|
| 69 |
+
tokenized (Gemma tokenizer) and encoded: load one source or all of them and start training.
|
| 70 |
+
|
| 71 |
+
**More than SHADOW needs.** SHADOW reads only the 512-bit codes, but where the licence allows we also publish the
|
| 72 |
+
encoders' vectors (int8, with a scale per token), so the same data works for other projects: your own model, image or
|
| 73 |
+
audio retrieval, embedding research, or studying how each kind of data shapes a model. Results on how each part of the
|
| 74 |
+
data shapes SHADOW will be published with the trained model.
|
| 75 |
+
|
| 76 |
+
| | |
|
| 77 |
+
|---|---|
|
| 78 |
+
| Text pretraining | **286B tokens** (web, Australian, legal, code, maths, books, arithmetic) |
|
| 79 |
+
| Pictures and video | about **9.5B** picture tokens, with about 9.9B tokens of captions and questions |
|
| 80 |
+
| Audio | about **37,700 hours** (about 3.4B sound tokens), with transcripts and captions |
|
| 81 |
+
| Maths for the end of training | about **10.7B tokens** |
|
| 82 |
+
| Tuning (SFT) | about 7.4M text conversations, 1.74M vision, 1.34M audio, 29k Australian conversations |
|
| 83 |
+
|
| 84 |
+
Browse the conversations in the dataset viewer above (Kimi, audio, vision and SmolTalk tuning sets). The token shards
|
| 85 |
+
are binary and are read with the code in [Loading](#loading).
|
| 86 |
+
|
| 87 |
+
## How SHADOW uses this data
|
| 88 |
+
|
| 89 |
+
```
|
| 90 |
+
text, pictures, video, sound ──> 512-bit codes from one frozen table ──> 135M ternary body ──> next token
|
| 91 |
+
│
|
| 92 |
+
memory: gist (RNN state, 26 KB) window (last 32k tokens) archive on disk + index Engram (4B rows on disk)
|
| 93 |
+
```
|
| 94 |
+
|
| 95 |
+
SHADOW keeps four kinds of memory, and each is taught by part of this data:
|
| 96 |
+
|
| 97 |
+
| memory | what it is | taught by |
|
| 98 |
+
|---|---|---|
|
| 99 |
+
| **Gist** | 21 RNN layers carry a fixed 26 KB summary of everything read, whatever the length; it can be saved to resume a conversation instantly | long documents: books, legal judgments, Hansard, long web pages (30B tokens in documents of 8k+) |
|
| 100 |
+
| **Window** | 3 attention layers see the last 32k tokens exactly | window stages from 2k to 32k on long documents |
|
| 101 |
+
| **Archive and index** | everything older stays on disk; a word-match search plus a trained index fetch the best 2,048 tokens back into the attention layers, best block last | archive practice on long documents: facts placed early, questions asked later (in a test, fetching lifted recall from 9% to 63%) |
|
| 102 |
+
| **Engram** | 4B hashed 2- and 3-token memories read from disk: facts and phrases | all the text (never picture or sound codes) |
|
| 103 |
+
|
| 104 |
+
Pictures, video and sound share the same 512-bit space as words, so their captions and transcripts (about 10B tokens)
|
| 105 |
+
teach the body what things look and sound like. More in the [guide](GUIDE.md).
|
| 106 |
+
|
| 107 |
+
## Use it for
|
| 108 |
+
|
| 109 |
+
- **Training a small model**: the text shards, with their per-shard token counts for a unigram prior.
|
| 110 |
+
- **Tuning only**: the conversation sets in `sft/`, including 29k Australian conversations.
|
| 111 |
+
- **Retrieval and embedding research**: picture, video and sound codes and the int8 encoder vectors, with their text.
|
| 112 |
+
- **Clean evaluation**: the held-out manifests and the lists in `decontamination/`.
|
| 113 |
+
- **The sets we may not post**: scripts in `rebuild/` that rebuild them from the original source.
