Training in progress, step 4660
Browse files- all_results.json +18 -0
- train_results.json +18 -0
- trainer_state.json +3063 -0
all_results.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"before_init_mem_cpu": 933711872,
|
| 3 |
+
"before_init_mem_gpu": 0,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"init_mem_cpu_alloc_delta": 2791841792,
|
| 6 |
+
"init_mem_cpu_peaked_delta": 413900800,
|
| 7 |
+
"init_mem_gpu_alloc_delta": 497994240,
|
| 8 |
+
"init_mem_gpu_peaked_delta": 512,
|
| 9 |
+
"train_loss": 1.525319737733178,
|
| 10 |
+
"train_mem_cpu_alloc_delta": 8646197248,
|
| 11 |
+
"train_mem_cpu_peaked_delta": 402849792,
|
| 12 |
+
"train_mem_gpu_alloc_delta": 2010885120,
|
| 13 |
+
"train_mem_gpu_peaked_delta": 15468135424,
|
| 14 |
+
"train_runtime": 218631.5085,
|
| 15 |
+
"train_samples": 3825056,
|
| 16 |
+
"train_samples_per_second": 174.954,
|
| 17 |
+
"train_steps_per_second": 0.021
|
| 18 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"before_init_mem_cpu": 933711872,
|
| 3 |
+
"before_init_mem_gpu": 0,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"init_mem_cpu_alloc_delta": 2791841792,
|
| 6 |
+
"init_mem_cpu_peaked_delta": 413900800,
|
| 7 |
+
"init_mem_gpu_alloc_delta": 497994240,
|
| 8 |
+
"init_mem_gpu_peaked_delta": 512,
|
| 9 |
+
"train_loss": 1.525319737733178,
|
| 10 |
+
"train_mem_cpu_alloc_delta": 8646197248,
|
| 11 |
+
"train_mem_cpu_peaked_delta": 402849792,
|
| 12 |
+
"train_mem_gpu_alloc_delta": 2010885120,
|
| 13 |
+
"train_mem_gpu_peaked_delta": 15468135424,
|
| 14 |
+
"train_runtime": 218631.5085,
|
| 15 |
+
"train_samples": 3825056,
|
| 16 |
+
"train_samples_per_second": 174.954,
|
| 17 |
+
"train_steps_per_second": 0.021
|
| 18 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3063 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.997992236648374,
|
| 5 |
+
"global_step": 4660,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 5.987339055793991e-06,
|
| 13 |
+
"loss": 1.5226,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 5.974678111587983e-06,
|
| 19 |
+
"loss": 1.5251,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 5.9620171673819746e-06,
|
| 25 |
+
"loss": 1.5261,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.09,
|
| 30 |
+
"learning_rate": 5.949356223175966e-06,
|
| 31 |
+
"loss": 1.5257,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.11,
|
| 36 |
+
"learning_rate": 5.936695278969957e-06,
|
| 37 |
+
"loss": 1.5235,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"learning_rate": 5.924034334763948e-06,
|
| 43 |
+
"loss": 1.5214,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15,
|
| 48 |
+
"learning_rate": 5.91137339055794e-06,
|
| 49 |
+
"loss": 1.5245,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 5.898712446351932e-06,
|
| 55 |
+
"loss": 1.5259,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.19,
|
| 60 |
+
"learning_rate": 5.8860515021459225e-06,
|
| 61 |
+
"loss": 1.5223,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.21,
|
| 66 |
+
"learning_rate": 5.873390557939914e-06,
|
| 67 |
+
"loss": 1.522,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.24,
|
| 72 |
+
"learning_rate": 5.860729613733905e-06,
|
| 73 |
+
"loss": 1.5269,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.26,
|
| 78 |
+
"learning_rate": 5.848068669527898e-06,
|
| 79 |
+
"loss": 1.5226,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.28,
|
| 84 |
+
"learning_rate": 5.835407725321889e-06,
|
| 85 |
+
"loss": 1.5234,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3,
|
| 90 |
+
"learning_rate": 5.8227467811158795e-06,
|
| 91 |
+
"loss": 1.5268,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.32,
|
| 96 |
+
"learning_rate": 5.810085836909871e-06,
|
| 97 |
+
"loss": 1.5222,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.34,
|
| 102 |
+
"learning_rate": 5.797424892703863e-06,
|
| 103 |
+
"loss": 1.5245,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.34,
|
| 108 |
+
"eval_loss": 1.4186874628067017,
|
| 109 |
+
"eval_runtime": 78.1328,
|
| 110 |
+
"eval_samples_per_second": 511.949,
|
| 111 |
+
"eval_steps_per_second": 7.999,
|
| 112 |
+
"step": 160
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.36,
|
| 116 |
+
"learning_rate": 5.784763948497855e-06,
|
| 117 |
+
"loss": 1.5271,
|
| 118 |
+
"step": 170
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.39,
|
| 122 |
+
"learning_rate": 5.772103004291846e-06,
|
| 123 |
+
"loss": 1.5254,
|
| 124 |
+
"step": 180
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.41,
|
| 128 |
+
"learning_rate": 5.759442060085837e-06,
|
| 129 |
+
"loss": 1.5261,
|
| 130 |
+
"step": 190
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.43,
|
| 134 |
+
"learning_rate": 5.746781115879828e-06,
|
| 135 |
+
"loss": 1.5286,
|
| 136 |
+
"step": 200
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.45,
|
| 140 |
+
"learning_rate": 5.73412017167382e-06,
|
| 141 |
+
"loss": 1.5227,
|
| 142 |
+
"step": 210
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.47,
|
| 146 |
+
"learning_rate": 5.721459227467812e-06,
|
| 147 |
+
"loss": 1.5246,
|
| 148 |
+
"step": 220
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.49,
|
| 152 |
+
"learning_rate": 5.708798283261803e-06,
|
| 153 |
+
"loss": 1.5228,
|
| 154 |
+
"step": 230
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.51,
|
| 158 |
+
"learning_rate": 5.696137339055794e-06,
|
| 159 |
+
"loss": 1.5238,
|
| 160 |
+
"step": 240
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.54,
|
| 164 |
+
"learning_rate": 5.683476394849785e-06,
|
| 165 |
+
"loss": 1.523,
|
| 166 |
+
"step": 250
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.56,
|
| 170 |
+
"learning_rate": 5.670815450643777e-06,
|
| 171 |
+
"loss": 1.5238,
|
| 172 |
+
"step": 260
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.58,
|
| 176 |
+
"learning_rate": 5.658154506437769e-06,
|
| 177 |
+
"loss": 1.5252,
|
| 178 |
+
"step": 270
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.6,
|
| 182 |
+
"learning_rate": 5.64549356223176e-06,
|
| 183 |
+
"loss": 1.5234,
|
| 184 |
+
"step": 280
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.62,
|
| 188 |
+
"learning_rate": 5.632832618025751e-06,
|
| 189 |
+
"loss": 1.527,
|
| 190 |
+
"step": 290
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.64,
|
| 194 |
+
"learning_rate": 5.6201716738197424e-06,
|
| 195 |
+
"loss": 1.5253,
|
| 196 |
+
"step": 300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.66,
|
| 200 |
+
"learning_rate": 5.607510729613734e-06,
|
| 201 |
+
"loss": 1.5302,
|
| 202 |
+
"step": 310
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.69,
|
| 206 |
+
"learning_rate": 5.594849785407725e-06,
|
| 207 |
+
"loss": 1.5245,
|
| 208 |
+
"step": 320
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.69,
|
| 212 |
+
"eval_loss": 1.4183368682861328,
|
| 213 |
+
"eval_runtime": 78.1992,
|
| 214 |
+
"eval_samples_per_second": 511.514,
|
| 215 |
+
"eval_steps_per_second": 7.992,
|
| 216 |
+
"step": 320
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.71,
|
| 220 |
+
"learning_rate": 5.582188841201717e-06,
|
| 221 |
+
"loss": 1.5235,
|
| 222 |
+
"step": 330
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.73,
|
| 226 |
+
"learning_rate": 5.569527896995708e-06,
|
| 227 |
+
"loss": 1.5255,
|
| 228 |
+
"step": 340
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.75,
|
| 232 |
+
"learning_rate": 5.5568669527897e-06,
|
| 233 |
+
"loss": 1.5268,
|
| 234 |
+
"step": 350
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.77,
|
| 238 |
+
"learning_rate": 5.544206008583691e-06,
|
| 239 |
+
"loss": 1.5232,
|
| 240 |
+
"step": 360
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.79,
|
| 244 |
+
"learning_rate": 5.531545064377683e-06,
|
| 245 |
+
"loss": 1.5254,
|
| 246 |
+
"step": 370
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.81,
|
| 250 |
+
"learning_rate": 5.518884120171674e-06,
|
| 251 |
+
"loss": 1.5265,
|
| 252 |
+
"step": 380
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.84,
|
| 256 |
+
"learning_rate": 5.506223175965666e-06,
|
| 257 |
+
"loss": 1.5246,
|
| 258 |
+
"step": 390
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.86,
|
| 262 |
+
"learning_rate": 5.493562231759657e-06,
|
| 263 |
+
"loss": 1.5238,
|
| 264 |
+
"step": 400
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.88,
|
| 268 |
+
"learning_rate": 5.480901287553648e-06,
|
| 269 |
+
"loss": 1.5267,
|
| 270 |
+
"step": 410
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.9,
|
| 274 |
+
"learning_rate": 5.468240343347639e-06,
|
| 275 |
+
"loss": 1.5257,
|
| 276 |
+
"step": 420
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.92,
|
| 280 |
+
"learning_rate": 5.455579399141631e-06,
|
| 281 |
+
"loss": 1.5254,
|
| 282 |
+
"step": 430
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.94,
|
| 286 |
+
