Upload training_state.json with huggingface_hub
Browse files- training_state.json +37 -0
training_state.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 100,
|
| 3 |
+
"tokens": 9830400,
|
| 4 |
+
"parameters": {
|
| 5 |
+
"total": 998694784,
|
| 6 |
+
"active": 340154240,
|
| 7 |
+
"shared": 10883968,
|
| 8 |
+
"active_path": 329270272
|
| 9 |
+
},
|
| 10 |
+
"batch_size": 96,
|
| 11 |
+
"sequence_length": 1024,
|
| 12 |
+
"loss": 3.15625,
|
| 13 |
+
"auxiliary_loss": 3.28125,
|
| 14 |
+
"gradient_norm": 2.203125,
|
| 15 |
+
"route": 0,
|
| 16 |
+
"lr_scale": 0.1,
|
| 17 |
+
"tokens_per_second": 250904.8699704037,
|
| 18 |
+
"elapsed_seconds": 36.73786387899963,
|
| 19 |
+
"average_patch_bytes": 9.319681456200227,
|
| 20 |
+
"patch_overflow_rate": 0.17708333333333334,
|
| 21 |
+
"gate": 0.462890625,
|
| 22 |
+
"peak_vram_gib": 31.73339605331421,
|
| 23 |
+
"documents": {
|
| 24 |
+
"HuggingFaceFW/fineweb-edu:sample-10BT": 921,
|
| 25 |
+
"HuggingFaceTB/finemath:finemath-4plus": 666,
|
| 26 |
+
"HuggingFaceTB/smollm-corpus:cosmopedia-v2": 526,
|
| 27 |
+
"open-r1/OpenR1-Math-220k:default": 398,
|
| 28 |
+
"open-r1/codeforces-cots:solutions_w_editorials_py_decontaminated": 8
|
| 29 |
+
},
|
| 30 |
+
"source_bytes": {
|
| 31 |
+
"HuggingFaceFW/fineweb-edu:sample-10BT": 4341513,
|
| 32 |
+
"HuggingFaceTB/finemath:finemath-4plus": 3189383,
|
| 33 |
+
"HuggingFaceTB/smollm-corpus:cosmopedia-v2": 1928354,
|
| 34 |
+
"open-r1/OpenR1-Math-220k:default": 523394,
|
| 35 |
+
"open-r1/codeforces-cots:solutions_w_editorials_py_decontaminated": 257377
|
| 36 |
+
}
|
| 37 |
+
}
|