Add files using upload-large-folder tool

.gitattributes

@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+visualizations/training_evolution_batch_1.gif filter=lfs diff=lfs merge=lfs -text
+visualizations/training_evolution_batch_2.gif filter=lfs diff=lfs merge=lfs -text
+visualizations/training_evolution_batch_3.gif filter=lfs diff=lfs merge=lfs -text
+visualizations/training_evolution_batch_4.gif filter=lfs diff=lfs merge=lfs -text
+visualizations/training_evolution_batch_5.gif filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,120 @@
+---
+license: mit
+tags:
+- chemistry
+- smiles
+- tokenization
+- dynamic-tokenization
+- h-net
+- hierarchical-networks
+- molecular-representation
+- polymer
+- mamba
+- transformer
+datasets:
+- PI1M
+language:
+- en
+pipeline_tag: feature-extraction
+---
+
+# PI1M-68M
+
+**H-Net model for dynamic SMILES tokenization**
+
+PI1M polymer dataset, 68M bytes (~1 epoch), 10x concatenation, 1-stage architecture
+
+## Model Details
+
+| Property | Value |
+|----------|-------|
+| **Architecture** | H-Net (Hierarchical Network) |
+| **Parameters** | ~350M |
+| **Dataset** | PI1M |
+| **Training Bytes** | 68M |
+| **Training Epochs** | 1 |
+| **Concatenation** | 10x SMILES per example |
+| **Architecture Variant** | 1-stage |
+
+### Architecture Layout
+
+1-stage: `['m4', ['T22'], 'm4']`
+
+- **Encoder**: 4 Mamba blocks for byte-level encoding
+- **Core**: 22 Transformer blocks with boundary prediction
+- **Decoder**: 4 Mamba blocks for final decoding
+
+## Files
+
+- `checkpoints/checkpoint_bytes_best.pt` - Best checkpoint (lowest validation loss)
+- `checkpoints/checkpoint_epoch_*.pt` - Epoch checkpoints
+- `metadata.json` - Training configuration and history
+- `test_smiles.txt` - Test SMILES used during training
+- `visualizations/` - Training evolution GIFs and prediction files
+
+## Usage
+
+```python
+import torch
+from pathlib import Path
+
+# Load checkpoint
+checkpoint_path = "checkpoints/checkpoint_bytes_best.pt"
+checkpoint = torch.load(checkpoint_path, map_location="cpu")
+
+# The checkpoint contains:
+# - 'model_state_dict': Model weights
+# - 'optimizer_state_dict': Optimizer state
+# - 'epoch': Training epoch
+# - 'metrics': Training metrics
+# - 'cumulative_training_bytes': Total bytes processed
+
+# Load into your H-Net model
+# model.load_state_dict(checkpoint['model_state_dict'])
+```
+
+## Performance
+
+### Tokenization Metrics (from paper)
+
+| Metric | Value |
+|--------|-------|
+| Bits-per-byte (BPB) | 0.83 |
+| Mean token length | 2.6 |
+
+### Property Prediction (embeddings)
+
+H-Net embeddings outperform RDKit descriptors on classification tasks:
+- BBBP: 0.950 AUC (vs 0.927 for RDKit)
+- HIV: 0.788 AUC (vs 0.760 for RDKit)
+
+## Citation
+
+```bibtex
+@inproceedings{hnet_smiles_2026,
+  title={Learning Chemical Grammar: Dynamic Tokenization for SMILES with Hierarchical Networks},
+  author={Anonymous},
+  booktitle={International Conference on Machine Learning (ICML)},
+  year={2026}
+}
+```
+
+## Related Models
+
+All models from the paper are available:
+
+**Polymer (PI1M) Models:**
+- [PI1M-68M](https://huggingface.co/jordiferrero/PI1M-68M) - 1 epoch, with concatenation
+- [PI1M-340M](https://huggingface.co/jordiferrero/PI1M-340M) - 5 epochs, with concatenation  
+- [PI1M-1B](https://huggingface.co/jordiferrero/PI1M-1B) - 22 epochs, with concatenation (best compression)
+- [PI1M-nocat](https://huggingface.co/jordiferrero/PI1M-nocat) - 5 epochs, no concatenation
+- [PI1M-2stg](https://huggingface.co/jordiferrero/PI1M-2stg) - 5 epochs, 2-stage architecture
+
+**Molecular (MOSES) Models:**
+- [MOSES-340M](https://huggingface.co/jordiferrero/MOSES-340M) - 5 epochs, with concatenation
+- [MOSES-nocat](https://huggingface.co/jordiferrero/MOSES-nocat) - 5 epochs, no concatenation
+- [MOSES-2stg](https://huggingface.co/jordiferrero/MOSES-2stg) - 5 epochs, 2-stage architecture
+
+## License
+
+MIT License