"learning_rate": 5.442918454935623e-06,
|
| 287 |
+
"loss": 1.5242,
|
| 288 |
+
"step": 440
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.96,
|
| 292 |
+
"learning_rate": 5.4302575107296144e-06,
|
| 293 |
+
"loss": 1.5268,
|
| 294 |
+
"step": 450
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.99,
|
| 298 |
+
"learning_rate": 5.417596566523605e-06,
|
| 299 |
+
"loss": 1.5252,
|
| 300 |
+
"step": 460
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.01,
|
| 304 |
+
"learning_rate": 5.404935622317596e-06,
|
| 305 |
+
"loss": 1.6674,
|
| 306 |
+
"step": 470
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.03,
|
| 310 |
+
"learning_rate": 5.392274678111588e-06,
|
| 311 |
+
"loss": 1.5259,
|
| 312 |
+
"step": 480
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.03,
|
| 316 |
+
"eval_loss": 1.4177254438400269,
|
| 317 |
+
"eval_runtime": 78.2595,
|
| 318 |
+
"eval_samples_per_second": 511.12,
|
| 319 |
+
"eval_steps_per_second": 7.986,
|
| 320 |
+
"step": 480
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.05,
|
| 324 |
+
"learning_rate": 5.37961373390558e-06,
|
| 325 |
+
"loss": 1.5217,
|
| 326 |
+
"step": 490
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.07,
|
| 330 |
+
"learning_rate": 5.3669527896995715e-06,
|
| 331 |
+
"loss": 1.5261,
|
| 332 |
+
"step": 500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.09,
|
| 336 |
+
"learning_rate": 5.354291845493562e-06,
|
| 337 |
+
"loss": 1.5241,
|
| 338 |
+
"step": 510
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.12,
|
| 342 |
+
"learning_rate": 5.341630901287553e-06,
|
| 343 |
+
"loss": 1.5241,
|
| 344 |
+
"step": 520
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.14,
|
| 348 |
+
"learning_rate": 5.328969957081545e-06,
|
| 349 |
+
"loss": 1.5233,
|
| 350 |
+
"step": 530
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.16,
|
| 354 |
+
"learning_rate": 5.316309012875537e-06,
|
| 355 |
+
"loss": 1.5245,
|
| 356 |
+
"step": 540
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.18,
|
| 360 |
+
"learning_rate": 5.303648068669528e-06,
|
| 361 |
+
"loss": 1.5237,
|
| 362 |
+
"step": 550
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.2,
|
| 366 |
+
"learning_rate": 5.2909871244635194e-06,
|
| 367 |
+
"loss": 1.5233,
|
| 368 |
+
"step": 560
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.22,
|
| 372 |
+
"learning_rate": 5.27832618025751e-06,
|
| 373 |
+
"loss": 1.5218,
|
| 374 |
+
"step": 570
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.24,
|
| 378 |
+
"learning_rate": 5.265665236051502e-06,
|
| 379 |
+
"loss": 1.5244,
|
| 380 |
+
"step": 580
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.27,
|
| 384 |
+
"learning_rate": 5.253004291845494e-06,
|
| 385 |
+
"loss": 1.5228,
|
| 386 |
+
"step": 590
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.29,
|
| 390 |
+
"learning_rate": 5.2403433476394856e-06,
|
| 391 |
+
"loss": 1.5268,
|
| 392 |
+
"step": 600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.31,
|
| 396 |
+
"learning_rate": 5.2276824034334765e-06,
|
| 397 |
+
"loss": 1.5246,
|
| 398 |
+
"step": 610
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.33,
|
| 402 |
+
"learning_rate": 5.215021459227467e-06,
|
| 403 |
+
"loss": 1.5205,
|
| 404 |
+
"step": 620
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.35,
|
| 408 |
+
"learning_rate": 5.20236051502146e-06,
|
| 409 |
+
"loss": 1.5252,
|
| 410 |
+
"step": 630
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.37,
|
| 414 |
+
"learning_rate": 5.189699570815451e-06,
|
| 415 |
+
"loss": 1.5265,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.37,
|
| 420 |
+
"eval_loss": 1.4185311794281006,
|
| 421 |
+
"eval_runtime": 78.202,
|
| 422 |
+
"eval_samples_per_second": 511.496,
|
| 423 |
+
"eval_steps_per_second": 7.992,
|
| 424 |
+
"step": 640
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.39,
|
| 428 |
+
"learning_rate": 5.177038626609442e-06,
|
| 429 |
+
"loss": 1.5229,
|
| 430 |
+
"step": 650
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.42,
|
| 434 |
+
"learning_rate": 5.1643776824034335e-06,
|
| 435 |
+
"loss": 1.5246,
|
| 436 |
+
"step": 660
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.44,
|
| 440 |
+
"learning_rate": 5.151716738197425e-06,
|
| 441 |
+
"loss": 1.5208,
|
| 442 |
+
"step": 670
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 1.46,
|
| 446 |
+
"learning_rate": 5.139055793991416e-06,
|
| 447 |
+
"loss": 1.523,
|
| 448 |
+
"step": 680
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.48,
|
| 452 |
+
"learning_rate": 5.126394849785408e-06,
|
| 453 |
+
"loss": 1.5222,
|
| 454 |
+
"step": 690
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.5,
|
| 458 |
+
"learning_rate": 5.113733905579399e-06,
|
| 459 |
+
"loss": 1.5223,
|
| 460 |
+
"step": 700
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.52,
|
| 464 |
+
"learning_rate": 5.1010729613733906e-06,
|
| 465 |
+
"loss": 1.5206,
|
| 466 |
+
"step": 710
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.54,
|
| 470 |
+
"learning_rate": 5.088412017167382e-06,
|
| 471 |
+
"loss": 1.5194,
|
| 472 |
+
"step": 720
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.57,
|
| 476 |
+
"learning_rate": 5.075751072961374e-06,
|
| 477 |
+
"loss": 1.5249,
|
| 478 |
+
"step": 730
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.59,
|
| 482 |
+
"learning_rate": 5.063090128755365e-06,
|
| 483 |
+
"loss": 1.5254,
|
| 484 |
+
"step": 740
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 1.61,
|
| 488 |
+
"learning_rate": 5.050429184549356e-06,
|
| 489 |
+
"loss": 1.521,
|
| 490 |
+
"step": 750
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.63,
|
| 494 |
+
"learning_rate": 5.037768240343348e-06,
|
| 495 |
+
"loss": 1.5255,
|
| 496 |
+
"step": 760
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.65,
|
| 500 |
+
"learning_rate": 5.025107296137339e-06,
|
| 501 |
+
"loss": 1.5238,
|
| 502 |
+
"step": 770
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.67,
|
| 506 |
+
"learning_rate": 5.01244635193133e-06,
|
| 507 |
+
"loss": 1.5261,
|
| 508 |
+
"step": 780
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.69,
|
| 512 |
+
"learning_rate": 4.999785407725322e-06,
|
| 513 |
+
"loss": 1.5233,
|
| 514 |
+
"step": 790
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.72,
|
| 518 |
+
"learning_rate": 4.987124463519313e-06,
|
| 519 |
+
"loss": 1.5245,
|
| 520 |
+
"step": 800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.72,
|
| 524 |
+
"eval_loss": 1.4190433025360107,
|
| 525 |
+
"eval_runtime": 78.1942,
|
| 526 |
+
"eval_samples_per_second": 511.547,
|
| 527 |
+
"eval_steps_per_second": 7.993,
|
| 528 |
+
"step": 800
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.74,
|
| 532 |
+
"learning_rate": 4.974463519313305e-06,
|
| 533 |
+
"loss": 1.5237,
|
| 534 |
+
"step": 810
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.76,
|
| 538 |
+
"learning_rate": 4.961802575107296e-06,
|
| 539 |
+
"loss": 1.5216,
|
| 540 |
+
"step": 820
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.78,
|
| 544 |
+
"learning_rate": 4.949141630901288e-06,
|
| 545 |
+
"loss": 1.5256,
|
| 546 |
+
"step": 830
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 1.8,
|
| 550 |
+
"learning_rate": 4.936480686695279e-06,
|
| 551 |
+
"loss": 1.5249,
|
| 552 |
+
"step": 840
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 1.82,
|
| 556 |
+
"learning_rate": 4.92381974248927e-06,
|
| 557 |
+
"loss": 1.5235,
|
| 558 |
+
"step": 850
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 1.84,
|
| 562 |
+
"learning_rate": 4.9111587982832626e-06,
|
| 563 |
+
"loss": 1.5255,
|
| 564 |
+
"step": 860
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.87,
|
| 568 |
+
"learning_rate": 4.8984978540772535e-06,
|
| 569 |
+
"loss": 1.5242,
|
| 570 |
+
"step": 870
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.89,
|
| 574 |
+
"learning_rate": 4.885836909871244e-06,
|
| 575 |
+
"loss": 1.5243,
|
| 576 |
+
"step": 880
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.91,
|
| 580 |
+
"learning_rate": 4.873175965665236e-06,
|
| 581 |
+
"loss": 1.5269,
|
| 582 |
+
"step": 890
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 1.93,
|
| 586 |
+
"learning_rate": 4.860515021459228e-06,
|
| 587 |
+
"loss": 1.5216,
|
| 588 |
+
"step": 900
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 1.95,
|
| 592 |
+
"learning_rate": 4.847854077253219e-06,
|
| 593 |
+
"loss": 1.5212,
|
| 594 |
+
"step": 910
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.97,
|
| 598 |
+
"learning_rate": 4.8351931330472105e-06,
|
| 599 |
+
"loss": 1.5254,
|
| 600 |
+
"step": 920
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.99,
|
| 604 |
+
"learning_rate": 4.822532188841201e-06,
|
| 605 |
+
"loss": 1.5241,
|
| 606 |
+
"step": 930
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.02,
