Add 40 model file(s)
Browse files- t1_pure_molecule_xxs1_seed1/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed1/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed1/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed1/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-10Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-12Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs1_seed2/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs1_seed2/config.json +111 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-10Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-12Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed1/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs2_seed1/config.json +111 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-10Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-12Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs2_seed2/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs2_seed2/config.json +111 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-10Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-12Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed1/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs3_seed1/config.json +111 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-10Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-12Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-1Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-2Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-5Blb.pth +3 -0
- t1_pure_molecule_xxs3_seed2/checkpoint-final.pth +3 -0
- t1_pure_molecule_xxs3_seed2/config.json +111 -0
- t1_pure_satellite_s_seed1/config.json +109 -0
t1_pure_molecule_xxs1_seed1/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06b1d8d90c57ddc64e7207c5691d2a1be16924befc84b4cc1ce3c51024644b94
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed1/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c44688621f38e68600a8ed0e5085d469e4da95c14ca7678111d332620a828b78
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed1/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fa328f780bd666d23b3d1540720f1fd25ed9c1a6344d6fb263a5c3143799123
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed1/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db5d1d151fec467bb8900f9e24a3080c1e39f7887db0db2d485879db4afc422e
|
| 3 |
+
size 428922544
|
t1_pure_molecule_xxs1_seed2/checkpoint-10Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee105d9f05da545454a4948f42b9c531981e47d2fc845fce00e71bc3917970dd
|
| 3 |
+
size 428922544
|
t1_pure_molecule_xxs1_seed2/checkpoint-12Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b586c77e9e4dd5cb113d51d4bc43a92338f89af1b36d4c9cd058f7eec1eda92
|
| 3 |
+
size 428922544
|
t1_pure_molecule_xxs1_seed2/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d0a6ac1dc0ef970215eb6db4b01f9129ade8a159af104fa368aca87fceca3b5
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed2/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab6fd6263365d4e3f602009e163f97ed6a3d8af2005b5f25be570a6b63221407
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed2/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f71d7f3c2ee8d265d08ee9024cb19099931b5f0235878019418dafc08c490eb7
|
| 3 |
+
size 428922021
|
t1_pure_molecule_xxs1_seed2/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c604aafdfca6b903458ef798021a11c9d818b2da4c8da056a518a550a06b5e5
|
| 3 |
+
size 428922544
|
t1_pure_molecule_xxs1_seed2/config.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_molecule_xxs1_seed2",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 36881728,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_molecule_xxs1_seed2",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 12,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"effective_batch_size": null,
|
| 13 |
+
"lr_batch_scaling": "sqrt",
|
| 14 |
+
"save_ckpt_freq": 1,
|
| 15 |
+
"save_ckpt_freq_tokens": 5,
|
| 16 |
+
"debug_batch_stats": 0,
|
| 17 |
+
"model": "fm_xxs1_12d_swiglu_nobias",
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"loss_type": "token",
|
| 20 |
+
"context_length": 2048,
|
| 21 |
+
"pos_emb_ablation": "vanilla",
|
| 22 |
+
"max_blocks": 32,
|
| 23 |
+
"intra_pos_emb": "default",
|
| 24 |
+
"video_pos_3d": false,
|
| 25 |
+
"finetune": "",
|
| 26 |
+
"opt": "adamw",
|
| 27 |
+
"opt_eps": 1e-08,
|
| 28 |
+
"opt_betas": [
|
| 29 |
+
0.9,
|
| 30 |
+
0.95
|
| 31 |
+
],
|
| 32 |
+
"compute_grad_norm": true,
|
| 33 |
+
"clip_grad": 1.0,
|
| 34 |
+
"skip_grad": null,
|
| 35 |
+
"momentum": 0.9,
|
| 36 |
+
"weight_decay": 0.1,
|
| 37 |
+
"weight_decay_end": 0.1,
|
| 38 |
+
"blr": 0.001,
|
| 39 |
+
"min_blr": 5e-05,
|
| 40 |
+
"scheduler": "wsd",
|
| 41 |
+
"decay_fraction": 0.1,
|
| 42 |
+
"warmup_steps": 250,
|
| 43 |
+
"cooldown_steps": -1,
|
| 44 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml",
|
| 45 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml",
|
| 46 |
+
"cross_eval_val": "",
|
| 47 |
+
"s3_endpoint": "",
|
| 48 |
+
"s3_data_endpoint": "",
|
| 49 |
+
"s3_multipart_chunksize_mb": 512,
|
| 50 |
+
"s3_multipart_threshold_mb": 512,
|
| 51 |
+
"s3_max_io_queue": 100,
|