metadata.json

@@ -0,0 +1,555 @@
+{
+  "run_name": "run_large_20251113_181705",
+  "timestamp": "20251113_181705",
+  "phase": "large",
+  "config": {
+    "arch_layout": [
+      "m4",
+      [
+        "T22"
+      ],
+      "m4"
+    ],
+    "d_model": [
+      1024,
+      1536
+    ],
+    "d_intermediate": [
+      0,
+      4096
+    ],
+    "vocab_size": 256,
+    "ssm_cfg": {
+      "chunk_size": 256,
+      "d_conv": 4,
+      "d_state": 128,
+      "expand": 2
+    },
+    "attn_cfg": {
+      "num_heads": [
+        16,
+        16
+      ],
+      "rotary_emb_dim": [
+        32,
+        48
+      ],
+      "window_size": [
+        1023,
+        -1
+      ]
+    },
+    "tie_embeddings": false
+  },
+  "training_args": {
+    "data": "datasets/PI1M/PI1M_v2.csv",
+    "max_samples": null,
+    "batch_size": 16,
+    "epochs": 1,
+    "lr": 0.0001,
+    "weight_decay": 0.1,
+    "gradient_accumulation": 8,
+    "concatenate": true,
+    "num_concatenate": 10,
+    "concatenate_separator": " ",
+    "checkpoint_bytes": 1000000,
+    "num_test_samples": 5,
+    "num_visualize": 5,
+    "skip_visualization": false
+  },
+  "dataset_info": {
+    "train_size": 99574,
+    "test_size": 5,
+    "test_smiles_file": "checkpoints/run_large_20251113_181705/test_smiles.txt"
+  },
+  "model_info": {
+    "num_parameters": 622923776,
+    "device": "cuda",
+    "dtype": "torch.bfloat16",
+    "use_amp": true
+  },
+  "training_history": [
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 1000000,
+      "cumulative_training_bytes": 1003954,
+      "metrics": {
+        "loss": 2.6488359265655053,
+        "ce_loss": 2.6388358778625953,
+        "lb_loss": 0.9999999945400325
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 2000000,
+      "cumulative_training_bytes": 2001594,
+      "metrics": {
+        "loss": 1.8856113029800299,
+        "ce_loss": 1.8756112833969465,
+        "lb_loss": 1.0000000022749864
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 3000000,
+      "cumulative_training_bytes": 3001614,
+      "metrics": {
+        "loss": 1.5395205251436501,
+        "ce_loss": 1.5295205152671756,
+        "lb_loss": 1.0000000034883125
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 4000000,
+      "cumulative_training_bytes": 4006760,
+      "metrics": {
+        "loss": 1.334807674393399,
+        "ce_loss": 1.324807669370229,
+        "lb_loss": 1.0000000032987304
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 5000000,
+      "cumulative_training_bytes": 5001224,
+      "metrics": {
+        "loss": 1.196902716196276,
+        "ce_loss": 1.1869027140672783,
+        "lb_loss": 1.0000000030075737
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 6000000,
+      "cumulative_training_bytes": 6005860,
+      "metrics": {
+        "loss": 1.0961166403096192,
+        "ce_loss": 1.0861166401273885,
+        "lb_loss": 1.0000000024297435
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 7000000,
+      "cumulative_training_bytes": 7003503,
+      "metrics": {
+        "loss": 1.0216845957959284,
+        "ce_loss": 1.0116845969945356,
+        "lb_loss": 1.0000000018239672
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 8000000,
+      "cumulative_training_bytes": 8006238,
+      "metrics": {
+        "loss": 0.9628897264856225,
+        "ce_loss": 0.9528897287284895,
+        "lb_loss": 1.0000000014815686
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 9000000,
+      "cumulative_training_bytes": 9000603,
+      "metrics": {
+        "loss": 0.9169575065171638,
+        "ce_loss": 0.9069575095663265,
+        "lb_loss": 1.0000000014698425
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 10000000,
+      "cumulative_training_bytes": 10007033,
+      "metrics": {
+        "loss": 0.8789942862426104,
+        "ce_loss": 0.8689942899464832,
+        "lb_loss": 1.0000000010936632
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 11000000,
+      "cumulative_training_bytes": 11004801,
+      "metrics": {
+        "loss": 0.8476108267841154,
+        "ce_loss": 0.837610831015299,
+        "lb_loss": 1.0000000000828995
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 12000000,
+      "cumulative_training_bytes": 12006473,
+      "metrics": {
+        "loss": 0.8209193093436105,
+        "ce_loss": 0.8109193140146683,
+        "lb_loss": 0.9999999999239737
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 13000000,