|
| 610 |
+
"learning_rate": 4.809871244635193e-06,
|
| 611 |
+
"loss": 1.6665,
|
| 612 |
+
"step": 940
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 2.04,
|
| 616 |
+
"learning_rate": 4.797210300429185e-06,
|
| 617 |
+
"loss": 1.5227,
|
| 618 |
+
"step": 950
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.06,
|
| 622 |
+
"learning_rate": 4.784549356223177e-06,
|
| 623 |
+
"loss": 1.5241,
|
| 624 |
+
"step": 960
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.06,
|
| 628 |
+
"eval_loss": 1.4172289371490479,
|
| 629 |
+
"eval_runtime": 78.2198,
|
| 630 |
+
"eval_samples_per_second": 511.379,
|
| 631 |
+
"eval_steps_per_second": 7.99,
|
| 632 |
+
"step": 960
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.08,
|
| 636 |
+
"learning_rate": 4.7718884120171676e-06,
|
| 637 |
+
"loss": 1.5248,
|
| 638 |
+
"step": 970
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.1,
|
| 642 |
+
"learning_rate": 4.7592274678111585e-06,
|
| 643 |
+
"loss": 1.5247,
|
| 644 |
+
"step": 980
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.12,
|
| 648 |
+
"learning_rate": 4.74656652360515e-06,
|
| 649 |
+
"loss": 1.5232,
|
| 650 |
+
"step": 990
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.15,
|
| 654 |
+
"learning_rate": 4.733905579399142e-06,
|
| 655 |
+
"loss": 1.5238,
|
| 656 |
+
"step": 1000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.17,
|
| 660 |
+
"learning_rate": 4.721244635193133e-06,
|
| 661 |
+
"loss": 1.5234,
|
| 662 |
+
"step": 1010
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.19,
|
| 666 |
+
"learning_rate": 4.708583690987125e-06,
|
| 667 |
+
"loss": 1.5218,
|
| 668 |
+
"step": 1020
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.21,
|
| 672 |
+
"learning_rate": 4.6959227467811155e-06,
|
| 673 |
+
"loss": 1.5213,
|
| 674 |
+
"step": 1030
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.23,
|
| 678 |
+
"learning_rate": 4.683261802575107e-06,
|
| 679 |
+
"loss": 1.5262,
|
| 680 |
+
"step": 1040
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.25,
|
| 684 |
+
"learning_rate": 4.670600858369099e-06,
|
| 685 |
+
"loss": 1.5232,
|
| 686 |
+
"step": 1050
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 2.27,
|
| 690 |
+
"learning_rate": 4.657939914163091e-06,
|
| 691 |
+
"loss": 1.5228,
|
| 692 |
+
"step": 1060
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.3,
|
| 696 |
+
"learning_rate": 4.645278969957082e-06,
|
| 697 |
+
"loss": 1.5241,
|
| 698 |
+
"step": 1070
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.32,
|
| 702 |
+
"learning_rate": 4.6326180257510726e-06,
|
| 703 |
+
"loss": 1.5196,
|
| 704 |
+
"step": 1080
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.34,
|
| 708 |
+
"learning_rate": 4.619957081545065e-06,
|
| 709 |
+
"loss": 1.5229,
|
| 710 |
+
"step": 1090
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.36,
|
| 714 |
+
"learning_rate": 4.607296137339056e-06,
|
| 715 |
+
"loss": 1.5212,
|
| 716 |
+
"step": 1100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.38,
|
| 720 |
+
"learning_rate": 4.594635193133047e-06,
|
| 721 |
+
"loss": 1.5233,
|
| 722 |
+
"step": 1110
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 2.4,
|
| 726 |
+
"learning_rate": 4.581974248927039e-06,
|
| 727 |
+
"loss": 1.5227,
|
| 728 |
+
"step": 1120
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 2.4,
|
| 732 |
+
"eval_loss": 1.4165441989898682,
|
| 733 |
+
"eval_runtime": 78.1888,
|
| 734 |
+
"eval_samples_per_second": 511.582,
|
| 735 |
+
"eval_steps_per_second": 7.993,
|
| 736 |
+
"step": 1120
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.42,
|
| 740 |
+
"learning_rate": 4.5693133047210304e-06,
|
| 741 |
+
"loss": 1.521,
|
| 742 |
+
"step": 1130
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 2.45,
|
| 746 |
+
"learning_rate": 4.556652360515021e-06,
|
| 747 |
+
"loss": 1.5211,
|
| 748 |
+
"step": 1140
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.47,
|
| 752 |
+
"learning_rate": 4.543991416309013e-06,
|
| 753 |
+
"loss": 1.5261,
|
| 754 |
+
"step": 1150
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 2.49,
|
| 758 |
+
"learning_rate": 4.531330472103004e-06,
|
| 759 |
+
"loss": 1.5206,
|
| 760 |
+
"step": 1160
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.51,
|
| 764 |
+
"learning_rate": 4.518669527896996e-06,
|
| 765 |
+
"loss": 1.5235,
|
| 766 |
+
"step": 1170
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.53,
|
| 770 |
+
"learning_rate": 4.5060085836909875e-06,
|
| 771 |
+
"loss": 1.5238,
|
| 772 |
+
"step": 1180
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.55,
|
| 776 |
+
"learning_rate": 4.493347639484979e-06,
|
| 777 |
+
"loss": 1.5195,
|
| 778 |
+
"step": 1190
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 2.57,
|
| 782 |
+
"learning_rate": 4.48068669527897e-06,
|
| 783 |
+
"loss": 1.523,
|
| 784 |
+
"step": 1200
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 2.6,
|
| 788 |
+
"learning_rate": 4.468025751072961e-06,
|
| 789 |
+
"loss": 1.5201,
|
| 790 |
+
"step": 1210
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.62,
|
| 794 |
+
"learning_rate": 4.455364806866953e-06,
|
| 795 |
+
"loss": 1.5235,
|
| 796 |
+
"step": 1220
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.64,
|
| 800 |
+
"learning_rate": 4.4427038626609445e-06,
|
| 801 |
+
"loss": 1.5208,
|
| 802 |
+
"step": 1230
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.66,
|
| 806 |
+
"learning_rate": 4.4300429184549354e-06,
|
| 807 |
+
"loss": 1.5261,
|
| 808 |
+
"step": 1240
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.68,
|
| 812 |
+
"learning_rate": 4.417381974248927e-06,
|
| 813 |
+
"loss": 1.5231,
|
| 814 |
+
"step": 1250
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.7,
|
| 818 |
+
"learning_rate": 4.404721030042918e-06,
|
| 819 |
+
"loss": 1.5216,
|
| 820 |
+
"step": 1260
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 2.72,
|
| 824 |
+
"learning_rate": 4.39206008583691e-06,
|
| 825 |
+
"loss": 1.5246,
|
| 826 |
+
"step": 1270
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.75,
|
| 830 |
+
"learning_rate": 4.379399141630902e-06,
|
| 831 |
+
"loss": 1.5226,
|
| 832 |
+
"step": 1280
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.75,
|
| 836 |
+
"eval_loss": 1.415173053741455,
|
| 837 |
+
"eval_runtime": 78.2044,
|
| 838 |
+
"eval_samples_per_second": 511.48,
|
| 839 |
+
"eval_steps_per_second": 7.992,
|
| 840 |
+
"step": 1280
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.77,
|
| 844 |
+
"learning_rate": 4.366738197424893e-06,
|
| 845 |
+
"loss": 1.5221,
|
| 846 |
+
"step": 1290
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.79,
|
| 850 |
+
"learning_rate": 4.354077253218884e-06,
|
| 851 |
+
"loss": 1.5248,
|
| 852 |
+
"step": 1300
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.81,
|
| 856 |
+
"learning_rate": 4.341416309012875e-06,
|
| 857 |
+
"loss": 1.5227,
|
| 858 |
+
"step": 1310
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.83,
|
| 862 |
+
"learning_rate": 4.328755364806868e-06,
|
| 863 |
+
"loss": 1.5244,
|
| 864 |
+
"step": 1320
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.85,
|
| 868 |
+
"learning_rate": 4.316094420600859e-06,
|
| 869 |
+
"loss": 1.5258,
|
| 870 |
+
"step": 1330
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.87,
|
| 874 |
+
"learning_rate": 4.3034334763948495e-06,
|
| 875 |
+
"loss": 1.5237,
|
| 876 |
+
"step": 1340
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 2.9,
|
| 880 |
+
"learning_rate": 4.290772532188841e-06,
|
| 881 |
+
"loss": 1.5248,
|
| 882 |
+
"step": 1350
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.92,
|
| 886 |
+
"learning_rate": 4.278111587982833e-06,
|
| 887 |
+
"loss": 1.524,
|
| 888 |
+
"step": 1360
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.94,
|
| 892 |
+
"learning_rate": 4.265450643776824e-06,
|
| 893 |
+
"loss": 1.523,
|
| 894 |
+
"step": 1370
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.96,
|
| 898 |
+
"learning_rate": 4.252789699570816e-06,
|
| 899 |
+
"loss": 1.5242,
|
| 900 |
+
"step": 1380
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.98,
|
| 904 |
+
"learning_rate": 4.2401287553648074e-06,
|
| 905 |
+
"loss": 1.5255,
|
| 906 |
+
"step": 1390
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 3.0,
|
| 910 |
+
"learning_rate": 4.2274678111587975e-06,
|
| 911 |
+
"loss": 1.6659,
|
| 912 |
+
"step": 1400
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.03,
|
| 916 |
+
"learning_rate": 4.21480686695279e-06,
|
| 917 |
+
"loss": 1.5233,
|
| 918 |
+
"step": 1410
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.05,
|
| 922 |
+
"learning_rate": 4.202145922746782e-06,
|
| 923 |
+
"loss": 1.5249,
|
| 924 |
+
"step": 1420
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 3.07,
|
| 928 |
+
"learning_rate": 4.189484978540773e-06,
|
| 929 |
+
"loss": 1.5248,