|
| 114 |
+
|
| 115 |
+
> **Find your way:** the [map below](#map) says where everything is; the [guide](GUIDE.md) explains every design
|
| 116 |
+
> decision and why; AI agents should read [AGENTS.md](AGENTS.md).
|
| 117 |
+
|
| 118 |
+
## Map
|
| 119 |
+
|
| 120 |
+
| I want... | go to |
|
| 121 |
+
|---|---|
|
| 122 |
+
| text for pretraining | `knowledge/`, `code/`, `math/`, `creative/`, `arithmetic/` |
|
| 123 |
+
| Australian text only | `knowledge/fineweb_edu_au/`, `knowledge/dclm_au/`, `knowledge/legal/`, `knowledge/hansard/`, `knowledge/spaceai/`, `knowledge/joeyllm/` |
|
| 124 |
+
| maths for the end of training (anneal) | `math/anneal/<source>/` |
|
| 125 |
+
| pictures and video | `vision/images/<source>/`, `vision/video/<source>/` |
|
| 126 |
+
| speech, sounds, music | `audio/speech/<source>/`, `audio/sounds/<source>/`, `audio/music/<source>/` |
|
| 127 |
+
| tuning conversations | `sft/<set>/`, `sft/vision/`, `sft/audio/`, `sft/kimi/` |
|
| 128 |
+
| the sets we may not post | `rebuild/` (scripts that rebuild them from the original source) |
|
| 129 |
+
| how to read the files | [Formats](#formats) and [Loading](#loading) below |
|
| 130 |
+
| what was kept out of training, and why | [Clean evaluation](#clean-evaluation) below |
|
| 131 |
+
| why it is built this way | [GUIDE.md](GUIDE.md) |
|
| 132 |
+
| what is done and what is next | [ROADMAP.md](ROADMAP.md) |
|
| 133 |
+
|
| 134 |
+
## What is in it
|
| 135 |
+
|
| 136 |
+
| part | tokens | sources |
|
| 137 |
+
|---|---|---|
|
| 138 |
+
| Knowledge, global | 107.6B | FineWeb-Edu 72.6B, DCLM-Baseline 35.0B |
|
| 139 |
+
| Knowledge, Australian | 34.3B | .au pages of FineWeb-Edu 24.3B and DCLM 6.0B, Open Australian Legal Corpus 2.06B, JoeyLLM Australian 0.99B, Australian Wikipedia 0.37B, SpaceAI Australian Corpus 0.37B, House of Representatives Hansard 1998 to 2025 0.23B, Australian Dictionary of Biography 0.02B |
|
| 140 |
+
| Code | 59.6B | OpenCoder FineWeb code 52.0B, OpenCoder annealing 7.5B |
|
| 141 |
+
| Maths | 53.9B | Nemotron-CC-Math 4plus |
|
| 142 |
+
| Creative writing | 3.6B | Cosmopedia stories, Project Gutenberg books, poetry |
|
| 143 |
+
| Arithmetic | 2.0B | generated by us, every answer checked |
|
| 144 |
+
| Maths for the anneal | about 10.7B | MegaMath Q&A, MathCoder2, TuluMath, DolminoSynthMath, GSM8K train |
|
| 145 |
+
| Pictures | about 7.0B picture tokens | FineVision, SynthChartNet, M4-Instruct, Atlas of Living Australia (256,874 photos), CORD-v2 |
|
| 146 |
+
| Video | about 2.5B picture tokens | VISTA-400K, LLaVA-Video-178K, Vript, Video-STaR, FineVideo, ShareGPT4Video, ActivityNet Captions, YouCook2, Australian Roads Dashcam |
|
| 147 |
+
| Audio | about 37,700 hours | Loquacious Set, LibriSpeech, Australian English speech, AudioSet with AudioSetCaps, Clotho, WavCaps (FreeSound, BBC Sound Effects, SoundBible), MusicCaps |
|
| 148 |
+
|
| 149 |
+
Tuning data (`sft/`): OpenCoder SFT, NuminaMath-CoT, Orca-Math, GSM8K train, ToolMind, Toucan, Glaive function calling,
|
| 150 |
+
When2Call, SmolTalk (without conversations that name other models), Australian alignment and legal QA, JSON extraction,
|
| 151 |
+
29k conversations we generated for SHADOW (everyday Australian life, dates and the limits of what it knows, who SHADOW
|
| 152 |
+
is, working with files, creative writing), and vision and audio conversations built from held-out shards.
