Training in progress, step 70000, checkpoint

last-checkpoint/generation_config.json

@@ -1,5 +1,4 @@
 {
-  "_from_model_config": true,
   "bos_token_id": 0,
   "do_sample": true,
   "eos_token_id": 2,

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d13683a9eb526c2a11e683a04934125cdfecdc6028c80a7aa087da26cca7114
+oid sha256:ee092c79e0b44f0dbff01ab92e7f2745a42f095c45606ecec1c465ecbfe2d1b3
 size 409608164

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:96ce67f9097c66e54e5a2f0ed74746f022227fc676edddfa5a5c87f451e263c6
+oid sha256:8461b39005aebca4cb2be6a57004dbbdfb857c61fa84fee72b747343bf9e4412
 size 814647162

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c97d92f20ad7b47d803d2c7fedc336e53e56e50a1239afbee7271f7ae31f6eaf
+oid sha256:ed15955fd3f000f6fbb48e527bdffe75e24c0ba876edf3b37cfb4740d80018cd
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84adf9c24f99fb570b7897d981d0c3d4eed31905ad350fe23cb92d1fd04362b7
+oid sha256:dbbf4e1e40678ab33c27d09cdf6c6af16557f9927a45e80add2e5a683e3f84f9
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.5317811614730878,
-  "best_model_checkpoint": "/mimer/NOBACKUP/groups/naiss2023-6-290/stefano/models//PROTAC-Splitter-EncoderDecoder-lr_cosine_restarts-opt25-rand-smiles/checkpoint-10000",
-  "epoch": 6.411521010061156,
+  "best_metric": 0.5904745042492918,
+  "best_model_checkpoint": "/mimer/NOBACKUP/groups/naiss2023-6-290/stefano/models//PROTAC-Splitter-EncoderDecoder-lr_cosine_restarts-rand-smiles/checkpoint-70000",
+  "epoch": 6.904714933912015,
   "eval_steps": 2500,
-  "global_step": 65000,
+  "global_step": 70000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1043,6 +1043,118 @@
       "learning_rate": 3.19382754340058e-05,
       "loss": 0.0008,
       "step": 65000
+    },
+    {
+      "epoch": 6.460840402446242,
+      "grad_norm": 0.022759919986128807,
+      "learning_rate": 2.0184879065157796e-05,
+      "loss": 0.0007,
+      "step": 65500
+    },
+    {
+      "epoch": 6.510159794831328,
+      "grad_norm": 0.021681690588593483,
+      "learning_rate": 9.497793269840106e-06,
+      "loss": 0.0006,
+      "step": 66000
+    },
+    {
+      "epoch": 6.5594791872164135,
+      "grad_norm": 0.016086483374238014,
+      "learning_rate": 2.243677730743965e-06,
+      "loss": 0.0005,
+      "step": 66500
+    },
+    {
+      "epoch": 6.608798579601499,
+      "grad_norm": 0.015512198209762573,
+      "learning_rate": 4.997217223258711e-05,
+      "loss": 0.0005,
+      "step": 67000
+    },
+    {
+      "epoch": 6.658117971986585,
+      "grad_norm": 0.05058777704834938,
+      "learning_rate": 4.670322405614625e-05,
+      "loss": 0.0022,
+      "step": 67500
+    },
+    {
+      "epoch": 6.707437364371671,
+      "grad_norm": 0.02859262190759182,
+      "learning_rate": 3.8644736479671964e-05,
+      "loss": 0.0018,
+      "step": 68000
+    },
+    {
+      "epoch": 6.756756756756757,
+      "grad_norm": 0.027067149057984352,
+      "learning_rate": 2.7552837001595315e-05,
+      "loss": 0.0008,
+      "step": 68500
+    },
+    {
+      "epoch": 6.806076149141843,
+      "grad_norm": 0.015645721927285194,
+      "learning_rate": 1.589561068663654e-05,
+      "loss": 0.0006,
+      "step": 69000
+    },
+    {
+      "epoch": 6.855395541526929,
+      "grad_norm": 0.022519439458847046,
+      "learning_rate": 6.254555247081101e-06,
+      "loss": 0.0005,
+      "step": 69500
+    },
+    {
+      "epoch": 6.904714933912015,
+      "grad_norm": 0.02764889970421791,
+      "learning_rate": 7.646864996253444e-07,
+      "loss": 0.0005,
+      "step": 70000
+    },
+    {
+      "epoch": 6.904714933912015,
+      "eval_all_ligands_equal": 0.5904745042492918,
+      "eval_e3_equal": 0.8272839943342776,
+      "eval_e3_graph_edit_distance": Infinity,
+      "eval_e3_graph_edit_distance_norm": Infinity,
+      "eval_e3_has_attachment_point(s)": 0.9890226628895185,
+      "eval_e3_heavy_atoms_difference": 0.3207330028328612,
+      "eval_e3_heavy_atoms_difference_norm": 0.0015200617544263726,
+      "eval_e3_tanimoto_similarity": 0.0,
+      "eval_e3_valid": 0.9890226628895185,
+      "eval_has_all_attachment_points": 0.9891111898016998,
+      "eval_has_three_substructures": 0.9995573654390935,
+      "eval_heavy_atoms_difference": 4.73371104815864,
+      "eval_heavy_atoms_difference_norm": 0.061486438625428276,
+      "eval_linker_equal": 0.8439270538243626,
+      "eval_linker_graph_edit_distance": 2.301699716713881e+61,
+      "eval_linker_graph_edit_distance_norm": Infinity,
+      "eval_linker_has_attachment_point(s)": 0.9976983002832861,
+      "eval_linker_heavy_atoms_difference": 0.20095609065155806,
+      "eval_linker_heavy_atoms_difference_norm": 0.0008608402807775271,
+      "eval_linker_tanimoto_similarity": 0.0,
+      "eval_linker_valid": 0.9976983002832861,
+      "eval_loss": 0.33629921078681946,
+      "eval_num_fragments": 2.999734419263456,
+      "eval_poi_equal": 0.7900141643059491,
+      "eval_poi_graph_edit_distance": Infinity,
+      "eval_poi_graph_edit_distance_norm": Infinity,
+      "eval_poi_has_attachment_point(s)": 0.9573300283286119,
+      "eval_poi_heavy_atoms_difference": 1.3261331444759208,
+      "eval_poi_heavy_atoms_difference_norm": 0.0403868595119425,
+      "eval_poi_tanimoto_similarity": 0.0,
+      "eval_poi_valid": 0.9573300283286119,
+      "eval_reassembly": 0.5982648725212465,
+      "eval_reassembly_nostereo": 0.6325247875354107,
+      "eval_runtime": 2333.1082,
+      "eval_samples_per_second": 4.842,
+      "eval_steps_per_second": 0.076,
+      "eval_tanimoto_similarity": 0.0,
+      "eval_valid": 0.9452018413597734,
+      "step": 70000
     }
   ],
   "logging_steps": 500,
@@ -1062,7 +1174,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.674161087565732e+17,
+  "total_flos": 6.110533685712476e+17,
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": null