|
| 930 |
+
"step": 1430
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 3.09,
|
| 934 |
+
"learning_rate": 4.176824034334764e-06,
|
| 935 |
+
"loss": 1.522,
|
| 936 |
+
"step": 1440
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.09,
|
| 940 |
+
"eval_loss": 1.418953537940979,
|
| 941 |
+
"eval_runtime": 78.1718,
|
| 942 |
+
"eval_samples_per_second": 511.693,
|
| 943 |
+
"eval_steps_per_second": 7.995,
|
| 944 |
+
"step": 1440
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 3.11,
|
| 948 |
+
"learning_rate": 4.164163090128755e-06,
|
| 949 |
+
"loss": 1.5241,
|
| 950 |
+
"step": 1450
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 3.13,
|
| 954 |
+
"learning_rate": 4.151502145922747e-06,
|
| 955 |
+
"loss": 1.5226,
|
| 956 |
+
"step": 1460
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 3.15,
|
| 960 |
+
"learning_rate": 4.138841201716738e-06,
|
| 961 |
+
"loss": 1.5231,
|
| 962 |
+
"step": 1470
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 3.18,
|
| 966 |
+
"learning_rate": 4.12618025751073e-06,
|
| 967 |
+
"loss": 1.5239,
|
| 968 |
+
"step": 1480
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.2,
|
| 972 |
+
"learning_rate": 4.113519313304721e-06,
|
| 973 |
+
"loss": 1.5217,
|
| 974 |
+
"step": 1490
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.22,
|
| 978 |
+
"learning_rate": 4.1008583690987124e-06,
|
| 979 |
+
"loss": 1.5206,
|
| 980 |
+
"step": 1500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 3.24,
|
| 984 |
+
"learning_rate": 4.088197424892704e-06,
|
| 985 |
+
"loss": 1.5232,
|
| 986 |
+
"step": 1510
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 3.26,
|
| 990 |
+
"learning_rate": 4.075536480686696e-06,
|
| 991 |
+
"loss": 1.5252,
|
| 992 |
+
"step": 1520
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.28,
|
| 996 |
+
"learning_rate": 4.062875536480687e-06,
|
| 997 |
+
"loss": 1.521,
|
| 998 |
+
"step": 1530
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 3.3,
|
| 1002 |
+
"learning_rate": 4.050214592274678e-06,
|
| 1003 |
+
"loss": 1.5205,
|
| 1004 |
+
"step": 1540
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 3.33,
|
| 1008 |
+
"learning_rate": 4.0375536480686695e-06,
|
| 1009 |
+
"loss": 1.5215,
|
| 1010 |
+
"step": 1550
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.35,
|
| 1014 |
+
"learning_rate": 4.024892703862661e-06,
|
| 1015 |
+
"loss": 1.5253,
|
| 1016 |
+
"step": 1560
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.37,
|
| 1020 |
+
"learning_rate": 4.012231759656652e-06,
|
| 1021 |
+
"loss": 1.5225,
|
| 1022 |
+
"step": 1570
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 3.39,
|
| 1026 |
+
"learning_rate": 3.999570815450644e-06,
|
| 1027 |
+
"loss": 1.523,
|
| 1028 |
+
"step": 1580
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 3.41,
|
| 1032 |
+
"learning_rate": 3.986909871244635e-06,
|
| 1033 |
+
"loss": 1.5241,
|
| 1034 |
+
"step": 1590
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.43,
|
| 1038 |
+
"learning_rate": 3.9742489270386265e-06,
|
| 1039 |
+
"loss": 1.5243,
|
| 1040 |
+
"step": 1600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 3.43,
|
| 1044 |
+
"eval_loss": 1.4176822900772095,
|
| 1045 |
+
"eval_runtime": 78.1608,
|
| 1046 |
+
"eval_samples_per_second": 511.766,
|
| 1047 |
+
"eval_steps_per_second": 7.996,
|
| 1048 |
+
"step": 1600
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 3.45,
|
| 1052 |
+
"learning_rate": 3.961587982832618e-06,
|
| 1053 |
+
"loss": 1.5235,
|
| 1054 |
+
"step": 1610
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 3.48,
|
| 1058 |
+
"learning_rate": 3.94892703862661e-06,
|
| 1059 |
+
"loss": 1.5224,
|
| 1060 |
+
"step": 1620
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 3.5,
|
| 1064 |
+
"learning_rate": 3.9362660944206e-06,
|
| 1065 |
+
"loss": 1.5192,
|
| 1066 |
+
"step": 1630
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.52,
|
| 1070 |
+
"learning_rate": 3.923605150214593e-06,
|
| 1071 |
+
"loss": 1.5186,
|
| 1072 |
+
"step": 1640
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 3.54,
|
| 1076 |
+
"learning_rate": 3.9109442060085844e-06,
|
| 1077 |
+
"loss": 1.5208,
|
| 1078 |
+
"step": 1650
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 3.56,
|
| 1082 |
+
"learning_rate": 3.898283261802575e-06,
|
| 1083 |
+
"loss": 1.5204,
|
| 1084 |
+
"step": 1660
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 3.58,
|
| 1088 |
+
"learning_rate": 3.885622317596566e-06,
|
| 1089 |
+
"loss": 1.5247,
|
| 1090 |
+
"step": 1670
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.6,
|
| 1094 |
+
"learning_rate": 3.872961373390559e-06,
|
| 1095 |
+
"loss": 1.5205,
|
| 1096 |
+
"step": 1680
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 3.63,
|
| 1100 |
+
"learning_rate": 3.86030042918455e-06,
|
| 1101 |
+
"loss": 1.5226,
|
| 1102 |
+
"step": 1690
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 3.65,
|
| 1106 |
+
"learning_rate": 3.847639484978541e-06,
|
| 1107 |
+
"loss": 1.5223,
|
| 1108 |
+
"step": 1700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.67,
|
| 1112 |
+
"learning_rate": 3.834978540772532e-06,
|
| 1113 |
+
"loss": 1.526,
|
| 1114 |
+
"step": 1710
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 3.69,
|
| 1118 |
+
"learning_rate": 3.822317596566524e-06,
|
| 1119 |
+
"loss": 1.523,
|
| 1120 |
+
"step": 1720
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 3.71,
|
| 1124 |
+
"learning_rate": 3.809656652360514e-06,
|
| 1125 |
+
"loss": 1.5209,
|
| 1126 |
+
"step": 1730
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 3.73,
|
| 1130 |
+
"learning_rate": 3.7969957081545068e-06,
|
| 1131 |
+
"loss": 1.5244,
|
| 1132 |
+
"step": 1740
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 3.75,
|
| 1136 |
+
"learning_rate": 3.784334763948498e-06,
|
| 1137 |
+
"loss": 1.5225,
|
| 1138 |
+
"step": 1750
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 3.78,
|
| 1142 |
+
"learning_rate": 3.771673819742489e-06,
|
| 1143 |
+
"loss": 1.5213,
|
| 1144 |
+
"step": 1760
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 3.78,
|
| 1148 |
+
"eval_loss": 1.4134083986282349,
|
| 1149 |
+
"eval_runtime": 78.1868,
|
| 1150 |
+
"eval_samples_per_second": 511.596,
|
| 1151 |
+
"eval_steps_per_second": 7.994,
|
| 1152 |
+
"step": 1760
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 3.8,
|
| 1156 |
+
"learning_rate": 3.7590128755364807e-06,
|
| 1157 |
+
"loss": 1.5239,
|
| 1158 |
+
"step": 1770
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 3.82,
|
| 1162 |
+
"learning_rate": 3.7463519313304725e-06,
|
| 1163 |
+
"loss": 1.5227,
|
| 1164 |
+
"step": 1780
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.84,
|
| 1168 |
+
"learning_rate": 3.733690987124464e-06,
|
| 1169 |
+
"loss": 1.5246,
|
| 1170 |
+
"step": 1790
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 3.86,
|
| 1174 |
+
"learning_rate": 3.721030042918455e-06,
|
| 1175 |
+
"loss": 1.5235,
|
| 1176 |
+
"step": 1800
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 3.88,
|
| 1180 |
+
"learning_rate": 3.7083690987124465e-06,
|
| 1181 |
+
"loss": 1.5228,
|
| 1182 |
+
"step": 1810
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 3.9,
|
| 1186 |
+
"learning_rate": 3.6957081545064374e-06,
|
| 1187 |
+
"loss": 1.5227,
|
| 1188 |
+
"step": 1820
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 3.93,
|
| 1192 |
+
"learning_rate": 3.683047210300429e-06,
|
| 1193 |
+
"loss": 1.5236,
|
| 1194 |
+
"step": 1830
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 3.95,
|
| 1198 |
+
"learning_rate": 3.670386266094421e-06,
|
| 1199 |
+
"loss": 1.5254,
|
| 1200 |
+
"step": 1840
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 3.97,
|
| 1204 |
+
"learning_rate": 3.657725321888412e-06,
|
| 1205 |
+
"loss": 1.5235,
|
| 1206 |
+
"step": 1850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.99,
|
| 1210 |
+
"learning_rate": 3.645064377682403e-06,
|
| 1211 |
+
"loss": 1.5245,
|
| 1212 |
+
"step": 1860
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 4.01,
|
| 1216 |
+
"learning_rate": 3.632403433476395e-06,
|
| 1217 |
+
"loss": 1.6644,
|
| 1218 |
+
"step": 1870
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 4.03,
|
| 1222 |
+
"learning_rate": 3.6197424892703866e-06,
|
| 1223 |
+
"loss": 1.5232,
|
| 1224 |
+
"step": 1880
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.06,
|
| 1228 |
+
"learning_rate": 3.6070815450643783e-06,
|
| 1229 |
+
"loss": 1.5218,
|
| 1230 |
+
"step": 1890
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 4.08,
|
| 1234 |
+
"learning_rate": 3.5944206008583692e-06,
|
| 1235 |
+
"loss": 1.5209,
|
| 1236 |
+
"step": 1900
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 4.1,
|
| 1240 |
+
"learning_rate": 3.5817596566523606e-06,
|
| 1241 |
+
"loss": 1.5217,
|
| 1242 |
+
"step": 1910