| 52 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 53 |
+
"eval_freq": 1,
|
| 54 |
+
"eval_freq_tokens": 2,
|
| 55 |
+
"dist_eval": true,
|
| 56 |
+
"eval": false,
|
| 57 |
+
"val_sampling_mode": "without_replacement",
|
| 58 |
+
"output_dir": "output/seed/t1_pure_molecule_xxs1_seed2",
|
| 59 |
+
"device": "cuda",
|
| 60 |
+
"seed": 2,
|
| 61 |
+
"resume": "",
|
| 62 |
+
"auto_resume": true,
|
| 63 |
+
"start_epoch": 0,
|
| 64 |
+
"num_workers": 10,
|
| 65 |
+
"pin_mem": true,
|
| 66 |
+
"find_unused_params": true,
|
| 67 |
+
"rlimit": 4096,
|
| 68 |
+
"print_all": false,
|
| 69 |
+
"show_user_warnings": false,
|
| 70 |
+
"s3_save_dir": "",
|
| 71 |
+
"dist_url": "env://",
|
| 72 |
+
"log_wandb": true,
|
| 73 |
+
"wandb_project": "loom",
|
| 74 |
+
"wandb_entity": null,
|
| 75 |
+
"wandb_run_name": "t1_pure_molecule_xxs1_seed2",
|
| 76 |
+
"wandb_group": "t1-seedvar-xxs1",
|
| 77 |
+
"dim": null,
|
| 78 |
+
"depth": null,
|
| 79 |
+
"num_heads": null,
|
| 80 |
+
"mlp_ratio": null,
|
| 81 |
+
"log_grad_norms_every": 0,
|
| 82 |
+
"save_at_tokens": "1,2,5,10,12",
|
| 83 |
+
"unigram_entropies": null,
|
| 84 |
+
"cooldown_branch": false,
|
| 85 |
+
"cooldown_tokens": 0,
|
| 86 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs1_seed2.yaml",
|
| 87 |
+
"rank": 0,
|
| 88 |
+
"world_size": 1,
|
| 89 |
+
"gpu": 0,
|
| 90 |
+
"distributed": true,
|
| 91 |
+
"dist_backend": "nccl",
|
| 92 |
+
"num_tasks": 1,
|
| 93 |
+
"sampling_mode": "without_replacement",
|
| 94 |
+
"max_micro_steps": 366211,
|
| 95 |
+
"lossbearing_target": 12000000000.0,
|
| 96 |
+
"lossbearing_mode": true,
|
| 97 |
+
"lr": 0.001,
|
| 98 |
+
"min_lr": 5e-05,
|
| 99 |
+
"global_micro_step": 366211,
|
| 100 |
+
"cum_lossbearing": 12000775806.0,
|
| 101 |
+
"cum_mod_tokens": {
|
| 102 |
+
"text": 881776720,
|
| 103 |
+
"tok_molecule": 11118999086
|
| 104 |
+
}
|
| 105 |
+
},
|
| 106 |
+
"source_configs": {
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs1_seed2.yaml": "# xxs1 \u03c3_seed variant of pure molecule (P1-ladder companion \u2014 the xxs1-scale training-seed\n# floor for the P1 sub-XS scaling anchors). Clone of run_xxs1_pure.yaml + seed 2;\n# seed0 = the existing output/p1/t1_pure_molecule_xxs1 anchor. Train-only self-cool \u2192 output/seed/.\n_base_: ../../../../models/dec/4m-xxs1-dec.yaml\nrun_name: t1_pure_molecule_xxs1_seed2\nseed: 2\ntotal_tokens: -1\nlossbearing_tokens: 12\ndata_config: cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,12\nwandb_group: t1-seedvar-xxs1\n",
|
| 108 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: molecule_zinc20m\n shards: data/shards/molecule_zinc20m_shuffled/train/shard_{000000..021643}.tar\n weight: 21643268\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc\n shards: data/shards/molecule_pubchemqc_reshuffled/train/shard_*.tar\n weight: 213751114\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20\n shards: data/shards/molecule_chebi20_shuffled/train/shard_{000000..000021}.tar\n weight: 21378\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap\n shards: data/shards/molecule_molcap_shuffled/train/shard_{000000..000070}.tar\n weight: 70805\n modality_tokens:\n tok_molecule: 29\n format: npz\nsampling_mode: without_replacement\n",
|
| 109 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: molecule_zinc20m_val\n shards: data/shards/molecule_zinc20m_shuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc_val\n shards: data/shards/molecule_pubchemqc_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20_val\n shards: data/shards/molecule_chebi20_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap_val\n shards: data/shards/molecule_molcap_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 29\n format: npz\n"
|
| 110 |
+
}
|
| 111 |
+
}
|
t1_pure_molecule_xxs2_seed1/checkpoint-10Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb410104583125357a42c291f7e2a5bd1c0817ca2e59406510195fe08ec44fb7
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed1/checkpoint-12Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43442cd54a487d4c91a471bc7d08ad5d057fa61255232cafa4902e325148cf09
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed1/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbe8b9b9a942d7114ac98c3c6d9dfac18c32d0e1e4cf9b1858683a8e7a6e05ee
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed1/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ccc8401bdf060cb890005fe5309398adc2cd00a1e4c1b39bf005aaa2600a255
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed1/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b78dff43af23c3e1e7a675dd8f5d8786942dc807b7314943ba20a869c4f51cce
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed1/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d8dbac3b41874289369c02545eab990e26b34c588068d15c8ddab53b4351433
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed1/config.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_molecule_xxs2_seed1",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 49021568,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_molecule_xxs2_seed1",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 12,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"effective_batch_size": null,
