Hugging Face's logo

genshiai-daichi
/
med-ruri-v3-70m-v2-from-ruri

Sentence Similarity
sentence-transformers
Safetensors
Japanese
modernbert
medical
japanese
ruri
embedding
text-embeddings-inference
Model card Files
xet
 Community
genshiai-daichi commited on
Commit
68fa8db
·
verified ·
1 Parent(s): acc436f

latest = step 3000 (nDCG@10=0.4497)

Browse files
Files changed (2) hide show
  1. model.safetensors +1 -1
  2. trainer_state.json +94 -5
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:33a3868e6fe2ae76d21a8cbf1319856f8fee95e6b00b6f1f9610130c50184270
3
  size 280019040
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21f547abc5e4116f4b37cf51af764d7b6347cd3bcd74732964b5ec09a95d6393
3
  size 280019040
trainer_state.json CHANGED
@@ -1,10 +1,10 @@
1
  {
2
- "best_global_step": 2500,
3
- "best_metric": 0.44659083423553114,
4
- "best_model_checkpoint": "checkpoints/med-ruri-v3-70m-v2-from-ruri/checkpoint-2500",
5
- "epoch": 0.4059100503328462,
6
  "eval_steps": 500,
7
- "global_step": 2500,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
@@ -453,6 +453,95 @@
453
  "eval_samples_per_second": 0.0,
454
  "eval_steps_per_second": 0.0,
455
  "step": 2500
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
456
  }
457
  ],
458
  "logging_steps": 50,
 
1
  {
2
+ "best_global_step": 3000,
3
+ "best_metric": 0.4496575061053132,
4
+ "best_model_checkpoint": "checkpoints/med-ruri-v3-70m-v2-from-ruri/checkpoint-3000",
5
+ "epoch": 0.4870920603994155,
6
  "eval_steps": 500,
7
+ "global_step": 3000,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
 
453
  "eval_samples_per_second": 0.0,
454
  "eval_steps_per_second": 0.0,
455
  "step": 2500
456
+ },
457
+ {
458
+ "epoch": 0.41402825133950316,
459
+ "grad_norm": 23.17564582824707,
460
+ "learning_rate": 6.169885489659888e-06,
461
+ "loss": 0.5776227569580078,
462
+ "step": 2550
463
+ },
464
+ {
465
+ "epoch": 0.4221464523461601,
466
+ "grad_norm": 28.088123321533203,
467
+ "learning_rate": 6.084430011963768e-06,
468
+ "loss": 0.61005126953125,
469
+ "step": 2600
470
+ },
471
+ {
472
+ "epoch": 0.430264653352817,
473
+ "grad_norm": 20.892213821411133,
474
+ "learning_rate": 5.998974534267648e-06,
475
+ "loss": 0.5894330978393555,
476
+ "step": 2650
477
+ },
478
+ {
479
+ "epoch": 0.43838285435947394,
480
+ "grad_norm": 29.862913131713867,
481
+ "learning_rate": 5.913519056571527e-06,
482
+ "loss": 0.5896952819824218,
483
+ "step": 2700
484
+ },
485
+ {
486
+ "epoch": 0.4465010553661309,
487
+ "grad_norm": 23.289093017578125,
488
+ "learning_rate": 5.828063578875406e-06,
489
+ "loss": 0.5747700500488281,
490
+ "step": 2750
491
+ },
492
+ {
493
+ "epoch": 0.45461925637278777,
494
+ "grad_norm": 24.890533447265625,
495
+ "learning_rate": 5.742608101179286e-06,
496
+ "loss": 0.5570634078979492,
497
+ "step": 2800
498
+ },
499
+ {
500
+ "epoch": 0.4627374573794447,
501
+ "grad_norm": 25.896406173706055,
502
+ "learning_rate": 5.657152623483165e-06,
503
+ "loss": 0.561774673461914,
504
+ "step": 2850
505
+ },
506
+ {
507
+ "epoch": 0.47085565838610166,
508
+ "grad_norm": 24.267595291137695,
509
+ "learning_rate": 5.571697145787045e-06,
510
+ "loss": 0.5851204299926758,
511
+ "step": 2900
512
+ },
513
+ {
514
+ "epoch": 0.47897385939275855,
515
+ "grad_norm": 25.277156829833984,
516
+ "learning_rate": 5.486241668090925e-06,
517
+ "loss": 0.5668196105957031,
518
+ "step": 2950
519
+ },
520
+ {
521
+ "epoch": 0.4870920603994155,
522
+ "grad_norm": 24.794769287109375,
523
+ "learning_rate": 5.400786190394805e-06,
524
+ "loss": 0.529843521118164,
525
+ "step": 3000
526
+ },
527
+ {
528
+ "epoch": 0.4870920603994155,
529
+ "eval_med-ir_cosine_accuracy@1": 0.2475,
530
+ "eval_med-ir_cosine_accuracy@10": 0.684,
531
+ "eval_med-ir_cosine_accuracy@5": 0.5575,
532
+ "eval_med-ir_cosine_map@10": 0.3764117063492064,
533
+ "eval_med-ir_cosine_mrr@10": 0.3764117063492073,
534
+ "eval_med-ir_cosine_ndcg@10": 0.4496575061053132,
535
+ "eval_med-ir_cosine_precision@1": 0.2475,
536
+ "eval_med-ir_cosine_precision@10": 0.06839999999999999,
537
+ "eval_med-ir_cosine_precision@5": 0.11149999999999999,
538
+ "eval_med-ir_cosine_recall@1": 0.2475,
539
+ "eval_med-ir_cosine_recall@10": 0.684,
540
+ "eval_med-ir_cosine_recall@5": 0.5575,
541
+ "eval_runtime": 53.1101,
542
+ "eval_samples_per_second": 0.0,
543
+ "eval_steps_per_second": 0.0,
544
+ "step": 3000
545
  }
546
  ],
547
  "logging_steps": 50,