+      "cumulative_training_bytes": 13001107,
+      "metrics": {
+        "loss": 0.7981791027302736,
+        "ce_loss": 0.7881791077738516,
+        "lb_loss": 0.9999999992628401
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 14000000,
+      "cumulative_training_bytes": 14006641,
+      "metrics": {
+        "loss": 0.7781868026581145,
+        "ce_loss": 0.7681868080235101,
+        "lb_loss": 0.9999999982728014
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 15000000,
+      "cumulative_training_bytes": 15000604,
+      "metrics": {
+        "loss": 0.7610264323705064,
+        "ce_loss": 0.7510264380107252,
+        "lb_loss": 0.999999998417037
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 16000000,
+      "cumulative_training_bytes": 16005428,
+      "metrics": {
+        "loss": 0.745549300003417,
+        "ce_loss": 0.7355493058879847,
+        "lb_loss": 0.9999999981739123
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 17000000,
+      "cumulative_training_bytes": 17003164,
+      "metrics": {
+        "loss": 0.7321184523850209,
+        "ce_loss": 0.7221184584835512,
+        "lb_loss": 0.9999999979585611
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 18000000,
+      "cumulative_training_bytes": 18007792,
+      "metrics": {
+        "loss": 0.7199148238339053,
+        "ce_loss": 0.7099148301254785,
+        "lb_loss": 0.9999999978196514
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 19000000,
+      "cumulative_training_bytes": 19000101,
+      "metrics": {
+        "loss": 0.7091557649007395,
+        "ce_loss": 0.6991557713623539,
+        "lb_loss": 0.9999999983663378
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 20000000,
+      "cumulative_training_bytes": 20004224,
+      "metrics": {
+        "loss": 0.699172711847022,
+        "ce_loss": 0.6891727184628637,
+        "lb_loss": 0.9999999983341734
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 21000000,
+      "cumulative_training_bytes": 21007087,
+      "metrics": {
+        "loss": 0.6900150315239183,
+        "ce_loss": 0.6800150382792562,
+        "lb_loss": 0.9999999978487569
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 22000000,
+      "cumulative_training_bytes": 22001426,
+      "metrics": {
+        "loss": 0.6818090505065546,
+        "ce_loss": 0.671809057387748,
+        "lb_loss": 0.9999999977593799
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 23000000,
+      "cumulative_training_bytes": 23005677,
+      "metrics": {
+        "loss": 0.6744740536304987,
+        "ce_loss": 0.6644740606274967,
+        "lb_loss": 0.9999999976586724
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 24000000,
+      "cumulative_training_bytes": 24006172,
+      "metrics": {
+        "loss": 0.6676162458873062,
+        "ce_loss": 0.6576162529904306,
+        "lb_loss": 0.9999999978895963
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 25000000,
+      "cumulative_training_bytes": 25003110,
+      "metrics": {
+        "loss": 0.661207509639434,
+        "ce_loss": 0.6512075168401714,
+        "lb_loss": 0.999999997937745
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 26000000,
+      "cumulative_training_bytes": 26006257,
+      "metrics": {
+        "loss": 0.6552165900745004,
+        "ce_loss": 0.6452165973653223,
+        "lb_loss": 0.9999999981225502
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 27000000,
+      "cumulative_training_bytes": 27002550,
+      "metrics": {
+        "loss": 0.6496880242396659,
+        "ce_loss": 0.639688031613269,
+        "lb_loss": 0.9999999981241521
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 28000000,
+      "cumulative_training_bytes": 28002689,
+      "metrics": {
+        "loss": 0.6444570045773614,
+        "ce_loss": 0.6344570120284309,
+        "lb_loss": 0.9999999981424468
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 29000000,
+      "cumulative_training_bytes": 29007301,
+      "metrics": {
+        "loss": 0.6395036750117906,
+        "ce_loss": 0.6295036825349697,
+        "lb_loss": 0.9999999981752075
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 30000000,
+      "cumulative_training_bytes": 30001779,
+      "metrics": {
+        "loss": 0.6350867093658836,
+        "ce_loss": 0.625086716955856,
+        "lb_loss": 0.9999999978859286
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 31000000,
+      "cumulative_training_bytes": 31005052,
