Update tokenizer, configuration and training reports
Browse files
reports/run/metrics.jsonl
CHANGED
|
@@ -21,3 +21,5 @@
|
|
| 21 |
{"step": 420, "tokens": 220200960, "loss": 4.741484451293945, "tokens_per_second": 1007498.0805184614, "lr": 0.000315, "gradient_norm": 1.2149815559387207, "benchmark": false}
|
| 22 |
{"step": 440, "tokens": 230686720, "loss": 4.623250198364258, "tokens_per_second": 1011900.1103533852, "lr": 0.00033, "gradient_norm": 1.3369032144546509, "benchmark": false}
|
| 23 |
{"step": 460, "tokens": 241172480, "loss": 4.510102844238281, "tokens_per_second": 988569.6225406213, "lr": 0.00034500000000000004, "gradient_norm": 1.1452369689941406, "benchmark": false}
|
|
|
|
|
|
|
|
|
| 21 |
{"step": 420, "tokens": 220200960, "loss": 4.741484451293945, "tokens_per_second": 1007498.0805184614, "lr": 0.000315, "gradient_norm": 1.2149815559387207, "benchmark": false}
|
| 22 |
{"step": 440, "tokens": 230686720, "loss": 4.623250198364258, "tokens_per_second": 1011900.1103533852, "lr": 0.00033, "gradient_norm": 1.3369032144546509, "benchmark": false}
|
| 23 |
{"step": 460, "tokens": 241172480, "loss": 4.510102844238281, "tokens_per_second": 988569.6225406213, "lr": 0.00034500000000000004, "gradient_norm": 1.1452369689941406, "benchmark": false}
|
| 24 |
+
{"step": 480, "tokens": 251658240, "loss": 4.4276611328125, "tokens_per_second": 989866.3254340036, "lr": 0.00035999999999999997, "gradient_norm": 0.9811086058616638, "benchmark": false}
|
| 25 |
+
{"step": 500, "tokens": 262144000, "loss": 4.3114372253417965, "tokens_per_second": 987284.1551733142, "lr": 0.000375, "gradient_norm": 0.8948100805282593, "benchmark": false}
|