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 4.12,
|
| 1246 |
+
"learning_rate": 3.5690987124463515e-06,
|
| 1247 |
+
"loss": 1.524,
|
| 1248 |
+
"step": 1920
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.12,
|
| 1252 |
+
"eval_loss": 1.4139704704284668,
|
| 1253 |
+
"eval_runtime": 78.1302,
|
| 1254 |
+
"eval_samples_per_second": 511.966,
|
| 1255 |
+
"eval_steps_per_second": 7.999,
|
| 1256 |
+
"step": 1920
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.14,
|
| 1260 |
+
"learning_rate": 3.5564377682403432e-06,
|
| 1261 |
+
"loss": 1.5244,
|
| 1262 |
+
"step": 1930
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.16,
|
| 1266 |
+
"learning_rate": 3.543776824034335e-06,
|
| 1267 |
+
"loss": 1.5248,
|
| 1268 |
+
"step": 1940
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 4.18,
|
| 1272 |
+
"learning_rate": 3.5311158798283267e-06,
|
| 1273 |
+
"loss": 1.5207,
|
| 1274 |
+
"step": 1950
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 4.21,
|
| 1278 |
+
"learning_rate": 3.5184549356223176e-06,
|
| 1279 |
+
"loss": 1.52,
|
| 1280 |
+
"step": 1960
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.23,
|
| 1284 |
+
"learning_rate": 3.505793991416309e-06,
|
| 1285 |
+
"loss": 1.5208,
|
| 1286 |
+
"step": 1970
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 4.25,
|
| 1290 |
+
"learning_rate": 3.4931330472103015e-06,
|
| 1291 |
+
"loss": 1.525,
|
| 1292 |
+
"step": 1980
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 4.27,
|
| 1296 |
+
"learning_rate": 3.4804721030042916e-06,
|
| 1297 |
+
"loss": 1.5242,
|
| 1298 |
+
"step": 1990
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 4.29,
|
| 1302 |
+
"learning_rate": 3.4678111587982833e-06,
|
| 1303 |
+
"loss": 1.5228,
|
| 1304 |
+
"step": 2000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.31,
|
| 1308 |
+
"learning_rate": 3.4551502145922747e-06,
|
| 1309 |
+
"loss": 1.5219,
|
| 1310 |
+
"step": 2010
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 4.33,
|
| 1314 |
+
"learning_rate": 3.4424892703862664e-06,
|
| 1315 |
+
"loss": 1.5212,
|
| 1316 |
+
"step": 2020
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 4.36,
|
| 1320 |
+
"learning_rate": 3.4298283261802573e-06,
|
| 1321 |
+
"loss": 1.523,
|
| 1322 |
+
"step": 2030
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.38,
|
| 1326 |
+
"learning_rate": 3.417167381974249e-06,
|
| 1327 |
+
"loss": 1.5245,
|
| 1328 |
+
"step": 2040
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 4.4,
|
| 1332 |
+
"learning_rate": 3.404506437768241e-06,
|
| 1333 |
+
"loss": 1.5223,
|
| 1334 |
+
"step": 2050
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 4.42,
|
| 1338 |
+
"learning_rate": 3.3918454935622317e-06,
|
| 1339 |
+
"loss": 1.5259,
|
| 1340 |
+
"step": 2060
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 4.44,
|
| 1344 |
+
"learning_rate": 3.379184549356223e-06,
|
| 1345 |
+
"loss": 1.5214,
|
| 1346 |
+
"step": 2070
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.46,
|
| 1350 |
+
"learning_rate": 3.3665236051502144e-06,
|
| 1351 |
+
"loss": 1.5223,
|
| 1352 |
+
"step": 2080
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 4.46,
|
| 1356 |
+
"eval_loss": 1.4173228740692139,
|
| 1357 |
+
"eval_runtime": 78.2047,
|
| 1358 |
+
"eval_samples_per_second": 511.478,
|
| 1359 |
+
"eval_steps_per_second": 7.992,
|
| 1360 |
+
"step": 2080
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.48,
|
| 1364 |
+
"learning_rate": 3.3538626609442057e-06,
|
| 1365 |
+
"loss": 1.5204,
|
| 1366 |
+
"step": 2090
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 4.51,
|
| 1370 |
+
"learning_rate": 3.3412017167381974e-06,
|
| 1371 |
+
"loss": 1.5211,
|
| 1372 |
+
"step": 2100
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 4.53,
|
| 1376 |
+
"learning_rate": 3.328540772532189e-06,
|
| 1377 |
+
"loss": 1.5202,
|
| 1378 |
+
"step": 2110
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 4.55,
|
| 1382 |
+
"learning_rate": 3.3158798283261805e-06,
|
| 1383 |
+
"loss": 1.5183,
|
| 1384 |
+
"step": 2120
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 4.57,
|
| 1388 |
+
"learning_rate": 3.3032188841201714e-06,
|
| 1389 |
+
"loss": 1.524,
|
| 1390 |
+
"step": 2130
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 4.59,
|
| 1394 |
+
"learning_rate": 3.290557939914164e-06,
|
| 1395 |
+
"loss": 1.5208,
|
| 1396 |
+
"step": 2140
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.61,
|
| 1400 |
+
"learning_rate": 3.277896995708154e-06,
|
| 1401 |
+
"loss": 1.5224,
|
| 1402 |
+
"step": 2150
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 4.63,
|
| 1406 |
+
"learning_rate": 3.265236051502146e-06,
|
| 1407 |
+
"loss": 1.5193,
|
| 1408 |
+
"step": 2160
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 4.66,
|
| 1412 |
+
"learning_rate": 3.2525751072961376e-06,
|
| 1413 |
+
"loss": 1.5231,
|
| 1414 |
+
"step": 2170
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 4.68,
|
| 1418 |
+
"learning_rate": 3.239914163090129e-06,
|
| 1419 |
+
"loss": 1.5252,
|
| 1420 |
+
"step": 2180
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 4.7,
|
| 1424 |
+
"learning_rate": 3.2272532188841198e-06,
|
| 1425 |
+
"loss": 1.5216,
|
| 1426 |
+
"step": 2190
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 4.72,
|
| 1430 |
+
"learning_rate": 3.2145922746781115e-06,
|
| 1431 |
+
"loss": 1.5217,
|
| 1432 |
+
"step": 2200
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 4.74,
|
| 1436 |
+
"learning_rate": 3.2019313304721033e-06,
|
| 1437 |
+
"loss": 1.5235,
|
| 1438 |
+
"step": 2210
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 4.76,
|
| 1442 |
+
"learning_rate": 3.189270386266094e-06,
|
| 1443 |
+
"loss": 1.5205,
|
| 1444 |
+
"step": 2220
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 4.78,
|
| 1448 |
+
"learning_rate": 3.176609442060086e-06,
|
| 1449 |
+
"loss": 1.524,
|
| 1450 |
+
"step": 2230
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 4.81,
|
| 1454 |
+
"learning_rate": 3.1639484978540773e-06,
|
| 1455 |
+
"loss": 1.5236,
|
| 1456 |
+
"step": 2240
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 4.81,
|
| 1460 |
+
"eval_loss": 1.4121308326721191,
|
| 1461 |
+
"eval_runtime": 78.2109,
|
| 1462 |
+
"eval_samples_per_second": 511.438,
|
| 1463 |
+
"eval_steps_per_second": 7.991,
|
| 1464 |
+
"step": 2240
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 4.83,
|
| 1468 |
+
"learning_rate": 3.151287553648068e-06,
|
| 1469 |
+
"loss": 1.5241,
|
| 1470 |
+
"step": 2250
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 4.85,
|
| 1474 |
+
"learning_rate": 3.13862660944206e-06,
|
| 1475 |
+
"loss": 1.5202,
|
| 1476 |
+
"step": 2260
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 4.87,
|
| 1480 |
+
"learning_rate": 3.1259656652360516e-06,
|
| 1481 |
+
"loss": 1.5246,
|
| 1482 |
+
"step": 2270
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 4.89,
|
| 1486 |
+
"learning_rate": 3.113304721030043e-06,
|
| 1487 |
+
"loss": 1.5213,
|
| 1488 |
+
"step": 2280
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 4.91,
|
| 1492 |
+
"learning_rate": 3.1006437768240343e-06,
|
| 1493 |
+
"loss": 1.5239,
|
| 1494 |
+
"step": 2290
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 4.93,
|
| 1498 |
+
"learning_rate": 3.0879828326180256e-06,
|
| 1499 |
+
"loss": 1.5219,
|
| 1500 |
+
"step": 2300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 4.96,
|
| 1504 |
+
"learning_rate": 3.0753218884120165e-06,
|
| 1505 |
+
"loss": 1.5227,
|
| 1506 |
+
"step": 2310
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 4.98,
|
| 1510 |
+
"learning_rate": 3.0626609442060083e-06,
|
| 1511 |
+
"loss": 1.5235,
|
| 1512 |
+
"step": 2320
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 5.0,
|
| 1516 |
+
"learning_rate": 3.05e-06,
|
| 1517 |
+
"loss": 1.5208,
|
| 1518 |
+
"step": 2330
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 5.02,
|
| 1522 |
+
"learning_rate": 3.0373390557939913e-06,
|
| 1523 |
+
"loss": 1.6633,
|
| 1524 |
+
"step": 2340
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 5.04,
|
| 1528 |
+
"learning_rate": 3.024678111587983e-06,
|
| 1529 |
+
"loss": 1.521,
|
| 1530 |
+
"step": 2350
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 5.06,
|
| 1534 |
+
"learning_rate": 3.012017167381974e-06,
|
| 1535 |
+
"loss": 1.5219,
|
| 1536 |
+
"step": 2360
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 5.09,
|
| 1540 |
+
"learning_rate": 2.999356223175966e-06,
|
| 1541 |
+
"loss": 1.5228,
|
| 1542 |
+
"step": 2370
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 5.11,
|
| 1546 |
+
"learning_rate": 2.986695278969957e-06,
|
| 1547 |
+
"loss": 1.5236,
|
| 1548 |
+
"step": 2380
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 5.13,
|
| 1552 |
+
"learning_rate": 2.9740343347639484e-06,
|
| 1553 |
+
"loss": 1.5201,
|
| 1554 |
+
"step": 2390