+      "metrics": {
+        "loss": 0.6308955363285395,
+        "ce_loss": 0.6208955439814815,
+        "lb_loss": 0.9999999979248754
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 32000000,
+      "cumulative_training_bytes": 32006260,
+      "metrics": {
+        "loss": 0.6268380693031025,
+        "ce_loss": 0.6168380770150682,
+        "lb_loss": 0.9999999979613814
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 33000000,
+      "cumulative_training_bytes": 33000074,
+      "metrics": {
+        "loss": 0.6230709754120722,
+        "ce_loss": 0.6130709831786543,
+        "lb_loss": 0.9999999981192038
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 34000000,
+      "cumulative_training_bytes": 34002192,
+      "metrics": {
+        "loss": 0.6197430997301775,
+        "ce_loss": 0.6097431075489754,
+        "lb_loss": 0.9999999979330972
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 35000000,
+      "cumulative_training_bytes": 35004902,
+      "metrics": {
+        "loss": 0.616444450246589,
+        "ce_loss": 0.6064444581146107,
+        "lb_loss": 0.99999999782284
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 36000000,
+      "cumulative_training_bytes": 36005669,
+      "metrics": {
+        "loss": 0.6132622880933641,
+        "ce_loss": 0.6032622960078673,
+        "lb_loss": 0.9999999978454626
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 37000000,
+      "cumulative_training_bytes": 37003862,
+      "metrics": {
+        "loss": 0.6103136071894452,
+        "ce_loss": 0.6003136151479106,
+        "lb_loss": 0.999999997447629
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 38000000,
+      "cumulative_training_bytes": 38000959,
+      "metrics": {
+        "loss": 0.607473284078163,
+        "ce_loss": 0.5974732920779613,
+        "lb_loss": 0.9999999972262947
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 39000000,
+      "cumulative_training_bytes": 39001989,
+      "metrics": {
+        "loss": 0.6047125624955227,
+        "ce_loss": 0.5947125705348381,
+        "lb_loss": 0.9999999972508162
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 40000000,
+      "cumulative_training_bytes": 40006663,
+      "metrics": {
+        "loss": 0.6021836798536883,
+        "ce_loss": 0.5921836879305397,
+        "lb_loss": 0.9999999969547569
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 41000000,
+      "cumulative_training_bytes": 41006405,
+      "metrics": {
+        "loss": 0.599793053990963,
+        "ce_loss": 0.5897930621032487,
+        "lb_loss": 0.9999999971399564
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 42000000,
+      "cumulative_training_bytes": 42002665,
+      "metrics": {
+        "loss": 0.5974936547002956,
+        "ce_loss": 0.5874936628463361,
+        "lb_loss": 0.9999999970556218
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 43000000,
+      "cumulative_training_bytes": 43005789,
+      "metrics": {
+        "loss": 0.595229192796245,
+        "ce_loss": 0.5852292009747196,
+        "lb_loss": 0.9999999970712334
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 44000000,
+      "cumulative_training_bytes": 44006011,
+      "metrics": {
+        "loss": 0.5930570741453018,
+        "ce_loss": 0.5830570823547321,
+        "lb_loss": 0.9999999973764832
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 45000000,
+      "cumulative_training_bytes": 45001756,
+      "metrics": {
+        "loss": 0.5909799169058214,
+        "ce_loss": 0.5809799251446071,
+        "lb_loss": 0.9999999973128222
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 46000000,
+      "cumulative_training_bytes": 46006542,
+      "metrics": {
+        "loss": 0.5889255487099525,
+        "ce_loss": 0.5789255569770344,
+        "lb_loss": 0.9999999973813236
+      }
+    },
+    {
+      "checkpoint_type": "bytes",
+      "bytes_threshold": 47000000,
+      "cumulative_training_bytes": 47000889,
+      "metrics": {
+        "loss": 0.5870973660748213,
+        "ce_loss": 0.5770973743687897,
+        "lb_loss": 0.9999999973008484
+      }
+    },
+    {
+      "epoch": 1,
+      "checkpoint_type": "epoch",
+      "metrics": {
+        "loss": 0.5858815966657623,
+        "ce_loss": 0.575881604976703,
+        "lb_loss": 0.9999999973281337,
+        "training_bytes": 47653402
+      },
+      "cumulative_training_bytes": 47653402,
+      "training_bytes_this_epoch": 47653402
+    }
+  ]
+}