|
| 153 |
+
|
| 154 |
+
Documents of 32k tokens or more hold about 6.6B tokens (books, legal judgments, Hansard, long web pages), enough to
|
| 155 |
+
train a 32k context window natively.
|
| 156 |
+
|
| 157 |
+
## Formats
|
| 158 |
+
|
| 159 |
+
**Text shards** (`<source>_w<worker>_<number>.bin`): raw tokens, 3 bytes each, little-endian; every document is
|
| 160 |
+
`<bos> text <eos>`. Next to each: a `.json` (token and document counts, source files) and a `.counts.npz` (how often
|
| 161 |
+
each token appears). Tokenizer: Gemma's 262,144 tokens (the EmbeddingGemma 2 tokenizer).
|
| 162 |
+
|
| 163 |
+
**Picture and video shards** (`.npz`): `codes` (one 512-bit code per picture token, 64 bytes), `vec` and `vec_scale`
|
| 164 |
+
(the vision tower's int8 vector per token, where the licence allows), `img` (per picture: first token, token count,
|
| 165 |
+
budget, width, height). A `.jsonl` gives each sample's text (captions, questions, answers), its pictures, and frame
|
| 166 |
+
times for video. Photos use up to 280 tokens, documents up to 1,120, video one frame a second with up to 64 frames.
|
| 167 |
+
|
| 168 |
+
**Audio shards** (`.npz`): `codes` (25 tokens a second, 64 bytes each) and `clips` (per clip: first token, token count,
|
| 169 |
+
length in hundredths of a second). A `.jsonl` gives each clip's id, source, kind (speech, sound, music) and text
|
| 170 |
+
(transcript, caption, or questions and answers). Large sources are split into `part_NN/` folders.
|
| 171 |
+
|
| 172 |
+
**Tuning** (`sft/<set>/part_NNN.jsonl.zst`): one conversation per line, `{"messages": [...], "tools": [...], "source": ...}`.
|
| 173 |
+
Vision and audio conversations point at their shard and picture or clip instead of copying codes.
|
| 174 |
+
|
| 175 |
+
All codes use one rule for words, pictures and sound: the sign of a fixed projection of EmbeddingGemma 2's vector
|
| 176 |
+
(see [GUIDE.md](GUIDE.md)).
|
| 177 |
+
|
| 178 |
+
## Loading
|
| 179 |
+
|
| 180 |
+
```python
|
| 181 |
+
import json, numpy as np
|
| 182 |
+
from huggingface_hub import hf_hub_download
|
| 183 |
+
R = "QLNI/shadow-assistant-data"
|
| 184 |
+
|
| 185 |
+
# a text shard -> token ids, split into documents
|
| 186 |
+
p = hf_hub_download(R, "knowledge/legal/legal_w00_00000.bin", repo_type="dataset")
|
| 187 |
+
b = np.fromfile(p, np.uint8).reshape(-1, 3).astype(np.int64)
|
| 188 |
+
ids = b[:, 0] | (b[:, 1] << 8) | (b[:, 2] << 16)
|
| 189 |
+
starts = np.flatnonzero(ids == 2) # <bos>
|
| 190 |
+
|
| 191 |
+
# an audio shard -> the codes of each clip, with its text
|
| 192 |
+
z = np.load(hf_hub_download(R, "audio/sounds/clotho/<shard>.npz", repo_type="dataset"))
|
| 193 |
+
lines = open(hf_hub_download(R, "audio/sounds/clotho/<shard>.jsonl", repo_type="dataset")).read().splitlines()
|
| 194 |
+
first, count, centis = z["clips"][0]
|
| 195 |
+
codes = np.unpackbits(z["codes"][first:first + count], axis=1) # [count, 512] bits
|
| 196 |
+
text = json.loads(lines[0])["texts"]
|
| 197 |
+
```
|
| 198 |
+
|
| 199 |
+
## Clean evaluation
|
| 200 |
+
|
| 201 |
+
- **Held-out tuning shards:** 10% of every picture, video and audio source is kept for tuning only. The lists are in
|
| 202 |
+
`sft/vision/vision_sft_manifest.json` and `sft/audio/audio_sft_manifest.json`; pretraining should not read them.