|
| 13 |
+
"lr_batch_scaling": "sqrt",
|
| 14 |
+
"save_ckpt_freq": 1,
|
| 15 |
+
"save_ckpt_freq_tokens": 5,
|
| 16 |
+
"debug_batch_stats": 0,
|
| 17 |
+
"model": "fm_xxs2_12d_swiglu_nobias",
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"loss_type": "token",
|
| 20 |
+
"context_length": 2048,
|
| 21 |
+
"pos_emb_ablation": "vanilla",
|
| 22 |
+
"max_blocks": 32,
|
| 23 |
+
"intra_pos_emb": "default",
|
| 24 |
+
"video_pos_3d": false,
|
| 25 |
+
"finetune": "",
|
| 26 |
+
"opt": "adamw",
|
| 27 |
+
"opt_eps": 1e-08,
|
| 28 |
+
"opt_betas": [
|
| 29 |
+
0.9,
|
| 30 |
+
0.95
|
| 31 |
+
],
|
| 32 |
+
"compute_grad_norm": true,
|
| 33 |
+
"clip_grad": 1.0,
|
| 34 |
+
"skip_grad": null,
|
| 35 |
+
"momentum": 0.9,
|
| 36 |
+
"weight_decay": 0.1,
|
| 37 |
+
"weight_decay_end": 0.1,
|
| 38 |
+
"blr": 0.001,
|
| 39 |
+
"min_blr": 5e-05,
|
| 40 |
+
"scheduler": "wsd",
|
| 41 |
+
"decay_fraction": 0.1,
|
| 42 |
+
"warmup_steps": 250,
|
| 43 |
+
"cooldown_steps": -1,
|
| 44 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml",
|
| 45 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml",
|
| 46 |
+
"cross_eval_val": "",
|
| 47 |
+
"s3_endpoint": "",
|
| 48 |
+
"s3_data_endpoint": "",
|
| 49 |
+
"s3_multipart_chunksize_mb": 512,
|
| 50 |
+
"s3_multipart_threshold_mb": 512,
|
| 51 |
+
"s3_max_io_queue": 100,
|
| 52 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 53 |
+
"eval_freq": 1,
|
| 54 |
+
"eval_freq_tokens": 2,
|
| 55 |
+
"dist_eval": true,
|
| 56 |
+
"eval": false,
|
| 57 |
+
"val_sampling_mode": "without_replacement",
|
| 58 |
+
"output_dir": "output/seed/t1_pure_molecule_xxs2_seed1",
|
| 59 |
+
"device": "cuda",
|
| 60 |
+
"seed": 1,
|
| 61 |
+
"resume": "",
|
| 62 |
+
"auto_resume": true,
|
| 63 |
+
"start_epoch": 0,
|
| 64 |
+
"num_workers": 10,
|
| 65 |
+
"pin_mem": true,
|
| 66 |
+
"find_unused_params": true,
|
| 67 |
+
"rlimit": 4096,
|
| 68 |
+
"print_all": false,
|
| 69 |
+
"show_user_warnings": false,
|
| 70 |
+
"s3_save_dir": "",
|
| 71 |
+
"dist_url": "env://",
|
| 72 |
+
"log_wandb": true,
|
| 73 |
+
"wandb_project": "loom",
|
| 74 |
+
"wandb_entity": null,
|
| 75 |
+
"wandb_run_name": "t1_pure_molecule_xxs2_seed1",
|
| 76 |
+
"wandb_group": "t1-seedvar-xxs2",
|
| 77 |
+
"dim": null,
|
| 78 |
+
"depth": null,
|
| 79 |
+
"num_heads": null,
|
| 80 |
+
"mlp_ratio": null,
|
| 81 |
+
"log_grad_norms_every": 0,
|
| 82 |
+
"save_at_tokens": "1,2,5,10,12",
|
| 83 |
+
"unigram_entropies": null,
|
| 84 |
+
"cooldown_branch": false,
|
| 85 |
+
"cooldown_tokens": 0,
|
| 86 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs2_seed1.yaml",
|
| 87 |
+
"rank": 0,
|
| 88 |
+
"world_size": 1,
|
| 89 |
+
"gpu": 0,
|
| 90 |
+
"distributed": true,
|
| 91 |
+
"dist_backend": "nccl",
|
| 92 |
+
"num_tasks": 1,
|
| 93 |
+
"sampling_mode": "without_replacement",
|
| 94 |
+
"max_micro_steps": 366211,
|
| 95 |
+
"lossbearing_target": 12000000000.0,
|
| 96 |
+
"lossbearing_mode": true,
|
| 97 |
+
"lr": 0.001,
|
| 98 |
+
"min_lr": 5e-05,
|
| 99 |
+
"global_micro_step": 366211,
|
| 100 |
+
"cum_lossbearing": 12000767199.0,
|
| 101 |
+
"cum_mod_tokens": {
|
| 102 |
+
"text": 881769414,
|
| 103 |
+
"tok_molecule": 11118997785
|
| 104 |
+
}
|
| 105 |
+
},
|
| 106 |
+
"source_configs": {
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs2_seed1.yaml": "# xxs2 \u03c3_seed variant of pure molecule (P1-ladder companion \u2014 the xxs2-scale training-seed\n# floor for the P1 sub-XS scaling anchors). Clone of run_xxs2_pure.yaml + seed 1;\n# seed0 = the existing output/p1/t1_pure_molecule_xxs2 anchor. Train-only self-cool \u2192 output/seed/.\n_base_: ../../../../models/dec/4m-xxs2-dec.yaml\nrun_name: t1_pure_molecule_xxs2_seed1\nseed: 1\ntotal_tokens: -1\nlossbearing_tokens: 12\ndata_config: cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,12\nwandb_group: t1-seedvar-xxs2\n",
|
| 108 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: molecule_zinc20m\n shards: data/shards/molecule_zinc20m_shuffled/train/shard_{000000..021643}.tar\n weight: 21643268\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc\n shards: data/shards/molecule_pubchemqc_reshuffled/train/shard_*.tar\n weight: 213751114\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20\n shards: data/shards/molecule_chebi20_shuffled/train/shard_{000000..000021}.tar\n weight: 21378\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap\n shards: data/shards/molecule_molcap_shuffled/train/shard_{000000..000070}.tar\n weight: 70805\n modality_tokens:\n tok_molecule: 29\n format: npz\nsampling_mode: without_replacement\n",
|
| 109 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: molecule_zinc20m_val\n shards: data/shards/molecule_zinc20m_shuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc_val\n shards: data/shards/molecule_pubchemqc_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20_val\n shards: data/shards/molecule_chebi20_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap_val\n shards: data/shards/molecule_molcap_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 29\n format: npz\n"
|
| 110 |
+
}
|
| 111 |
+
}
|