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 5.15,
|
| 1558 |
+
"learning_rate": 2.96137339055794e-06,
|
| 1559 |
+
"loss": 1.5239,
|
| 1560 |
+
"step": 2400
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 5.15,
|
| 1564 |
+
"eval_loss": 1.4185978174209595,
|
| 1565 |
+
"eval_runtime": 78.1188,
|
| 1566 |
+
"eval_samples_per_second": 512.041,
|
| 1567 |
+
"eval_steps_per_second": 8.001,
|
| 1568 |
+
"step": 2400
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 5.17,
|
| 1572 |
+
"learning_rate": 2.9487124463519315e-06,
|
| 1573 |
+
"loss": 1.521,
|
| 1574 |
+
"step": 2410
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 5.19,
|
| 1578 |
+
"learning_rate": 2.9360515021459232e-06,
|
| 1579 |
+
"loss": 1.5202,
|
| 1580 |
+
"step": 2420
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 5.21,
|
| 1584 |
+
"learning_rate": 2.923390557939914e-06,
|
| 1585 |
+
"loss": 1.5206,
|
| 1586 |
+
"step": 2430
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 5.24,
|
| 1590 |
+
"learning_rate": 2.9107296137339054e-06,
|
| 1591 |
+
"loss": 1.5233,
|
| 1592 |
+
"step": 2440
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 5.26,
|
| 1596 |
+
"learning_rate": 2.898068669527897e-06,
|
| 1597 |
+
"loss": 1.5213,
|
| 1598 |
+
"step": 2450
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 5.28,
|
| 1602 |
+
"learning_rate": 2.8854077253218885e-06,
|
| 1603 |
+
"loss": 1.5242,
|
| 1604 |
+
"step": 2460
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 5.3,
|
| 1608 |
+
"learning_rate": 2.8727467811158794e-06,
|
| 1609 |
+
"loss": 1.5217,
|
| 1610 |
+
"step": 2470
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 5.32,
|
| 1614 |
+
"learning_rate": 2.860085836909871e-06,
|
| 1615 |
+
"loss": 1.5208,
|
| 1616 |
+
"step": 2480
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 5.34,
|
| 1620 |
+
"learning_rate": 2.8474248927038625e-06,
|
| 1621 |
+
"loss": 1.5247,
|
| 1622 |
+
"step": 2490
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 5.36,
|
| 1626 |
+
"learning_rate": 2.8347639484978542e-06,
|
| 1627 |
+
"loss": 1.5238,
|
| 1628 |
+
"step": 2500
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 5.39,
|
| 1632 |
+
"learning_rate": 2.8221030042918456e-06,
|
| 1633 |
+
"loss": 1.5214,
|
| 1634 |
+
"step": 2510
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 5.41,
|
| 1638 |
+
"learning_rate": 2.8094420600858365e-06,
|
| 1639 |
+
"loss": 1.523,
|
| 1640 |
+
"step": 2520
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 5.43,
|
| 1644 |
+
"learning_rate": 2.7967811158798286e-06,
|
| 1645 |
+
"loss": 1.5212,
|
| 1646 |
+
"step": 2530
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 5.45,
|
| 1650 |
+
"learning_rate": 2.7841201716738195e-06,
|
| 1651 |
+
"loss": 1.5213,
|
| 1652 |
+
"step": 2540
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.47,
|
| 1656 |
+
"learning_rate": 2.7714592274678113e-06,
|
| 1657 |
+
"loss": 1.522,
|
| 1658 |
+
"step": 2550
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 5.49,
|
| 1662 |
+
"learning_rate": 2.7587982832618026e-06,
|
| 1663 |
+
"loss": 1.5203,
|
| 1664 |
+
"step": 2560
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 5.49,
|
| 1668 |
+
"eval_loss": 1.4153515100479126,
|
| 1669 |
+
"eval_runtime": 78.1738,
|
| 1670 |
+
"eval_samples_per_second": 511.68,
|
| 1671 |
+
"eval_steps_per_second": 7.995,
|
| 1672 |
+
"step": 2560
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 5.51,
|
| 1676 |
+
"learning_rate": 2.746137339055794e-06,
|
| 1677 |
+
"loss": 1.5194,
|
| 1678 |
+
"step": 2570
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 5.54,
|
| 1682 |
+
"learning_rate": 2.7334763948497857e-06,
|
| 1683 |
+
"loss": 1.5199,
|
| 1684 |
+
"step": 2580
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 5.56,
|
| 1688 |
+
"learning_rate": 2.7208154506437766e-06,
|
| 1689 |
+
"loss": 1.5209,
|
| 1690 |
+
"step": 2590
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 5.58,
|
| 1694 |
+
"learning_rate": 2.7081545064377683e-06,
|
| 1695 |
+
"loss": 1.5241,
|
| 1696 |
+
"step": 2600
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 5.6,
|
| 1700 |
+
"learning_rate": 2.6954935622317597e-06,
|
| 1701 |
+
"loss": 1.5186,
|
| 1702 |
+
"step": 2610
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 5.62,
|
| 1706 |
+
"learning_rate": 2.682832618025751e-06,
|
| 1707 |
+
"loss": 1.5232,
|
| 1708 |
+
"step": 2620
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 5.64,
|
| 1712 |
+
"learning_rate": 2.6701716738197427e-06,
|
| 1713 |
+
"loss": 1.5215,
|
| 1714 |
+
"step": 2630
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 5.66,
|
| 1718 |
+
"learning_rate": 2.6575107296137336e-06,
|
| 1719 |
+
"loss": 1.5262,
|
| 1720 |
+
"step": 2640
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 5.69,
|
| 1724 |
+
"learning_rate": 2.644849785407725e-06,
|
| 1725 |
+
"loss": 1.5242,
|
| 1726 |
+
"step": 2650
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 5.71,
|
| 1730 |
+
"learning_rate": 2.6321888412017167e-06,
|
| 1731 |
+
"loss": 1.5209,
|
| 1732 |
+
"step": 2660
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 5.73,
|
| 1736 |
+
"learning_rate": 2.619527896995708e-06,
|
| 1737 |
+
"loss": 1.5205,
|
| 1738 |
+
"step": 2670
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 5.75,
|
| 1742 |
+
"learning_rate": 2.6068669527896998e-06,
|
| 1743 |
+
"loss": 1.5235,
|
| 1744 |
+
"step": 2680
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 5.77,
|
| 1748 |
+
"learning_rate": 2.594206008583691e-06,
|
| 1749 |
+
"loss": 1.5214,
|
| 1750 |
+
"step": 2690
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 5.79,
|
| 1754 |
+
"learning_rate": 2.581545064377682e-06,
|
| 1755 |
+
"loss": 1.5238,
|
| 1756 |
+
"step": 2700
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 5.81,
|
| 1760 |
+
"learning_rate": 2.5688841201716738e-06,
|
| 1761 |
+
"loss": 1.5238,
|
| 1762 |
+
"step": 2710
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 5.84,
|
| 1766 |
+
"learning_rate": 2.556223175965665e-06,
|
| 1767 |
+
"loss": 1.522,
|
| 1768 |
+
"step": 2720
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 5.84,
|
| 1772 |
+
"eval_loss": 1.416185736656189,
|
| 1773 |
+
"eval_runtime": 78.1964,
|
| 1774 |
+
"eval_samples_per_second": 511.533,
|
| 1775 |
+
"eval_steps_per_second": 7.993,
|
| 1776 |
+
"step": 2720
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 5.86,
|
| 1780 |
+
"learning_rate": 2.543562231759657e-06,
|
| 1781 |
+
"loss": 1.5245,
|
| 1782 |
+
"step": 2730
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 5.88,
|
| 1786 |
+
"learning_rate": 2.530901287553648e-06,
|
| 1787 |
+
"loss": 1.5218,
|
| 1788 |
+
"step": 2740
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 5.9,
|
| 1792 |
+
"learning_rate": 2.518240343347639e-06,
|
| 1793 |
+
"loss": 1.5231,
|
| 1794 |
+
"step": 2750
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 5.92,
|
| 1798 |
+
"learning_rate": 2.5055793991416312e-06,
|
| 1799 |
+
"loss": 1.5205,
|
| 1800 |
+
"step": 2760
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 5.94,
|
| 1804 |
+
"learning_rate": 2.492918454935622e-06,
|
| 1805 |
+
"loss": 1.523,
|
| 1806 |
+
"step": 2770
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 5.96,
|
| 1810 |
+
"learning_rate": 2.480257510729614e-06,
|
| 1811 |
+
"loss": 1.5246,
|
| 1812 |
+
"step": 2780
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 5.99,
|
| 1816 |
+
"learning_rate": 2.467596566523605e-06,
|
| 1817 |
+
"loss": 1.5226,
|
| 1818 |
+
"step": 2790
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 6.01,
|
| 1822 |
+
"learning_rate": 2.4549356223175965e-06,
|
| 1823 |
+
"loss": 1.6635,
|
| 1824 |
+
"step": 2800
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 6.03,
|
| 1828 |
+
"learning_rate": 2.4422746781115883e-06,
|
| 1829 |
+
"loss": 1.524,
|
| 1830 |
+
"step": 2810
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 6.05,
|
| 1834 |
+
"learning_rate": 2.429613733905579e-06,
|
| 1835 |
+
"loss": 1.522,
|
| 1836 |
+
"step": 2820
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.07,
|
| 1840 |
+
"learning_rate": 2.416952789699571e-06,
|
| 1841 |
+
"loss": 1.5204,
|
| 1842 |
+
"step": 2830
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 6.09,
|
| 1846 |
+
"learning_rate": 2.4042918454935622e-06,
|
| 1847 |
+
"loss": 1.5187,
|
| 1848 |
+
"step": 2840
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 6.12,
|
| 1852 |
+
"learning_rate": 2.3916309012875536e-06,
|
| 1853 |
+
"loss": 1.5209,
|
| 1854 |
+
"step": 2850
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 6.14,
|
| 1858 |
+
"learning_rate": 2.3789699570815453e-06,
|
| 1859 |
+
"loss": 1.5242,
|
| 1860 |
+
"step": 2860
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 6.16,
|
| 1864 |
+
"learning_rate": 2.3663090128755362e-06,
|
| 1865 |
+
"loss": 1.5214,
|
| 1866 |
+