|
| 203 |
+
- **Benchmark test data kept out at build time:** LibriSpeech test, Clotho evaluation, AudioCaps test and validation,
|
| 204 |
+
MusicCaps evaluation, and WavCaps' AudioCaps/Clotho blacklist.
|
| 205 |
+
- **Skip lists for overlaps found after building:** HellaSwag's validation and test items come from 3,299 ActivityNet
|
| 206 |
+
videos that are in our video set; ESC-50, UrbanSound8K and VGGSound share 2,219 FreeSound clips and 15,510 AudioSet
|
| 207 |
+
clips with our audio. Both lists are in `decontamination/`. A 13-gram scan of all text against nine text benchmarks (HumanEval,
|
| 208 |
+
MBPP, GSM8K, MMLU, ARC, HellaSwag, PIQA, WinoGrande, OpenBookQA) will be added there before training.
|
| 209 |
+
- **Disclosed, not removed:** FineVision's authors report that 1.02% of its images match test images of 66 vision
|
| 210 |
+
benchmarks; we use FineVision as released.
|
| 211 |
+
|
| 212 |
+
## Published here or rebuilt
|
| 213 |
+
|
| 214 |
+
Published here, as tokens and codes (never raw pictures or audio): everything above except the sets below.
|
| 215 |
+
|
| 216 |
+
Rebuilt from the original source with the scripts in `rebuild/`, because their terms do not allow reposting:
|
| 217 |
+
Nemotron-CC-Math, Australian Wikipedia (CC BY-SA), Australian Dictionary of Biography (CC BY-NC-ND), and the video sets
|
| 218 |
+
for research use only or without a stated licence (LLaVA-Video-178K, Vript, Video-STaR, ActivityNet Captions,
|
| 219 |
+
YouCook2). MusicCaps codes are kept private; its rebuild script will follow.
|
| 220 |
+
|
| 221 |
+
## Left out on purpose
|
| 222 |
+
|
| 223 |
+
- News sites and other sites whose terms forbid bulk downloading.
|
| 224 |
+
- Sets that only ship file ids, and tuning sets written by models whose licences follow the trained model.
|
| 225 |
+
- MovieChat-1K (no licence) and MAmmoTH-VL (mostly repeats FineVision).
|
| 226 |
+
- Trove digitised newspapers: the National Library of Australia declined our request for an exemption to its API terms.
|
| 227 |
+
- Senate Hansard, House of Representatives Hansard before 1998, AustLII, and several picture collections: not used.
|
| 228 |
+
|
| 229 |
+
## Licence
|
| 230 |
+
|
| 231 |
+
QLNI Non-Commercial Research Licence (see LICENSE.md): free for non-commercial research and education, with
|
| 232 |
+
attribution. Each part also keeps the terms of its source.
|
| 233 |
+
|
| 234 |
+
## Thanks
|
| 235 |
+
|
| 236 |
+
Compute from [CloudRift](https://cloudrift.ai) through their AI Builder Grant, for the picture, video and audio encodes
|
| 237 |
+
and the test runs. The House of Representatives Hansard comes from Lindsay Katz and Rohan Alexander's database of the
|
| 238 |
+
Australian Parliamentary Debates ([Zenodo 17351233](https://zenodo.org/records/17351233), CC BY 4.0). Every source
|
| 239 |
+
above belongs to its creators; see each source's own page.
|
| 240 |
+
|
| 241 |
+
## Citation
|
| 242 |
+
|
| 243 |
+
The frozen vocabulary this data is built for: [Embeddings Are Overpriced?](https://arxiv.org/abs/2610.22428).