t1_pure_molecule_xxs2_seed2/checkpoint-10Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e87600e7d454423e9b595177c7879366d9530a2cb380f8f7423d02653134e31
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed2/checkpoint-12Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d29291bf16920fb77e8aebebf8ce1917ec7f8e79426851616e00a256a0379b47
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed2/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b34b23b04aff88219c7d884ea8188d96099b0c38453ccd01e6bc7d5ce7b02dcb
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed2/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2dac5abfac54ed0227c9f9066e7c0bdd8161ea04f721e570c92a31e69990bff3
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed2/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a45f4d63c82a551d3b879778f2726d7433813a6fad3dba3c31c33c10cab15cc6
|
| 3 |
+
size 571407781
|
t1_pure_molecule_xxs2_seed2/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f38a26a2e6abf69266f92597ee863e3bbbd026f00d15abf33ea6dbe78234f34
|
| 3 |
+
size 571408304
|
t1_pure_molecule_xxs2_seed2/config.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_molecule_xxs2_seed2",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 49021568,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_molecule_xxs2_seed2",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 12,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"effective_batch_size": null,
|
| 13 |
+
"lr_batch_scaling": "sqrt",
|
| 14 |
+
"save_ckpt_freq": 1,
|
| 15 |
+
"save_ckpt_freq_tokens": 5,
|
| 16 |
+
"debug_batch_stats": 0,
|
| 17 |
+
"model": "fm_xxs2_12d_swiglu_nobias",
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"loss_type": "token",
|
| 20 |
+
"context_length": 2048,
|
| 21 |
+
"pos_emb_ablation": "vanilla",
|
| 22 |
+
"max_blocks": 32,
|
| 23 |
+
"intra_pos_emb": "default",
|
| 24 |
+
"video_pos_3d": false,
|
| 25 |
+
"finetune": "",
|
| 26 |
+
"opt": "adamw",
|
| 27 |
+
"opt_eps": 1e-08,
|
| 28 |
+
"opt_betas": [
|
| 29 |
+
0.9,
|
| 30 |
+
0.95
|
| 31 |
+
],
|
| 32 |
+
"compute_grad_norm": true,
|
| 33 |
+
"clip_grad": 1.0,
|
| 34 |
+
"skip_grad": null,
|
| 35 |
+
"momentum": 0.9,
|
| 36 |
+
"weight_decay": 0.1,
|
| 37 |
+
"weight_decay_end": 0.1,
|
| 38 |
+
"blr": 0.001,
|
| 39 |
+
"min_blr": 5e-05,
|
| 40 |
+
"scheduler": "wsd",
|
| 41 |
+
"decay_fraction": 0.1,
|
| 42 |
+
"warmup_steps": 250,
|
| 43 |
+
"cooldown_steps": -1,
|
| 44 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml",
|
| 45 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml",
|
| 46 |
+
"cross_eval_val": "",
|
| 47 |
+
"s3_endpoint": "",
|
| 48 |
+
"s3_data_endpoint": "",
|
| 49 |
+
"s3_multipart_chunksize_mb": 512,
|
| 50 |
+
"s3_multipart_threshold_mb": 512,
|
| 51 |
+
"s3_max_io_queue": 100,
|
| 52 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 53 |
+
"eval_freq": 1,
|
| 54 |
+
"eval_freq_tokens": 2,
|
| 55 |
+
"dist_eval": true,
|
| 56 |
+
"eval": false,
|
| 57 |
+
"val_sampling_mode": "without_replacement",
|
| 58 |
+
"output_dir": "output/seed/t1_pure_molecule_xxs2_seed2",
|
| 59 |
+
"device": "cuda",
|
| 60 |
+
"seed": 2,
|
| 61 |
+
"resume": "",
|
| 62 |
+
"auto_resume": true,
|
| 63 |
+
"start_epoch": 0,
|
| 64 |
+
"num_workers": 10,
|
| 65 |
+
"pin_mem": true,
|
| 66 |
+
"find_unused_params": true,
|
| 67 |
+
"rlimit": 4096,
|
| 68 |
+
"print_all": false,
|
| 69 |
+
"show_user_warnings": false,
|
| 70 |
+
"s3_save_dir": "",
|
| 71 |
+
"dist_url": "env://",
|
| 72 |
+
"log_wandb": true,
|
| 73 |
+
"wandb_project": "loom",
|
| 74 |
+
"wandb_entity": null,
|
| 75 |
+
"wandb_run_name": "t1_pure_molecule_xxs2_seed2",
|
| 76 |
+
"wandb_group": "t1-seedvar-xxs2",
|
| 77 |
+
"dim": null,
|
| 78 |
+
"depth": null,
|
| 79 |
+
"num_heads": null,
|
| 80 |
+
"mlp_ratio": null,
|
| 81 |
+
"log_grad_norms_every": 0,
|
| 82 |
+
"save_at_tokens": "1,2,5,10,12",
|
| 83 |
+
"unigram_entropies": null,
|
| 84 |
+
"cooldown_branch": false,
|
| 85 |
+
"cooldown_tokens": 0,
|
| 86 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs2_seed2.yaml",
|
| 87 |
+
"rank": 0,
|
| 88 |
+
"world_size": 1,
|
| 89 |
+
"gpu": 0,
|
| 90 |
+
"distributed": true,
|
| 91 |
+
"dist_backend": "nccl",
|
| 92 |
+
"num_tasks": 1,
|
| 93 |
+
"sampling_mode": "without_replacement",
|
| 94 |
+
"max_micro_steps": 366211,
|
| 95 |
+
"lossbearing_target": 12000000000.0,
|
| 96 |
+
"lossbearing_mode": true,
|
| 97 |
+
"lr": 0.001,
|
| 98 |
+
"min_lr": 5e-05,
|
| 99 |
+
"global_micro_step": 366211,
|
| 100 |
+
"cum_lossbearing": 12000758024.0,
|
| 101 |
+
"cum_mod_tokens": {
|
| 102 |
+
"text": 881775534,
|
| 103 |
+
"tok_molecule": 11118982490
|
| 104 |
+
}
|
| 105 |
+
},
|
| 106 |
+
"source_configs": {
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs2_seed2.yaml": "# xxs2 \u03c3_seed variant of pure molecule (P1-ladder companion \u2014 the xxs2-scale training-seed\n# floor for the P1 sub-XS scaling anchors). Clone of run_xxs2_pure.yaml + seed 2;\n# seed0 = the existing output/p1/t1_pure_molecule_xxs2 anchor. Train-only self-cool \u2192 output/seed/.\n_base_: ../../../../models/dec/4m-xxs2-dec.yaml\nrun_name: t1_pure_molecule_xxs2_seed2\nseed: 2\ntotal_tokens: -1\nlossbearing_tokens: 12\ndata_config: cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,12\nwandb_group: t1-seedvar-xxs2\n",