test_smiles.txt

@@ -0,0 +1,5 @@
+*C(=O)c1ccc(C(=O)Nc2ccc(C(*)=O)cc2N)cc1 *OP(=O)(Oc1ccc(I)cc1)Oc1ccc(C(C)(C)c2ccc(*)cc2)c([N+](=O)[O-])c1 */C=C/C(C=C)CCNC(=O)CCCCCCCCCCC(=O)N* *C(=O)Oc1cc(C)c(*)cc1I *Oc1ccc(Oc2ccc(Nc3ccc(NC(=O)c4ccc(*)cc4)cc3)cc2)cc1 *CC(*)(C)C(=O)OCCCCOc1ccc(-c2ccccc2)c(Cl)c1 */N=N/c1ccccc1C(C)CCC(*)=O *CC(*)C(=O)OCCN(CCC)c1ccc(-c2ccc(-c3ccc(C#N)cc3)cc2)cc1 *CCCCCCCCCCC(C)NC(=O)CCCCCCCCCCCCCCCCC(=O)N* *O[Si](*)(C)CCCCCCCCC(=O)OCCCCCCOc1ccc(/C=C/c2ccc(OCCCC)cc2)cc1
+*Oc1ccc(C(C)(C)c2ccc(OC(=O)COc3ccc([PH](*)(=O)=O)cc3)cc2)cc1 *C(=O)NSCCCCCCCCCCNC(=O)C(*)(C)CC *Oc1ccc(C(C)(CC)c2ccc(Oc3ccc(C(=O)c4ccc(NC(=O)c5cc(C(*)=O)cc(C(F)(F)F)c5)cc4)cc3)cc2)cc1 *C(=O)NCc1ccc(C(C)(C)c2ccc(*)o2)c(C)c1 *CC(*)OC(C#N)C[N+](c1ccccc1)c1ccc([N+](=O)[O-])cc1 *CCOCCOP(=O)(Oc1ccc(C#N)cc1)Oc1ccc(O*)cc1 *Oc1ccc(S(=O)(=O)c2ccc(Oc3ccc(/C=C/c4cc(C)c(*)c(OC)c4)cc3C)cc2)cc1Cl *CC(*)(C)C(=O)NCCCCCOc1ccc(/N=N/c2ccc(F)cc2)cc1 *CCOCCOCCOCCOCCOc1ccc(/N=C/CCCCCCCCCCCC/N=N/c2ccc(OCCCCCCCCCC)cc2)cc1* *CC(*)C(=O)NC(C#N)C(=O)OCCCCCCCCC
+*Oc1ccc(CC(NC(=O)Oc2ccc(OC)cc2)C(*)C)cc1 *C=C(CC)S(=O)(=O)c1ccc(*)cc1 *CC(CCC(*)c1ccccc1-c1ccccc1)COc1ccc2ccccc2c1 *CC(*)(COCC)C(=O)OCCCCCOc1ccc(OC(=O)c2ccc(OCCCCC)cc2)cc1 *CCCCCCCCCCCCCCCCCCNC(=O)CCCCC(=O)NCCC(*)=O *Oc1ccc(Oc2ccc(S(=O)(=O)c3ccc(Oc4ccc(C(=O)C(=O)c5ccc(*)c(Br)c5)c(Cl)c4)nc3)cc2)cc1 *CCCCCCCCCCC(=O)NCCC(=O)NNC(=O)CCCO* *N=P(*)(OCCOCCOC)OCCOCC(O)CO *Nc1ccc(NC(=O)C(C)C(*)=CC(=O)Nc2ccc(C#N)cc2)cc1 *O[Si](*)(C)CCCCCCCCCCCOc1ccc(OC(=O)c2ccc(OCc3ccc(OCCCCCCCCCCCCCC)c(C(=O)OCCCC)c3)cc2)cc1
+*CCCCCCC(=O)Oc1ccc(OC(=O)c2ccc(/C=[SH]/c3ccc(O*)cc3)cc2)cc1 *c1cc(OCC)c(CCCCCCCC)c(*)c1CCCCCC *CCCCCCOC(=O)c1ccc(C(=O)NCCCCNC(=O)CCCO*)cc1 */C=N/c1ccc(Cc2ccc(CSC(=O)c3ccc(NC(=O)CCCCCCCC(=O)Nc4cccc(C(*)=O)c4)cc3)cc2)cc1 *CC(*)(CC(=O)OCCOCCCCC[N+](=O)[O-])NC(C)C */C=C/C=C/CC=C(C=C(C#N)C#N)/C=C/C=C/C=C/C=C/CNC(=O)CCCCCCCCCCCCCCCCCCC* *CC(*)C(=O)c1ccc(C(=O)OCCCCCCCC)cc1 *CCCOC(=O)SCCCCOCCCCCCCCC(=O)O* *CCCCCCCCCCCCCCNC(=O)CCSCCCCCC(=O)S* *O[Si](*)(CCOCCOc1ccc(S(=O)(=O)O)cc1)c1ccc([N+](=O)[O-])cc1
+*CCOC(=O)C(CCCCCCc1ccccc1)C(*)C(=O)O *O[Si](C)(C)COC(=O)NCCCCCCCCCCCCCCNC(=O)OCCCCCCOc1ccc(-c2ccc(OC(=O)c3ccc(C(*)=O)cc3)cc2)cc1 *OC(C[SH]C(=O)OC)C(*)OC(=O)c1ccc(O)cc1 *CCCCCCCC(=O)SCCC* *CC(*)c1cc(CO)c(OC)c(O)c1Br *CC(*)(C)C(=O)OCCOC(=O)N(C)CCN(C)c1ccc(OC)cc1 *Oc1ccc(/N=C/c2ccc([Si](*)(C)OCCOCCOCCOCCOCCOCCOCCOCCOCCOCCOCCOC)cc2)cc1 *CCCCCCNC(=O)CCc1ccc(CCNC(=O)O*)cc1 *N=P(*)(OCCOc1ccc(Cl)cc1)Oc1ccc(COC(=O)c2ccccc2)cc1 *CC/C=C(\C(=*)C#N)C(C)=S

