| { |
| "step": 39000, |
| "tokens": 3833856000, |
| "parameters": { |
| "total": 1043390336, |
| "active": 384849792, |
| "shared": 55579520, |
| "active_path": 329270272 |
| }, |
| "batch_size": 96, |
| "sequence_length": 1024, |
| "loss": 3.296875, |
| "auxiliary_loss": 3.90625, |
| "gradient_norm": 0.279296875, |
| "route": 2, |
| "lr_scale": 0.7106121392357294, |
| "tokens_per_second": 280915.36663150863, |
| "elapsed_seconds": 13862.435317914002, |
| "average_patch_bytes": 4.383288001070139, |
| "tokenizer": "mistralai/Mistral-7B-v0.1", |
| "patch_coalesced_rate": 0.5729166666666666, |
| "gate": 0.515625, |
| "peak_vram_gib": 53.84470319747925, |
| "documents": { |
| "HuggingFaceFW/fineweb-edu:sample-10BT": 1239006, |
| "HuggingFaceTB/finemath:finemath-4plus": 792924, |
| "HuggingFaceTB/smollm-corpus:cosmopedia-v2": 491075, |
| "open-r1/OpenR1-Math-220k:default": 1016956, |
| "open-r1/codeforces-cots:solutions_w_editorials_py_decontaminated": 11706 |
| }, |
| "source_bytes": { |
| "HuggingFaceFW/fineweb-edu:sample-10BT": 5880657332, |
| "HuggingFaceTB/finemath:finemath-4plus": 4048334700, |
| "HuggingFaceTB/smollm-corpus:cosmopedia-v2": 1840112188, |
| "open-r1/OpenR1-Math-220k:default": 1288078700, |
| "open-r1/codeforces-cots:solutions_w_editorials_py_decontaminated": 552098223 |
| }, |
| "validation_loss": 3.78125 |
| } |