|
| 108 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: molecule_zinc20m\n shards: data/shards/molecule_zinc20m_shuffled/train/shard_{000000..021643}.tar\n weight: 21643268\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc\n shards: data/shards/molecule_pubchemqc_reshuffled/train/shard_*.tar\n weight: 213751114\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20\n shards: data/shards/molecule_chebi20_shuffled/train/shard_{000000..000021}.tar\n weight: 21378\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap\n shards: data/shards/molecule_molcap_shuffled/train/shard_{000000..000070}.tar\n weight: 70805\n modality_tokens:\n tok_molecule: 29\n format: npz\nsampling_mode: without_replacement\n",
|
| 109 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: molecule_zinc20m_val\n shards: data/shards/molecule_zinc20m_shuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc_val\n shards: data/shards/molecule_pubchemqc_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20_val\n shards: data/shards/molecule_chebi20_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap_val\n shards: data/shards/molecule_molcap_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 29\n format: npz\n"
|
| 110 |
+
}
|
| 111 |
+
}
|
t1_pure_molecule_xxs3_seed1/checkpoint-10Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a9ca19f3ab09b34b3c4cde45507d48d873e7d7c503653c9a4c98c977b8dc4ad
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed1/checkpoint-12Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6631b2672dd9463d89d46b2ec6ccbb473092c9b4fe46a016be31abac3eed2097
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed1/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d32948fd3588bf6f7fd87001b453ec91a4e92402fa0882cc9d7949f26b861d1e
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed1/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:20fa63815919c47200cd7e01f5a71e034f8b53369aebf26d4549efed17e880a0
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed1/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5263edd6c9ea259bfe3afdead09d70a954ce763a7aeb7a07ed9bb502fa2a5385
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed1/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6fccb112b1e78f535eb48d98fc038c9b8d885e625696b85795d74f8ec53a9e5
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed1/config.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_molecule_xxs3_seed1",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 62342592,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_molecule_xxs3_seed1",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 12,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"effective_batch_size": null,
|
| 13 |
+
"lr_batch_scaling": "sqrt",
|
| 14 |
+
"save_ckpt_freq": 1,
|
| 15 |
+
"save_ckpt_freq_tokens": 5,
|
| 16 |
+
"debug_batch_stats": 0,
|
| 17 |
+
"model": "fm_xxs3_12d_swiglu_nobias",
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"loss_type": "token",
|
| 20 |
+
"context_length": 2048,
|
| 21 |
+
"pos_emb_ablation": "vanilla",
|
| 22 |
+
"max_blocks": 32,
|
| 23 |
+
"intra_pos_emb": "default",
|
| 24 |
+
"video_pos_3d": false,
|
| 25 |
+
"finetune": "",
|
| 26 |
+
"opt": "adamw",
|
| 27 |
+
"opt_eps": 1e-08,
|
| 28 |
+
"opt_betas": [
|
| 29 |
+
0.9,
|
| 30 |
+
0.95
|
| 31 |
+
],
|
| 32 |
+
"compute_grad_norm": true,
|
| 33 |
+
"clip_grad": 1.0,
|
| 34 |
+
"skip_grad": null,
|
| 35 |
+
"momentum": 0.9,
|
| 36 |
+
"weight_decay": 0.1,
|
| 37 |
+
"weight_decay_end": 0.1,
|
| 38 |
+
"blr": 0.001,
|
| 39 |
+
"min_blr": 5e-05,
|
| 40 |
+
"scheduler": "wsd",
|
| 41 |
+
"decay_fraction": 0.1,
|
| 42 |
+
"warmup_steps": 250,
|
| 43 |
+
"cooldown_steps": -1,
|
| 44 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml",
|
| 45 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml",
|
| 46 |
+
"cross_eval_val": "",
|
| 47 |
+
"s3_endpoint": "",
|
| 48 |
+
"s3_data_endpoint": "",
|
| 49 |
+
"s3_multipart_chunksize_mb": 512,
|
| 50 |
+
"s3_multipart_threshold_mb": 512,
|
| 51 |
+
"s3_max_io_queue": 100,
|
| 52 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 53 |
+
"eval_freq": 1,
|
| 54 |
+
"eval_freq_tokens": 2,
|
| 55 |
+
"dist_eval": true,
|
| 56 |
+
"eval": false,
|
| 57 |
+
"val_sampling_mode": "without_replacement",
|
| 58 |
+
"output_dir": "output/seed/t1_pure_molecule_xxs3_seed1",
|
| 59 |
+
"device": "cuda",
|
| 60 |
+
"seed": 1,
|
| 61 |
+
"resume": "",
|
| 62 |
+
"auto_resume": true,
|
| 63 |
+
"start_epoch": 0,
|
| 64 |
+
"num_workers": 10,
|
| 65 |
+
"pin_mem": true,
|
| 66 |
+
"find_unused_params": true,
|
| 67 |
+
"rlimit": 4096,
|
| 68 |
+
"print_all": false,
|
| 69 |
+
"show_user_warnings": false,
|
| 70 |
+
"s3_save_dir": "",
|
| 71 |
+
"dist_url": "env://",
|
| 72 |
+
"log_wandb": true,
|
| 73 |
+
"wandb_project": "loom",
|
| 74 |
+
"wandb_entity": null,
|
| 75 |
+
"wandb_run_name": "t1_pure_molecule_xxs3_seed1",
|
| 76 |
+
"wandb_group": "t1-seedvar-xxs3",
|
| 77 |
+
"dim": null,
|
| 78 |
+
"depth": null,
|
| 79 |
+
"num_heads": null,
|
| 80 |
+
"mlp_ratio": null,
|
| 81 |
+
"log_grad_norms_every": 0,
|
| 82 |
+
"save_at_tokens": "1,2,5,10,12",