|
| 244 |
+
|
| 245 |
+
```bibtex
|
| 246 |
+
@misc{qlni2026shadowdata,
|
| 247 |
+
title = {SHADOW Assistant Data},
|
| 248 |
+
author = {Bathula, Sai Kiran},
|
| 249 |
+
year = {2026},
|
| 250 |
+
publisher = {QLNI},
|
| 251 |
+
url = {https://huggingface.co/datasets/QLNI/shadow-assistant-data}
|
| 252 |
+
}
|
| 253 |
+
```
|
| 254 |
+
When you use this data, please mention QLNI and CloudRift, for example: "Training data from SHADOW Assistant by QLNI
|
| 255 |
+
(https://huggingface.co/datasets/QLNI/shadow-assistant-data), encoded with compute from CloudRift (https://cloudrift.ai)."
|
| 256 |
+
|
| 257 |
+
## Contact
|
| 258 |
+
|
| 259 |
+
Sai Kiran Bathula, QLNI, New South Wales, Australia. saikiranbathula@qlni.io
|
ROADMAP.md
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# SHADOW Assistant roadmap
|
| 2 |
+
|
| 3 |
+
Every step from data to release. Updated 2026-10-31.
|
| 4 |
+
|
| 5 |
+
**Overall: 16 of 46 steps done** `██████████░░░░░░░░░░░░░░░░░░░░` 35%
|
| 6 |
+
|
| 7 |
+
| phase | progress | done |
|
| 8 |
+
|---|---|---|
|
| 9 |
+
| 1. [Data](#1-data) | `████████████████████` | 9 / 9 |
|
| 10 |
+
| 2. [Proof and speed](#2-proof-and-speed) | `█████████████████░░░` | 5 / 6 |
|
| 11 |
+
| 3. [Main pretraining](#3-main-pretraining) | `░░░░░░░░░░░░░░░░░░░░` | 0 / 5 |
|
| 12 |
+
| 4. [Base benchmarks](#4-base-benchmarks) | `░░░░░░░░░░░░░░░░░░░░` | 0 / 6 |
|
| 13 |
+
| 5. [Tuning (SFT)](#5-tuning-sft) | `░░░░░░░░░░░░░░░░░░░░` | 0 / 4 |
|
| 14 |
+
| 6. [Final benchmarks](#6-final-benchmarks) | `░░░░░░░░░░░░░░░░░░░░` | 0 / 2 |
|
| 15 |
+
| 7. [Runtimes and ports](#7-runtimes-and-ports) | `████░░░░░░░░░░░░░░░░` | 2 / 9 |
|
| 16 |
+
| 8. [Release](#8-release) | `░░░░░░░░░░░░░░░░░░░░` | 0 / 5 |
|
| 17 |
+
|
| 18 |
+
## Now
|
| 19 |
+
|
| 20 |
+
- ⏭️ next: **Main-run trainer** (window stages to 32k, anneal mix, skip lists, multi-GPU, streaming, resume)
|
| 21 |
+
- ⏳ waiting: **Compute for the main run** (4x RTX PRO 6000, about 9 days)
|
| 22 |
+
|
| 23 |
+
## 1. Data
|
| 24 |
+
|
| 25 |
+
| status | step | details |
|
| 26 |
+
|---|---|---|
|
| 27 |
+
| ✅ done | Text pretraining data | 286B tokens, 15 sources, Australian 34B |
|
| 28 |
+
| ✅ done | Pictures and video | about 9.5B picture tokens, 256,874 ALA photos |
|
| 29 |
+
| ✅ done | Audio | about 37,700 hours incl. WavCaps |
|
| 30 |
+
| ✅ done | Anneal maths complete | about 10.7B tokens |
|
| 31 |
+
| ✅ done | Tuning data | text 7.4M, vision 1.74M, audio 1.34M, Kimi 28,992 |
|
| 32 |
+
| ✅ done | Final unigram prior | 285.7B tokens, 22,394 shards |
|
| 33 |
+
| ✅ done | Benchmark skip lists | HellaSwag (3,299 videos), ESC-50/UrbanSound8K/VGGSound clips |
|
| 34 |
+
| ✅ done | Pending sources closed | Trove declined; data final |
|
| 35 |
+
| ✅ done | Public data card | README, GUIDE, AGENTS, decontamination lists |
|
| 36 |
+
|
| 37 |
+
## 2. Proof and speed
|
| 38 |
+
|
| 39 |
+
| status | step | details |
|
| 40 |
+
|---|---|---|
|
| 41 |
+
| ✅ done | Archive proof | int4 KV, word-match search, best block last |
|
| 42 |
+
| ✅ done | Full-size test run | 135M + 4B Engram, all senses, 0.85B tokens, passed |
|
| 43 |
+
| ✅ done | Fetch on the trained model | fact recall 9% to 63% of digits |
|
| 44 |
+
| ✅ done | Training speed kernels | fused GRU, fused readout, compile: 7.2x |