visualizations/predictions/predictions_bytes_1,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5129d55d322a8cb1089f3eb2cf001276c1a44aafad254859c356c28c0d07d5c
+size 6498

visualizations/predictions/predictions_bytes_10,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbc5a5ed89346fb8f9a2d8b2d65bc07400314b0762dc1a5a500c25048d2d2e60
+size 6678

visualizations/predictions/predictions_bytes_11,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b60c16f48bed5cc763aa88be5653c0bae8e110d67b32a7600b81aa721307ae55
+size 6674

visualizations/predictions/predictions_bytes_12,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05ca2def1913c5b1a5466bfdc9497111307b100744cf394439f8517258858518
+size 6716

visualizations/predictions/predictions_bytes_13,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:653cf8b71e142006fb57d7daab1b4cc59f308424ba2ed916da42750ff16f9571
+size 6676

visualizations/predictions/predictions_bytes_14,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42a682b9101c8d4dd1f969f75678505500ea9c3cf37c575833237d4e7d9acc84
+size 6675

visualizations/predictions/predictions_bytes_15,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b25aeb94d4f7873c80746a7fdf17888e5b91d60b0ca50e0ae27cef30c3e0cba0
+size 6703

visualizations/predictions/predictions_bytes_16,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c8c73a9fe286ba7ee7878b84e3c084637b07da33b47a46db4a705944bb76b63
+size 6688

visualizations/predictions/predictions_bytes_17,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbeadb2add37180f19720dc0bf0e0c97ee33499f02d4ab9113dba5839d4e36af
+size 6662

visualizations/predictions/predictions_bytes_18,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85614683e6f3e8835c157eb63cb8987cd5a4d14b3a72a55e049aa65c1ecacb84
+size 6694

visualizations/predictions/predictions_bytes_19,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb410a24687f152383bfabf934e18e5241e24cb0584acb88f05158f8799b1d28
+size 6691

visualizations/predictions/predictions_bytes_2,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e731ebc4e0814a02ec32bd6653e8d9e554da4ccd6b8038053a78b01b469da6e
+size 6440

visualizations/predictions/predictions_bytes_20,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89138c658d3e996f0ea2946f65ee62f047de33e6213b2e6c713cc27309e26e14
+size 6719

visualizations/predictions/predictions_bytes_21,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bc6df407ce5a22d070f2e424eb7b25b59899011f63f78ba0fb05720c6f52c1d
+size 6677

visualizations/predictions/predictions_bytes_22,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d48e9b08638033a7128af0738e5b64dc984c03fe20b4fe434d0eba276ece623
+size 6666

visualizations/predictions/predictions_bytes_23,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05c6746b3a6d80a31542c242d1d2e665f815d1ca03b9fde33fce5a1196cda5c2
+size 6729

visualizations/predictions/predictions_bytes_24,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f4b28a35c29fa294f2b2331b47ffbe886514e3216c7c63aac9209870d555282
+size 6759

visualizations/predictions/predictions_bytes_25,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfebee0852090cb8d750ee0dc63140b983f3d8bea5d3a93f8af5ab42741e690e
+size 6779

visualizations/predictions/predictions_bytes_26,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:538f7f3437bf4af210e9b13f2349981989126e6707f4138bbb0c5dafb4e8131a
+size 6724