|
| 83 |
+
"unigram_entropies": null,
|
| 84 |
+
"cooldown_branch": false,
|
| 85 |
+
"cooldown_tokens": 0,
|
| 86 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs3_seed1.yaml",
|
| 87 |
+
"rank": 0,
|
| 88 |
+
"world_size": 1,
|
| 89 |
+
"gpu": 0,
|
| 90 |
+
"distributed": true,
|
| 91 |
+
"dist_backend": "nccl",
|
| 92 |
+
"num_tasks": 1,
|
| 93 |
+
"sampling_mode": "without_replacement",
|
| 94 |
+
"max_micro_steps": 366211,
|
| 95 |
+
"lossbearing_target": 12000000000.0,
|
| 96 |
+
"lossbearing_mode": true,
|
| 97 |
+
"lr": 0.001,
|
| 98 |
+
"min_lr": 5e-05,
|
| 99 |
+
"global_micro_step": 366211,
|
| 100 |
+
"cum_lossbearing": 12000782833.0,
|
| 101 |
+
"cum_mod_tokens": {
|
| 102 |
+
"text": 881770354,
|
| 103 |
+
"tok_molecule": 11119012479
|
| 104 |
+
}
|
| 105 |
+
},
|
| 106 |
+
"source_configs": {
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs3_seed1.yaml": "# xxs3 \u03c3_seed variant of pure molecule (P1-ladder companion \u2014 the xxs3-scale training-seed\n# floor for the P1 sub-XS scaling anchors). Clone of run_xxs3_pure.yaml + seed 1;\n# seed0 = the existing output/p1/t1_pure_molecule_xxs3 anchor. Train-only self-cool \u2192 output/seed/.\n_base_: ../../../../models/dec/4m-xxs3-dec.yaml\nrun_name: t1_pure_molecule_xxs3_seed1\nseed: 1\ntotal_tokens: -1\nlossbearing_tokens: 12\ndata_config: cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,12\nwandb_group: t1-seedvar-xxs3\n",
|
| 108 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: molecule_zinc20m\n shards: data/shards/molecule_zinc20m_shuffled/train/shard_{000000..021643}.tar\n weight: 21643268\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc\n shards: data/shards/molecule_pubchemqc_reshuffled/train/shard_*.tar\n weight: 213751114\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20\n shards: data/shards/molecule_chebi20_shuffled/train/shard_{000000..000021}.tar\n weight: 21378\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap\n shards: data/shards/molecule_molcap_shuffled/train/shard_{000000..000070}.tar\n weight: 70805\n modality_tokens:\n tok_molecule: 29\n format: npz\nsampling_mode: without_replacement\n",
|
| 109 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: molecule_zinc20m_val\n shards: data/shards/molecule_zinc20m_shuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc_val\n shards: data/shards/molecule_pubchemqc_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20_val\n shards: data/shards/molecule_chebi20_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap_val\n shards: data/shards/molecule_molcap_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 29\n format: npz\n"
|
| 110 |
+
}
|
| 111 |
+
}
|
t1_pure_molecule_xxs3_seed2/checkpoint-10Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c82c20cb43bc22db5aa803ba6a8f302769ae122d340b8927713391e35c403822
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed2/checkpoint-12Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7d6774f632408b39d2d5a7b86a26d09467aa8cfcfdada154f11cecd703e46a3
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed2/checkpoint-1Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb15640b76fe03c2dd50426488c06813d2e556fb3fbc7ab9dfea4f4af6146ab0
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed2/checkpoint-2Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e9923772ec119cc70bbf9d1c129b1979af1da4071d6706eaa6634f1a8b0a8bf
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed2/checkpoint-5Blb.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffa3b6e054205d97ba1c426c33ffada914ba5f823964ff2bc7e56f8d5233df40
|
| 3 |
+
size 728067749
|
t1_pure_molecule_xxs3_seed2/checkpoint-final.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b0b89a01a8651fdc0a528d1e38e4c7167e65f5d99ef7a63aba8bfd76f04c70e
|
| 3 |
+
size 728068272
|
t1_pure_molecule_xxs3_seed2/config.json
ADDED
|
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_molecule_xxs3_seed2",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 62342592,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_molecule_xxs3_seed2",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 12,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"effective_batch_size": null,
|
| 13 |
+
"lr_batch_scaling": "sqrt",
|
| 14 |
+
"save_ckpt_freq": 1,
|
| 15 |
+
"save_ckpt_freq_tokens": 5,
|
| 16 |
+
"debug_batch_stats": 0,
|
| 17 |
+
"model": "fm_xxs3_12d_swiglu_nobias",
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"loss_type": "token",
|
| 20 |
+
"context_length": 2048,
|
| 21 |
+
"pos_emb_ablation": "vanilla",
|
| 22 |
+
"max_blocks": 32,
|
| 23 |
+
"intra_pos_emb": "default",
|
| 24 |
+
"video_pos_3d": false,
|
| 25 |
+
"finetune": "",
|
| 26 |
+
"opt": "adamw",
|
| 27 |
+
"opt_eps": 1e-08,
|
| 28 |
+
"opt_betas": [
|
| 29 |
+
0.9,
|
| 30 |
+
0.95
|
| 31 |
+
],
|
| 32 |
+
"compute_grad_norm": true,
|
| 33 |
+
"clip_grad": 1.0,
|
| 34 |
+
"skip_grad": null,
|
| 35 |
+
"momentum": 0.9,
|
| 36 |
+
"weight_decay": 0.1,
|
| 37 |
+
"weight_decay_end": 0.1,
|
| 38 |
+
"blr": 0.001,
|
| 39 |
+
"min_blr": 5e-05,
|
| 40 |
+
"scheduler": "wsd",
|
| 41 |
+
"decay_fraction": 0.1,
|
| 42 |
+
"warmup_steps": 250,
|
| 43 |
+
"cooldown_steps": -1,