|
| 45 |
+
| ✅ done | CPU speed up to 128k | 551 tok/s empty, 141 at 128k |
|
| 46 |
+
| ⬜ to do | Training speed at 32k | on the main-run GPUs |
|
| 47 |
+
|
| 48 |
+
## 3. Main pretraining
|
| 49 |
+
|
| 50 |
+
| status | step | details |
|
| 51 |
+
|---|---|---|
|
| 52 |
+
| ⏭️ next | Main-run trainer | window stages to 32k, anneal mix, skip lists, multi-GPU, streaming, resume |
|
| 53 |
+
| ⏳ waiting | Compute for the main run | 4x RTX PRO 6000, about 9 days |
|
| 54 |
+
| ⬜ to do | Text decontamination scan | 13-grams vs 9 benchmarks over all 286B tokens |
|
| 55 |
+
| ⬜ to do | Pretraining, 2k to 12k | one pass of the corpus |
|
| 56 |
+
| ⬜ to do | 32k stage and anneal | 40% long documents, 60% maths, code, best text |
|
| 57 |
+
|
| 58 |
+
## 4. Base benchmarks
|
| 59 |
+
|
| 60 |
+
| status | step | details |
|
| 61 |
+
|---|---|---|
|
| 62 |
+
| ⬜ to do | Evaluation harness | runs every benchmark on each checkpoint |
|
| 63 |
+
| ⬜ to do | Quality for the size | HellaSwag, ARC, PIQA, ArithMark, MMLU, GSM8K, code |
|
| 64 |
+
| ⬜ to do | Window: 32k to 128k | needle in a haystack, RULER, YaRN to 128k |
|
| 65 |
+
| ⬜ to do | Memory beyond the window | BABILong, LongMemEval, own archive test |
|
| 66 |
+
| ⬜ to do | Pictures and video | MMMU, DocVQA, ChartQA, Video-MME, COCO |
|
| 67 |
+
| ⬜ to do | Audio | LibriSpeech, AudioCaps, Clotho, MusicCaps, MMAU |
|
| 68 |
+
|
| 69 |
+
## 5. Tuning (SFT)
|
| 70 |
+
|
| 71 |
+
| status | step | details |
|
| 72 |
+
|---|---|---|
|
| 73 |
+
| ⬜ to do | SFT trainer | chat template, loss on answers, Engram frozen |
|
| 74 |
+
| ⬜ to do | Text SFT | 7.1M conversations |
|
| 75 |
+
| ⬜ to do | Media SFT | pictures, video, sound and text together (SmolVLM2 mix) |
|
| 76 |
+
| ⬜ to do | Kimi SFT | Australian life, identity, files, time, creative |
|
| 77 |
+
|
| 78 |
+
## 6. Final benchmarks
|
| 79 |
+
|
| 80 |
+
| status | step | details |
|
| 81 |
+
|---|---|---|
|
| 82 |
+
| ⬜ to do | All benchmarks again | on the tuned model, plus IFEval |
|
| 83 |
+
| ⬜ to do | Real-use demos | chat, photos, documents, sound, memory |
|
| 84 |
+
|
| 85 |
+
## 7. Runtimes and ports
|
| 86 |
+
|
| 87 |
+
| status | step | details |
|
| 88 |
+
|---|---|---|
|
| 89 |
+
| ✅ done | CPU kernels at real size | ternary, int4 attention, RNN, Engram, eye |
|
| 90 |
+
| ✅ done | Check which runtimes accept new models | llama.cpp takes new models (fork first, then pull requests); ONNX via a custom-op package; OpenVINO via optimum-intel |
|
| 91 |
+
| ⬜ to do | C++ runtime with real weights | every feature, the speed headline |
|
| 92 |
+
| ⬜ to do | Hugging Face model code | safetensors + modelling code; unblocks every converter |
|
| 93 |
+
| ⬜ to do | llama.cpp fork, then pull requests | CPU text body first (TQ2_0, q4_0 KV), then GRU op, Engram, eye and ear; also reaches Ollama |
|
| 94 |
+
| ⬜ to do | MLX (Apple) | single-file model, BitNet kernel exists |
|
| 95 |
+
| ⬜ to do | ONNX custom-op package | stateful RNN, host code for Engram and archive |
|
| 96 |
+
| ⬜ to do | OpenVINO | issue first with optimum-intel; Intel CPUs and GPUs |
|
| 97 |
+
| ⬜ to do | Optional GPU path for eye and ear | about 20x faster sound encoding |
|
| 98 |
+
|
| 99 |
+
## 8. Release
|
| 100 |
+
|
| 101 |