"step": 2870
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 6.18,
|
| 1870 |
+
"learning_rate": 2.3536480686695275e-06,
|
| 1871 |
+
"loss": 1.5209,
|
| 1872 |
+
"step": 2880
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 6.18,
|
| 1876 |
+
"eval_loss": 1.4153697490692139,
|
| 1877 |
+
"eval_runtime": 78.1036,
|
| 1878 |
+
"eval_samples_per_second": 512.14,
|
| 1879 |
+
"eval_steps_per_second": 8.002,
|
| 1880 |
+
"step": 2880
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 6.2,
|
| 1884 |
+
"learning_rate": 2.3409871244635193e-06,
|
| 1885 |
+
"loss": 1.5226,
|
| 1886 |
+
"step": 2890
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 6.22,
|
| 1890 |
+
"learning_rate": 2.3283261802575106e-06,
|
| 1891 |
+
"loss": 1.5218,
|
| 1892 |
+
"step": 2900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.24,
|
| 1896 |
+
"learning_rate": 2.3156652360515024e-06,
|
| 1897 |
+
"loss": 1.5258,
|
| 1898 |
+
"step": 2910
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.27,
|
| 1902 |
+
"learning_rate": 2.3030042918454937e-06,
|
| 1903 |
+
"loss": 1.5218,
|
| 1904 |
+
"step": 2920
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 6.29,
|
| 1908 |
+
"learning_rate": 2.2903433476394846e-06,
|
| 1909 |
+
"loss": 1.5227,
|
| 1910 |
+
"step": 2930
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 6.31,
|
| 1914 |
+
"learning_rate": 2.2776824034334763e-06,
|
| 1915 |
+
"loss": 1.521,
|
| 1916 |
+
"step": 2940
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 6.33,
|
| 1920 |
+
"learning_rate": 2.2650214592274677e-06,
|
| 1921 |
+
"loss": 1.521,
|
| 1922 |
+
"step": 2950
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 6.35,
|
| 1926 |
+
"learning_rate": 2.2523605150214594e-06,
|
| 1927 |
+
"loss": 1.5237,
|
| 1928 |
+
"step": 2960
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 6.37,
|
| 1932 |
+
"learning_rate": 2.2396995708154507e-06,
|
| 1933 |
+
"loss": 1.5222,
|
| 1934 |
+
"step": 2970
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.39,
|
| 1938 |
+
"learning_rate": 2.2270386266094416e-06,
|
| 1939 |
+
"loss": 1.5219,
|
| 1940 |
+
"step": 2980
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.42,
|
| 1944 |
+
"learning_rate": 2.214377682403434e-06,
|
| 1945 |
+
"loss": 1.5237,
|
| 1946 |
+
"step": 2990
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 6.44,
|
| 1950 |
+
"learning_rate": 2.2017167381974247e-06,
|
| 1951 |
+
"loss": 1.5193,
|
| 1952 |
+
"step": 3000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 6.46,
|
| 1956 |
+
"learning_rate": 2.1890557939914165e-06,
|
| 1957 |
+
"loss": 1.525,
|
| 1958 |
+
"step": 3010
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 6.48,
|
| 1962 |
+
"learning_rate": 2.176394849785408e-06,
|
| 1963 |
+
"loss": 1.5192,
|
| 1964 |
+
"step": 3020
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 6.5,
|
| 1968 |
+
"learning_rate": 2.1637339055793987e-06,
|
| 1969 |
+
"loss": 1.5195,
|
| 1970 |
+
"step": 3030
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 6.52,
|
| 1974 |
+
"learning_rate": 2.151072961373391e-06,
|
| 1975 |
+
"loss": 1.5196,
|
| 1976 |
+
"step": 3040
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 6.52,
|
| 1980 |
+
"eval_loss": 1.4152772426605225,
|
| 1981 |
+
"eval_runtime": 78.1214,
|
| 1982 |
+
"eval_samples_per_second": 512.024,
|
| 1983 |
+
"eval_steps_per_second": 8.0,
|
| 1984 |
+
"step": 3040
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 6.54,
|
| 1988 |
+
"learning_rate": 2.1384120171673818e-06,
|
| 1989 |
+
"loss": 1.5185,
|
| 1990 |
+
"step": 3050
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 6.57,
|
| 1994 |
+
"learning_rate": 2.1257510729613735e-06,
|
| 1995 |
+
"loss": 1.5209,
|
| 1996 |
+
"step": 3060
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 6.59,
|
| 2000 |
+
"learning_rate": 2.113090128755365e-06,
|
| 2001 |
+
"loss": 1.5231,
|
| 2002 |
+
"step": 3070
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 6.61,
|
| 2006 |
+
"learning_rate": 2.100429184549356e-06,
|
| 2007 |
+
"loss": 1.5206,
|
| 2008 |
+
"step": 3080
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 6.63,
|
| 2012 |
+
"learning_rate": 2.087768240343348e-06,
|
| 2013 |
+
"loss": 1.5227,
|
| 2014 |
+
"step": 3090
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 6.65,
|
| 2018 |
+
"learning_rate": 2.075107296137339e-06,
|
| 2019 |
+
"loss": 1.5225,
|
| 2020 |
+
"step": 3100
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 6.67,
|
| 2024 |
+
"learning_rate": 2.06244635193133e-06,
|
| 2025 |
+
"loss": 1.5236,
|
| 2026 |
+
"step": 3110
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 6.69,
|
| 2030 |
+
"learning_rate": 2.049785407725322e-06,
|
| 2031 |
+
"loss": 1.5218,
|
| 2032 |
+
"step": 3120
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 6.72,
|
| 2036 |
+
"learning_rate": 2.037124463519313e-06,
|
| 2037 |
+
"loss": 1.5207,
|
| 2038 |
+
"step": 3130
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 6.74,
|
| 2042 |
+
"learning_rate": 2.024463519313305e-06,
|
| 2043 |
+
"loss": 1.5236,
|
| 2044 |
+
"step": 3140
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 6.76,
|
| 2048 |
+
"learning_rate": 2.0118025751072963e-06,
|
| 2049 |
+
"loss": 1.5204,
|
| 2050 |
+
"step": 3150
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 6.78,
|
| 2054 |
+
"learning_rate": 1.999141630901287e-06,
|
| 2055 |
+
"loss": 1.5204,
|
| 2056 |
+
"step": 3160
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 6.8,
|
| 2060 |
+
"learning_rate": 1.9864806866952794e-06,
|
| 2061 |
+
"loss": 1.5234,
|
| 2062 |
+
"step": 3170
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 6.82,
|
| 2066 |
+
"learning_rate": 1.9738197424892703e-06,
|
| 2067 |
+
"loss": 1.5226,
|
| 2068 |
+
"step": 3180
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 6.84,
|
| 2072 |
+
"learning_rate": 1.961158798283262e-06,
|
| 2073 |
+
"loss": 1.5203,
|
| 2074 |
+
"step": 3190
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 6.87,
|
| 2078 |
+
"learning_rate": 1.9484978540772533e-06,
|
| 2079 |
+
"loss": 1.5209,
|
| 2080 |
+
"step": 3200
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 6.87,
|
| 2084 |
+
"eval_loss": 1.4121816158294678,
|
| 2085 |
+
"eval_runtime": 78.0644,
|
| 2086 |
+
"eval_samples_per_second": 512.398,
|
| 2087 |
+
"eval_steps_per_second": 8.006,
|
| 2088 |
+
"step": 3200
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 6.89,
|
| 2092 |
+
"learning_rate": 1.9358369098712447e-06,
|
| 2093 |
+
"loss": 1.5213,
|
| 2094 |
+
"step": 3210
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 6.91,
|
| 2098 |
+
"learning_rate": 1.9231759656652364e-06,
|
| 2099 |
+
"loss": 1.524,
|
| 2100 |
+
"step": 3220
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 6.93,
|
| 2104 |
+
"learning_rate": 1.9105150214592273e-06,
|
| 2105 |
+
"loss": 1.5224,
|
| 2106 |
+
"step": 3230
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 6.95,
|
| 2110 |
+
"learning_rate": 1.8978540772532188e-06,
|
| 2111 |
+
"loss": 1.5196,
|
| 2112 |
+
"step": 3240
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 6.97,
|
| 2116 |
+
"learning_rate": 1.8851931330472106e-06,
|
| 2117 |
+
"loss": 1.5237,
|
| 2118 |
+
"step": 3250
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 6.99,
|
| 2122 |
+
"learning_rate": 1.8725321888412017e-06,
|
| 2123 |
+
"loss": 1.5227,
|
| 2124 |
+
"step": 3260
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 7.02,
|
| 2128 |
+
"learning_rate": 1.8598712446351935e-06,
|
| 2129 |
+
"loss": 1.6653,
|
| 2130 |
+
"step": 3270
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.04,
|
| 2134 |
+
"learning_rate": 1.8472103004291844e-06,
|
| 2135 |
+
"loss": 1.5187,
|
| 2136 |
+
"step": 3280
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 7.06,
|
| 2140 |
+
"learning_rate": 1.8345493562231757e-06,
|
| 2141 |
+
"loss": 1.5202,
|
| 2142 |
+
"step": 3290
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 7.08,
|
| 2146 |
+
"learning_rate": 1.8218884120171672e-06,
|
| 2147 |
+
"loss": 1.5237,
|
| 2148 |
+
"step": 3300
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 7.1,
|
| 2152 |
+
"learning_rate": 1.809227467811159e-06,
|
| 2153 |
+
"loss": 1.5209,
|
| 2154 |
+
"step": 3310
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 7.12,
|
| 2158 |
+
"learning_rate": 1.7965665236051505e-06,
|
| 2159 |
+
"loss": 1.5229,
|
| 2160 |
+
"step": 3320
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 7.15,
|
| 2164 |
+
"learning_rate": 1.7839055793991418e-06,
|
| 2165 |
+
"loss": 1.52,
|
| 2166 |
+
"step": 3330
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 7.17,
|
| 2170 |
+
"learning_rate": 1.7712446351931327e-06,
|
| 2171 |
+
"loss": 1.5195,
|
| 2172 |
+
"step": 3340
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.19,
|
| 2176 |
+
"learning_rate": 1.7585836909871243e-06,
|
| 2177 |
+
