| { | |
| "name": "mars", | |
| "optimizer": "mars", | |
| "status": "ok", | |
| "diverged_at_step": null, | |
| "tokens_seen": 37806080, | |
| "steps": 4615, | |
| "total_steps": 4615, | |
| "token_frac": 1.0, | |
| "train_tokens_per_epoch": 37813248, | |
| "final_val_loss": 4.083916366833703, | |
| "final_bleu": 3.121595483131984, | |
| "best_val_loss": 4.083916366833703, | |
| "train_time_s": 1512.0, | |
| "wall_time_s": 1655.0, | |
| "sec_per_step": 0.32762018043336966, | |
| "opt_step_ms": 41.85049692734608, | |
| "optimizer_state_bytes": 500379920, | |
| "peak_mem_gb": 3.12, | |
| "total_params": 41698304, | |
| "hparams": { | |
| "name": "mars", | |
| "lr": 0.002, | |
| "betas": [ | |
| 0.9, | |
| 0.95 | |
| ], | |
| "eps": 1e-08, | |
| "weight_decay": 0.1, | |
| "gamma": 0.025, | |
| "warmup_steps": 300, | |
| "schedule": "cosine", | |
| "min_lr_frac": 0.1, | |
| "batch_tokens": 8192, | |
| "grad_clip": 1.0 | |
| }, | |
| "history": [ | |
| { | |
| "val_loss": 5.5644740222707245, | |
| "val_ppl": 260.98789378221204, | |
| "bleu": 1.3933804109520285, | |
| "bleu_human": 0.33815425204965927, | |
| "bleu_n": 497, | |
| "fraction": 0.1, | |
| "step": 462, | |
| "tokens_seen": 3784704, | |
| "train_time_s": 151.0, | |
| "wall_time_s": 164.6 | |
| }, | |
| { | |
| "val_loss": 5.113784666392663, | |
| "val_ppl": 166.29854990254802, | |
| "bleu": 2.6254619253922757, | |
| "bleu_human": 0.537738897277919, | |
| "bleu_n": 497, | |
| "fraction": 0.2, | |
| "step": 923, | |
| "tokens_seen": 7561216, | |
| "train_time_s": 302.1, | |
| "wall_time_s": 329.7 | |
| }, | |
| { | |
| "val_loss": 4.862805895018096, | |
| "val_ppl": 129.38673885545705, | |
| "bleu": 2.1623057665899803, | |
| "bleu_human": 0.3577395786141275, | |
| "bleu_n": 497, | |
| "fraction": 0.3, | |
| "step": 1385, | |
| "tokens_seen": 11345920, | |
| "train_time_s": 453.6, | |
| "wall_time_s": 495.5 | |
| }, | |
| { | |
| "val_loss": 4.689373418502267, | |
| "val_ppl": 108.78499578263424, | |
| "bleu": 3.088920894442003, | |
| "bleu_human": 0.6577046757389468, | |
| "bleu_n": 497, | |
| "fraction": 0.4, | |
| "step": 1846, | |
| "tokens_seen": 15122432, | |
| "train_time_s": 604.6, | |
| "wall_time_s": 660.9 | |
| }, | |
| { | |
| "val_loss": 4.5270724500977835, | |
| "val_ppl": 92.48740287394348, | |
| "bleu": 2.8014124634589574, | |
| "bleu_human": 0.558930500151262, | |
| "bleu_n": 497, | |
| "fraction": 0.5, | |
| "step": 2308, | |
| "tokens_seen": 18907136, | |
| "train_time_s": 755.9, | |
| "wall_time_s": 826.6 | |
| }, | |
| { | |
| "val_loss": 4.389742649881667, | |
| "val_ppl": 80.6196688292514, | |
| "bleu": 2.9227675880427477, | |
| "bleu_human": 0.49167359125807264, | |
| "bleu_n": 497, | |
| "fraction": 0.6, | |
| "step": 2770, | |
| "tokens_seen": 22691840, | |
| "train_time_s": 907.2, | |
| "wall_time_s": 992.5 | |
| }, | |
| { | |
| "val_loss": 4.282210729879905, | |
| "val_ppl": 72.40032076805875, | |
| "bleu": 2.8767315006059877, | |
| "bleu_human": 0.4865830846335313, | |
| "bleu_n": 497, | |
| "fraction": 0.7, | |
| "step": 3231, | |
| "tokens_seen": 26468352, | |
| "train_time_s": 1058.2, | |
| "wall_time_s": 1157.8 | |
| }, | |
| { | |
| "val_loss": 4.188797038677841, | |
| "val_ppl": 65.9434158486703, | |
| "bleu": 2.796587438315598, | |
| "bleu_human": 0.5671621194475668, | |
| "bleu_n": 497, | |
| "fraction": 0.8, | |
| "step": 3693, | |
| "tokens_seen": 30253056, | |
| "train_time_s": 1209.6, | |
| "wall_time_s": 1323.6 | |
| }, | |
| { | |
| "val_loss": 4.122510206694202, | |
| "val_ppl": 61.713962795464134, | |
| "bleu": 2.8315733169687958, | |
| "bleu_human": 0.5846257376730375, | |
| "bleu_n": 497, | |
| "fraction": 0.9, | |
| "step": 4154, | |
| "tokens_seen": 34029568, | |
| "train_time_s": 1360.7, | |
| "wall_time_s": 1489.3 | |
| }, | |
| { | |
| "val_loss": 4.083916366833703, | |
| "val_ppl": 59.37755938359462, | |
| "bleu": 3.121595483131984, | |
| "bleu_human": 0.5471262061546667, | |
| "bleu_n": 497, | |
| "fraction": 1.0, | |
| "step": 4615, | |
| "tokens_seen": 37806080, | |
| "train_time_s": 1512.0, | |
| "wall_time_s": 1655.0 | |
| } | |
| ], | |
| "data_meta": { | |
| "n_sources": 8290, | |
| "n_train_sources": 7627, | |
| "n_val_sources": 166, | |
| "n_test_sources": 497, | |
| "train_docs": 61016, | |
| "val_docs": 1328, | |
| "train_tokens": 37813248, | |
| "val_tokens": 833536, | |
| "reference_authors": [ | |
| "chunk_2", | |
| "Meta-Llama-3-70B", | |
| "Meta-Llama-3-70B-Instruct", | |
| "Meta-Llama-3-8B", | |
| "Meta-Llama-3-8B-Instruct", | |
| "gpt-4o-2024-08-06", | |
| "gpt-4o-mini-2024-07-18" | |
| ], | |
| "tokens_per_word": 1.2684311662630021 | |
| }, | |
| "wandb_url": "https://wandb.ai/frenchknuckles-iiit-hyderabad/anlp-a2/runs/dec1586b", | |
| "meta": { | |
| "git_commit": "6a52970-dirty", | |
| "host": "HP-Victus-16", | |
| "platform": "Windows-11-10.0.26200-SP0", | |
| "python": "3.13.12", | |
| "torch": "2.14.0+cu126", | |
| "cuda": "12.6", | |
| "started_at": "2026-10-03T03:35:29+05:30", | |
| "gpu": "NVIDIA GeForce RTX 3050 6GB Laptop GPU", | |
| "gpu_memory_gb": 6.44 | |
| } | |
| } |