|
| 44 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml",
|
| 45 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml",
|
| 46 |
+
"cross_eval_val": "",
|
| 47 |
+
"s3_endpoint": "",
|
| 48 |
+
"s3_data_endpoint": "",
|
| 49 |
+
"s3_multipart_chunksize_mb": 512,
|
| 50 |
+
"s3_multipart_threshold_mb": 512,
|
| 51 |
+
"s3_max_io_queue": 100,
|
| 52 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 53 |
+
"eval_freq": 1,
|
| 54 |
+
"eval_freq_tokens": 2,
|
| 55 |
+
"dist_eval": true,
|
| 56 |
+
"eval": false,
|
| 57 |
+
"val_sampling_mode": "without_replacement",
|
| 58 |
+
"output_dir": "output/seed/t1_pure_molecule_xxs3_seed2",
|
| 59 |
+
"device": "cuda",
|
| 60 |
+
"seed": 2,
|
| 61 |
+
"resume": "",
|
| 62 |
+
"auto_resume": true,
|
| 63 |
+
"start_epoch": 0,
|
| 64 |
+
"num_workers": 10,
|
| 65 |
+
"pin_mem": true,
|
| 66 |
+
"find_unused_params": true,
|
| 67 |
+
"rlimit": 4096,
|
| 68 |
+
"print_all": false,
|
| 69 |
+
"show_user_warnings": false,
|
| 70 |
+
"s3_save_dir": "",
|
| 71 |
+
"dist_url": "env://",
|
| 72 |
+
"log_wandb": true,
|
| 73 |
+
"wandb_project": "loom",
|
| 74 |
+
"wandb_entity": null,
|
| 75 |
+
"wandb_run_name": "t1_pure_molecule_xxs3_seed2",
|
| 76 |
+
"wandb_group": "t1-seedvar-xxs3",
|
| 77 |
+
"dim": null,
|
| 78 |
+
"depth": null,
|
| 79 |
+
"num_heads": null,
|
| 80 |
+
"mlp_ratio": null,
|
| 81 |
+
"log_grad_norms_every": 0,
|
| 82 |
+
"save_at_tokens": "1,2,5,10,12",
|
| 83 |
+
"unigram_entropies": null,
|
| 84 |
+
"cooldown_branch": false,
|
| 85 |
+
"cooldown_tokens": 0,
|
| 86 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs3_seed2.yaml",
|
| 87 |
+
"rank": 0,
|
| 88 |
+
"world_size": 1,
|
| 89 |
+
"gpu": 0,
|
| 90 |
+
"distributed": true,
|
| 91 |
+
"dist_backend": "nccl",
|
| 92 |
+
"num_tasks": 1,
|
| 93 |
+
"sampling_mode": "without_replacement",
|
| 94 |
+
"max_micro_steps": 366211,
|
| 95 |
+
"lossbearing_target": 12000000000.0,
|
| 96 |
+
"lossbearing_mode": true,
|
| 97 |
+
"lr": 0.001,
|
| 98 |
+
"min_lr": 5e-05,
|
| 99 |
+
"global_micro_step": 366211,
|
| 100 |
+
"cum_lossbearing": 12000765924.0,
|
| 101 |
+
"cum_mod_tokens": {
|
| 102 |
+
"text": 881775682,
|
| 103 |
+
"tok_molecule": 11118990242
|
| 104 |
+
}
|
| 105 |
+
},
|
| 106 |
+
"source_configs": {
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/run_xxs3_seed2.yaml": "# xxs3 \u03c3_seed variant of pure molecule (P1-ladder companion \u2014 the xxs3-scale training-seed\n# floor for the P1 sub-XS scaling anchors). Clone of run_xxs3_pure.yaml + seed 2;\n# seed0 = the existing output/p1/t1_pure_molecule_xxs3 anchor. Train-only self-cool \u2192 output/seed/.\n_base_: ../../../../models/dec/4m-xxs3-dec.yaml\nrun_name: t1_pure_molecule_xxs3_seed2\nseed: 2\ntotal_tokens: -1\nlossbearing_tokens: 12\ndata_config: cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,12\nwandb_group: t1-seedvar-xxs3\n",
|
| 108 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: molecule_zinc20m\n shards: data/shards/molecule_zinc20m_shuffled/train/shard_{000000..021643}.tar\n weight: 21643268\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc\n shards: data/shards/molecule_pubchemqc_reshuffled/train/shard_*.tar\n weight: 213751114\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20\n shards: data/shards/molecule_chebi20_shuffled/train/shard_{000000..000021}.tar\n weight: 21378\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap\n shards: data/shards/molecule_molcap_shuffled/train/shard_{000000..000070}.tar\n weight: 70805\n modality_tokens:\n tok_molecule: 29\n format: npz\nsampling_mode: without_replacement\n",
|
| 109 |
+
"cfgs/default/4m/experiments/t1/unimodal/molecule/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 molecule\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_molecule\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: molecule_zinc20m_val\n shards: data/shards/molecule_zinc20m_shuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 24\n format: npz\n- name: molecule_pubchemqc_val\n shards: data/shards/molecule_pubchemqc_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_molecule: 14\n format: npz\n- name: molecule_chebi20_val\n shards: data/shards/molecule_chebi20_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 26\n format: npz\n- name: molecule_molcap_val\n shards: data/shards/molecule_molcap_shuffled/val/shard_{000000..000001}.tar\n weight: 2000\n modality_tokens:\n tok_molecule: 29\n format: npz\n"
|
| 110 |
+
}
|
| 111 |
+
}
|
t1_pure_satellite_s_seed1/config.json
ADDED
|
@@ -0,0 +1,109 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "t1_pure_satellite_s_seed1",
|
| 3 |
+
"epoch": 0,
|
| 4 |
+
"n_params": 167075616,
|
| 5 |
+
"args": {
|
| 6 |
+
"run_name": "t1_pure_satellite_s_seed1",
|
| 7 |
+
"batch_size": 32,
|
| 8 |
+
"epochs": 1,
|
| 9 |
+
"total_tokens": -1,
|
| 10 |
+
"lossbearing_tokens": 20,
|
| 11 |
+
"accum_iter": 32,
|
| 12 |
+
"save_ckpt_freq": 1,
|
| 13 |
+
"save_ckpt_freq_tokens": 5,
|
| 14 |
+
"debug_batch_stats": 0,
|
| 15 |
+
"model": "fm_small_12d_swiglu_nobias",
|
| 16 |
+
"dtype": "bfloat16",
|
| 17 |
+
"loss_type": "token",
|
| 18 |
+
"context_length": 2048,
|
| 19 |
+