visualizations/predictions/predictions_bytes_27,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40b381dd6c95cef5b1be86014330b8e83cb64fb6e3d4f1b7b29aeb43913d64f5
+size 6757

visualizations/predictions/predictions_bytes_28,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a26bd0262070757736a2a78618684958132a510f7c1b4d1aa73e9b213fa98851
+size 6753

visualizations/predictions/predictions_bytes_29,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7234c85a643a7530a1a4af7b56671d023db10c3b94f28d13d1e59750fb76e56
+size 6732

visualizations/predictions/predictions_bytes_3,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2593e4be6eab0d5f21de1c5098c6a9074939dd52f98660294e3f1d17af038ef
+size 6519

visualizations/predictions/predictions_bytes_30,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efb3cf76d40574b92e3112d2c7122f5dd54279e0448ccdc582a60c38265755e5
+size 6717

visualizations/predictions/predictions_bytes_31,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:924cca01c1c2b45d8ccc17472518fdf369184e713d7388b6f6fd8dc130b251df
+size 6699

visualizations/predictions/predictions_bytes_32,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:227659306372c81cca4c22d26430a7bf0bab1e582c176620a8d89c951da10adc
+size 6715

visualizations/predictions/predictions_bytes_33,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fed38c8f19e3a48da759ecc990525665406da114992728b1f1eac0d376dc752
+size 6739

visualizations/predictions/predictions_bytes_34,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3000fdc1b0ff877cd7954e443fc7e3f1a7b4aeaab9741b97b16cac62c0163ba
+size 6758

visualizations/predictions/predictions_bytes_35,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b31386c1aded2769ad6355bc477e49d52f52a399daf93f52ca195ea88703e0b5
+size 6759

visualizations/predictions/predictions_bytes_36,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7bf6be9e49b2909e28b600af4154c196f51314820890757e54920c8aff31f91
+size 6788

visualizations/predictions/predictions_bytes_37,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c30ed0ba30afc1da211ce370efc2738a912a8c0a1faeda281a2443f4f0beed0c
+size 6733

visualizations/predictions/predictions_bytes_38,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b283ee7ad862e3cc7c5be70078bc06918bed097afdcd6bfea24396ec06529317
+size 6745

visualizations/predictions/predictions_bytes_39,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:402ba4aa3c880156dedaae1cde361ecb0479278c3ad1171807b3e1cb1fd79f15
+size 6737

visualizations/predictions/predictions_bytes_4,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb840202f462be195c3e5a4f021b7500d19a8cd3a5335a2b7b2d4c9bd4b0217e
+size 6522

visualizations/predictions/predictions_bytes_40,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a654e8845b39cbe78e3c22813f4f36210f011cd20fc3be3555beefef2d04901
+size 6721

visualizations/predictions/predictions_bytes_41,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2594c7d9206a5902be50cb736cb0d9c468fae0ee604c5e51e6c2c693fb9018ef
+size 6728

visualizations/predictions/predictions_bytes_42,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0476d5372ea65f863023352983dcbfefeb3619c24c55a11043796236df5b38fd
+size 6737

visualizations/predictions/predictions_bytes_5,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e2f47f27931e0b9dcc1ce24de048d1c8c98569faea95163a9679a8d951c2d7b
+size 6569

visualizations/predictions/predictions_bytes_6,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c7436a07b0b2eb5c80c73add03bd6f604fc479c3bced17ecafe63c17d21c1e7
+size 6600

visualizations/predictions/predictions_bytes_7,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed0abcb24ecee41f4fa345b8e8c344a4172f02671ae0e3158de759e8f9d51f8e
+size 6683

visualizations/predictions/predictions_bytes_8,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:841c4bc88aeed1021a69889f3e79ec6da373a9f5ef15a5e08e857f6182a72097
+size 6651

visualizations/predictions/predictions_bytes_9,000,000.pkl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df36ac3a5bad398f9ac1d0c0106a00ac3114f97c9742aeabc915c44cd5d70bab
+size 6695