+
| status | step | details |
|
| 102 |
+
|---|---|---|
|
| 103 |
+
| ⬜ to do | Model on Hugging Face | weights, model card, CloudRift credit |
|
| 104 |
+
| ⬜ to do | Public code repository | runtime and docs, one clean history |
|
| 105 |
+
| ⬜ to do | Open SLM Leaderboard entry | sub-150M class |
|
| 106 |
+
| ⬜ to do | Papers | paper 2: one space for all senses; index and archive |
|
| 107 |
+
| ⬜ to do | Announce | after measured results |
|
| 108 |
+
|
| 109 |
+
Status: ✅ done, 🔄 running, ⏭️ next, ⬜ to do, ⏳ waiting.
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c8c4766bc1f6bdaa43ea1ab963654a8bf1f6285c82a10b2d0c609fec4949e4c
|
| 3 |
+
size 750079734
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9abb1b5a106edfbefa232835ee44b5e6e241647d4bb4b0a21dbed0f05495956
|
| 3 |
+
size 11094
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250026578, "documents": 2223104, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1317bf70f5bea20cac7ccb1e4be092b33e1705b6c074694b40d24e4004335874
|
| 3 |
+
size 249934941
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31c901ba5213f32ebc64c70e2f4eb08a10ca276beb2126a119b55764adca204e
|
| 3 |
+
size 10629
|
arithmetic/arithmetic_generated/arithmetic_generated_w00_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83311647, "documents": 740096, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:debf1e9d11d024c48b243b6a0f83a980f674397de3683efc120d74692385a4ca
|
| 3 |
+
size 750052770
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d2e7bf74911dd21a3b5fdc4a976e6530b52f78f9a1a05fdfece3243a1a9f7b14
|
| 3 |
+
size 11095
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250017590, "documents": 2222336, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac4be1fb6b5a19248b889b04af2baf0144719e33882c2ed2568c1f7037bbb688
|
| 3 |
+
size 250001934
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:685d64f7518575739bb33027f77fdf3caf08d67f421e0ee2500d4bcd56103368
|
| 3 |
+
size 10645
|
arithmetic/arithmetic_generated/arithmetic_generated_w01_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83333978, "documents": 740608, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a8511b9c585b1d38038d3a1b2f1917df12e2a33b6ee030b19dad2727033a7e0
|
| 3 |
+
size 750069609
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5334e911243501d96d9bee1825d90fa336b9d629d2589ae189c0447ebd581800
|
| 3 |
+
size 11062
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250023203, "documents": 2221568, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c42e78e79f9464f23f6e45d9f4c1f8bc52b0042a23731de3208a9b15ba6ddbc0
|
| 3 |
+
size 250009623
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fee9a82696a7c832c0ff143f17528741d797de6b86b8b5018ee626daca1f0eda
|
| 3 |
+
size 10653
|
arithmetic/arithmetic_generated/arithmetic_generated_w02_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83336541, "documents": 740352, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be96a7d44eb863b880b3ae094fc37b30a15fad2f4f04a1908158cbd5431c34f7
|
| 3 |
+
size 750068973
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94c4a799ce87301e148c720a575d08f95d0d3d2e6ff88af9085a49689583e2ed
|
| 3 |
+
size 11089
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250022991, "documents": 2222848, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea6f40d7c35c5b2cb6e62d7217c5a1d10aa28b3dd2991df373618eb82ef5e63c