"pos_emb_ablation": "vanilla",
|
| 20 |
+
"max_blocks": 32,
|
| 21 |
+
"intra_pos_emb": "default",
|
| 22 |
+
"video_pos_3d": false,
|
| 23 |
+
"finetune": "",
|
| 24 |
+
"opt": "adamw",
|
| 25 |
+
"opt_eps": 1e-08,
|
| 26 |
+
"opt_betas": [
|
| 27 |
+
0.9,
|
| 28 |
+
0.95
|
| 29 |
+
],
|
| 30 |
+
"compute_grad_norm": true,
|
| 31 |
+
"clip_grad": 1.0,
|
| 32 |
+
"skip_grad": null,
|
| 33 |
+
"momentum": 0.9,
|
| 34 |
+
"weight_decay": 0.1,
|
| 35 |
+
"weight_decay_end": 0.1,
|
| 36 |
+
"blr": 0.001,
|
| 37 |
+
"min_blr": 5e-05,
|
| 38 |
+
"scheduler": "wsd",
|
| 39 |
+
"decay_fraction": 0.1,
|
| 40 |
+
"warmup_steps": 250,
|
| 41 |
+
"cooldown_steps": -1,
|
| 42 |
+
"data_config": "cfgs/default/4m/experiments/t1/unimodal/satellite/data_train_pure.yaml",
|
| 43 |
+
"data_config_val": "cfgs/default/4m/experiments/t1/unimodal/satellite/data_val_pure.yaml",
|
| 44 |
+
"cross_eval_val": "",
|
| 45 |
+
"s3_endpoint": "",
|
| 46 |
+
"s3_data_endpoint": "",
|
| 47 |
+
"s3_multipart_chunksize_mb": 512,
|
| 48 |
+
"s3_multipart_threshold_mb": 512,
|
| 49 |
+
"s3_max_io_queue": 100,
|
| 50 |
+
"text_tokenizer_path": "fourm/utils/tokenizer/trained/smollm_tokenizer.json",
|
| 51 |
+
"eval_freq": 1,
|
| 52 |
+
"eval_freq_tokens": 2,
|
| 53 |
+
"dist_eval": true,
|
| 54 |
+
"eval": false,
|
| 55 |
+
"val_sampling_mode": "without_replacement",
|
| 56 |
+
"output_dir": "output/seed/t1_pure_satellite_s_seed1",
|
| 57 |
+
"device": "cuda",
|
| 58 |
+
"seed": 1,
|
| 59 |
+
"resume": "",
|
| 60 |
+
"auto_resume": true,
|
| 61 |
+
"start_epoch": 0,
|
| 62 |
+
"num_workers": 10,
|
| 63 |
+
"pin_mem": true,
|
| 64 |
+
"find_unused_params": true,
|
| 65 |
+
"rlimit": 4096,
|
| 66 |
+
"print_all": false,
|
| 67 |
+
"show_user_warnings": false,
|
| 68 |
+
"s3_save_dir": "",
|
| 69 |
+
"dist_url": "env://",
|
| 70 |
+
"log_wandb": true,
|
| 71 |
+
"wandb_project": "loom",
|
| 72 |
+
"wandb_entity": null,
|
| 73 |
+
"wandb_run_name": "nr_t1_pure_satellite_s_seed1",
|
| 74 |
+
"wandb_group": "t1-seedvar",
|
| 75 |
+
"dim": null,
|
| 76 |
+
"depth": null,
|
| 77 |
+
"num_heads": null,
|
| 78 |
+
"mlp_ratio": null,
|
| 79 |
+
"log_grad_norms_every": 0,
|
| 80 |
+
"save_at_tokens": "1,2,5,10,15,20",
|
| 81 |
+
"unigram_entropies": null,
|
| 82 |
+
"cooldown_branch": false,
|
| 83 |
+
"cooldown_tokens": 0,
|
| 84 |
+
"config_path": "cfgs/default/4m/experiments/t1/unimodal/satellite/run_seed1.yaml",
|
| 85 |
+
"rank": 0,
|
| 86 |
+
"world_size": 1,
|
| 87 |
+
"gpu": 0,
|
| 88 |
+
"distributed": true,
|
| 89 |
+
"dist_backend": "nccl",
|
| 90 |
+
"num_tasks": 1,
|
| 91 |
+
"sampling_mode": "without_replacement",
|
| 92 |
+
"max_micro_steps": 610352,
|
| 93 |
+
"lossbearing_target": 20000000000.0,
|
| 94 |
+
"lossbearing_mode": true,
|
| 95 |
+
"lr": 0.001,
|
| 96 |
+
"min_lr": 5e-05,
|
| 97 |
+
"global_micro_step": 610352,
|
| 98 |
+
"cum_lossbearing": 20000203776.0,
|
| 99 |
+
"cum_mod_tokens": {
|
| 100 |
+
"text": 154441728,
|
| 101 |
+
"tok_satellite@128": 19845762048
|
| 102 |
+
}
|
| 103 |
+
},
|
| 104 |
+
"source_configs": {
|
| 105 |
+
"cfgs/default/4m/experiments/t1/unimodal/satellite/run_seed1.yaml": "# Seed variant of pure satellite-S for satellite \u03c3_seed (D4-1 #1b \u2014 the run-to-run floor\n# gating satellite-recipient interaction cells; video is the video\u21903d reversal's recipient).\n# Clone of run_pure.yaml + seed 1. seed0 = output/nr/t1_pure_satellite_s. Train-only self-cool.\n_base_: ../../../../models/dec/4m-s-dec.yaml\nrun_name: t1_pure_satellite_s_seed1\nseed: 1\ntotal_tokens: -1\nlossbearing_tokens: 20\ndata_config: cfgs/default/4m/experiments/t1/unimodal/satellite/data_train_pure.yaml\ndata_config_val: cfgs/default/4m/experiments/t1/unimodal/satellite/data_val_pure.yaml\nwarmup_steps: 250\neval_freq_tokens: 2\nsave_ckpt_freq_tokens: 5\nsave_at_tokens: 1,2,5,10,15,20\nwandb_group: t1-seedvar\n",
|
| 106 |
+
"cfgs/default/4m/experiments/t1/unimodal/satellite/data_train_pure.yaml": "# T1 pure unimodal training data \u2014 satellite\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_satellite@128\nepoch_size: 50000000\nsampler_type: temperature\ntemperature_tau: 1.0\nsources:\n- name: satellite_v2\n shards: data/shards/satellite_v2_reshuffled/train/shard_*.tar\n weight: 22408820\n modality_tokens:\n tok_satellite@128: 256\n format: npz\n- name: satellite_ben\n shards: data/shards/satellite_ben_shuffled/train/shard_{000000..000470}.tar\n weight: 470438\n modality_tokens:\n tok_satellite@128: 256\n format: npz\nsampling_mode: without_replacement\n",
|
| 107 |
+
"cfgs/default/4m/experiments/t1/unimodal/satellite/data_val_pure.yaml": "# T1 pure unimodal validation data \u2014 satellite\n# Text stripped from all sources (pure modality-only condition)\n\nmodalities:\n- tok_satellite@128\nval_steps: 200\nsampler_type: uniform\nsources:\n- name: satellite_v2_val\n shards: data/shards/satellite_v2_reshuffled/val/shard_{000000..000049}.tar\n weight: 50000\n modality_tokens:\n tok_satellite@128: 256\n format: npz\n- name: satellite_ben_val\n shards: data/shards/satellite_ben_shuffled/val/shard_{000000..000009}.tar\n weight: 9600\n modality_tokens:\n tok_satellite@128: 256\n format: npz\n"
|
| 108 |
+
}
|
| 109 |
+
}
|