|
| 3 |
+
size 249953895
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4ff09b9e1a1de7ef4ea088ee79b5edb3f9e50c8f4fcdc554b2c89ca7fc67b92
|
| 3 |
+
size 10641
|
arithmetic/arithmetic_generated/arithmetic_generated_w03_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83317965, "documents": 741120, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0deef87365db1b74da21a0349bcbf1c4f47fbc5615f5b5b9a29e5021a1eaff0
|
| 3 |
+
size 750012729
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92689da6f2f56ae74c2ebf55af618d2eb5878e138605fbd87da9b3b3749a82c0
|
| 3 |
+
size 11124
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250004243, "documents": 2221824, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:810718252a74067115fe54eabcaf0cecc0f1f12f0771ccd46dcd4d143e425ce1
|
| 3 |
+
size 250066602
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15c4d2c522b9ef29492086e7dba1fff4e2afb2854d10f85ec82d275ae4573713
|
| 3 |
+
size 10617
|
arithmetic/arithmetic_generated/arithmetic_generated_w04_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83355534, "documents": 740864, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59049963df8427efff81aa4b8957baedbf8d976c144c712c08ac9d7e4e06b8cf
|
| 3 |
+
size 750016821
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7c8104b99319ad72adc77e2e22ee8f984e48cb22c336d69b298b5ff86fe6a04
|
| 3 |
+
size 11104
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00000.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 250005607, "documents": 2222336, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:848c2ce90c2348eca70b1fa319cfce97939d0c8cc236cb1a1f5831d5270049a6
|
| 3 |
+
size 249988920
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a10adb5833ab53eff041f4d436b4b278ec5e9e6d3c4120662b4b96dee1be25a
|
| 3 |
+
size 10678
|
arithmetic/arithmetic_generated/arithmetic_generated_w05_00001.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"tokens": 83329640, "documents": 740608, "bytes_per_token": 3, "bos": 2, "eos": 1, "source_files": ["generated"]}
|
audio/sounds/audioset/g1_000_000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea04eb323533718922a31d90cb3bc54419344f16a0c09b312d00555f15061fdd
|
| 3 |
+
size 12568
|
audio/sounds/audioset/g1_000_000.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
audio/sounds/audioset/g1_000_000.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b83de5c3d64fd0174cb329573ade2cb1ccf56aa551e4d23c5095f3f34e93a3f8
|
| 3 |
+
size 286624838
|
audio/sounds/audioset/g1_001_000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:29af8f574394fce2ea81958f6ee2d71032e69ceb24f782b9f91eef6c869558f8
|
| 3 |
+
size 12535
|
audio/sounds/audioset/g1_001_000.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
audio/sounds/audioset/g1_001_000.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b4260a2d83181e9a935869fcecc76654dca6bc183dfc4c8a425656198260690
|
| 3 |
+
size 286123158
|
audio/sounds/audioset/g1_002_000.counts.npz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e29c9c3bf1c20a121fdf275e883a34fae9bc38b9f39c4f26e1c791054518feb7
|
| 3 |
+
size 12687
|
audio/sounds/audioset/g1_002_000.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|