"loss": 1.5216,
|
| 2178 |
+
"step": 3350
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 7.21,
|
| 2182 |
+
"learning_rate": 1.745922746781116e-06,
|
| 2183 |
+
"loss": 1.5202,
|
| 2184 |
+
"step": 3360
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 7.21,
|
| 2188 |
+
"eval_loss": 1.4146045446395874,
|
| 2189 |
+
"eval_runtime": 78.1595,
|
| 2190 |
+
"eval_samples_per_second": 511.774,
|
| 2191 |
+
"eval_steps_per_second": 7.996,
|
| 2192 |
+
"step": 3360
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 7.23,
|
| 2196 |
+
"learning_rate": 1.7332618025751075e-06,
|
| 2197 |
+
"loss": 1.5226,
|
| 2198 |
+
"step": 3370
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 7.25,
|
| 2202 |
+
"learning_rate": 1.7206008583690989e-06,
|
| 2203 |
+
"loss": 1.5221,
|
| 2204 |
+
"step": 3380
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 7.27,
|
| 2208 |
+
"learning_rate": 1.7079399141630898e-06,
|
| 2209 |
+
"loss": 1.5253,
|
| 2210 |
+
"step": 3390
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 7.3,
|
| 2214 |
+
"learning_rate": 1.6952789699570822e-06,
|
| 2215 |
+
"loss": 1.5236,
|
| 2216 |
+
"step": 3400
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 7.32,
|
| 2220 |
+
"learning_rate": 1.6826180257510726e-06,
|
| 2221 |
+
"loss": 1.5209,
|
| 2222 |
+
"step": 3410
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 7.34,
|
| 2226 |
+
"learning_rate": 1.6699570815450646e-06,
|
| 2227 |
+
"loss": 1.5211,
|
| 2228 |
+
"step": 3420
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 7.36,
|
| 2232 |
+
"learning_rate": 1.657296137339056e-06,
|
| 2233 |
+
"loss": 1.5218,
|
| 2234 |
+
"step": 3430
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 7.38,
|
| 2238 |
+
"learning_rate": 1.6446351931330472e-06,
|
| 2239 |
+
"loss": 1.5233,
|
| 2240 |
+
"step": 3440
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 7.4,
|
| 2244 |
+
"learning_rate": 1.6319742489270388e-06,
|
| 2245 |
+
"loss": 1.5219,
|
| 2246 |
+
"step": 3450
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 7.42,
|
| 2250 |
+
"learning_rate": 1.6193133047210297e-06,
|
| 2251 |
+
"loss": 1.523,
|
| 2252 |
+
"step": 3460
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 7.45,
|
| 2256 |
+
"learning_rate": 1.6066523605150216e-06,
|
| 2257 |
+
"loss": 1.5204,
|
| 2258 |
+
"step": 3470
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 7.47,
|
| 2262 |
+
"learning_rate": 1.593991416309013e-06,
|
| 2263 |
+
"loss": 1.5226,
|
| 2264 |
+
"step": 3480
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 7.49,
|
| 2268 |
+
"learning_rate": 1.5813304721030043e-06,
|
| 2269 |
+
"loss": 1.5178,
|
| 2270 |
+
"step": 3490
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 7.51,
|
| 2274 |
+
"learning_rate": 1.5686695278969958e-06,
|
| 2275 |
+
"loss": 1.5184,
|
| 2276 |
+
"step": 3500
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 7.53,
|
| 2280 |
+
"learning_rate": 1.556008583690987e-06,
|
| 2281 |
+
"loss": 1.5205,
|
| 2282 |
+
"step": 3510
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 7.55,
|
| 2286 |
+
"learning_rate": 1.543347639484978e-06,
|
| 2287 |
+
"loss": 1.5192,
|
| 2288 |
+
"step": 3520
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 7.55,
|
| 2292 |
+
"eval_loss": 1.4141193628311157,
|
| 2293 |
+
"eval_runtime": 78.1244,
|
| 2294 |
+
"eval_samples_per_second": 512.004,
|
| 2295 |
+
"eval_steps_per_second": 8.0,
|
| 2296 |
+
"step": 3520
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 7.57,
|
| 2300 |
+
"learning_rate": 1.53068669527897e-06,
|
| 2301 |
+
"loss": 1.5211,
|
| 2302 |
+
"step": 3530
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 7.6,
|
| 2306 |
+
"learning_rate": 1.5180257510729613e-06,
|
| 2307 |
+
"loss": 1.5212,
|
| 2308 |
+
"step": 3540
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 7.62,
|
| 2312 |
+
"learning_rate": 1.505364806866953e-06,
|
| 2313 |
+
"loss": 1.5226,
|
| 2314 |
+
"step": 3550
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 7.64,
|
| 2318 |
+
"learning_rate": 1.4927038626609444e-06,
|
| 2319 |
+
"loss": 1.5204,
|
| 2320 |
+
"step": 3560
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 7.66,
|
| 2324 |
+
"learning_rate": 1.4800429184549353e-06,
|
| 2325 |
+
"loss": 1.5258,
|
| 2326 |
+
"step": 3570
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 7.68,
|
| 2330 |
+
"learning_rate": 1.467381974248927e-06,
|
| 2331 |
+
"loss": 1.5217,
|
| 2332 |
+
"step": 3580
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 7.7,
|
| 2336 |
+
"learning_rate": 1.4547210300429184e-06,
|
| 2337 |
+
"loss": 1.5181,
|
| 2338 |
+
"step": 3590
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 7.72,
|
| 2342 |
+
"learning_rate": 1.4420600858369101e-06,
|
| 2343 |
+
"loss": 1.5192,
|
| 2344 |
+
"step": 3600
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 7.75,
|
| 2348 |
+
"learning_rate": 1.4293991416309015e-06,
|
| 2349 |
+
"loss": 1.5235,
|
| 2350 |
+
"step": 3610
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 7.77,
|
| 2354 |
+
"learning_rate": 1.4167381974248924e-06,
|
| 2355 |
+
"loss": 1.5178,
|
| 2356 |
+
"step": 3620
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 7.79,
|
| 2360 |
+
"learning_rate": 1.4040772532188841e-06,
|
| 2361 |
+
"loss": 1.5228,
|
| 2362 |
+
"step": 3630
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 7.81,
|
| 2366 |
+
"learning_rate": 1.3914163090128754e-06,
|
| 2367 |
+
"loss": 1.5212,
|
| 2368 |
+
"step": 3640
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 7.83,
|
| 2372 |
+
"learning_rate": 1.3787553648068672e-06,
|
| 2373 |
+
"loss": 1.5251,
|
| 2374 |
+
"step": 3650
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 7.85,
|
| 2378 |
+
"learning_rate": 1.3660944206008585e-06,
|
| 2379 |
+
"loss": 1.5254,
|
| 2380 |
+
"step": 3660
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 7.87,
|
| 2384 |
+
"learning_rate": 1.3534334763948494e-06,
|
| 2385 |
+
"loss": 1.5213,
|
| 2386 |
+
"step": 3670
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 7.9,
|
| 2390 |
+
"learning_rate": 1.3407725321888416e-06,
|
| 2391 |
+
"loss": 1.5215,
|
| 2392 |
+
"step": 3680
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 7.9,
|
| 2396 |
+
"eval_loss": 1.4122941493988037,
|
| 2397 |
+
"eval_runtime": 78.0947,
|
| 2398 |
+
"eval_samples_per_second": 512.199,
|
| 2399 |
+
"eval_steps_per_second": 8.003,
|
| 2400 |
+
"step": 3680
|
| 2401 |
+
},
|
| 2402 |
+
{
|
| 2403 |
+
"epoch": 7.92,
|
| 2404 |
+
"learning_rate": 1.3281115879828325e-06,
|
| 2405 |
+
"loss": 1.5199,
|
| 2406 |
+
"step": 3690
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 7.94,
|
| 2410 |
+
"learning_rate": 1.3154506437768242e-06,
|
| 2411 |
+
"loss": 1.5236,
|
| 2412 |
+
"step": 3700
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 7.96,
|
| 2416 |
+
"learning_rate": 1.3027896995708156e-06,
|
| 2417 |
+
"loss": 1.5216,
|
| 2418 |
+
"step": 3710
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 7.98,
|
| 2422 |
+
"learning_rate": 1.2901287553648069e-06,
|
| 2423 |
+
"loss": 1.5219,
|
| 2424 |
+
"step": 3720
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 8.0,
|
| 2428 |
+
"learning_rate": 1.2774678111587986e-06,
|
| 2429 |
+
"loss": 1.664,
|
| 2430 |
+
"step": 3730
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 8.03,
|
| 2434 |
+
"learning_rate": 1.2648068669527895e-06,
|
| 2435 |
+
"loss": 1.5229,
|
| 2436 |
+
"step": 3740
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 8.05,
|
| 2440 |
+
"learning_rate": 1.2521459227467809e-06,
|
| 2441 |
+
"loss": 1.5241,
|
| 2442 |
+
"step": 3750
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 8.07,
|
| 2446 |
+
"learning_rate": 1.2394849785407726e-06,
|
| 2447 |
+
"loss": 1.5221,
|
| 2448 |
+
"step": 3760
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 8.09,
|
| 2452 |
+
"learning_rate": 1.226824034334764e-06,
|
| 2453 |
+
"loss": 1.5236,
|
| 2454 |
+
"step": 3770
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 8.11,
|
| 2458 |
+
"learning_rate": 1.2141630901287557e-06,
|
| 2459 |
+
"loss": 1.5215,
|
| 2460 |
+
"step": 3780
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 8.13,
|
| 2464 |
+
"learning_rate": 1.201502145922747e-06,
|
| 2465 |
+
"loss": 1.5207,
|
| 2466 |
+
"step": 3790
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 8.15,
|
| 2470 |
+
"learning_rate": 1.188841201716738e-06,
|
| 2471 |
+
"loss": 1.5206,
|
| 2472 |
+
"step": 3800
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 8.18,
|
| 2476 |
+
"learning_rate": 1.1761802575107297e-06,
|
| 2477 |
+
"loss": 1.5198,
|
| 2478 |
+
"step": 3810
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 8.2,
|
| 2482 |
+
"learning_rate": 1.163519313304721e-06,
|
| 2483 |
+
"loss": 1.518,
|
| 2484 |
+
"step": 3820
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 8.22,
|
| 2488 |
+
"learning_rate": 1.1508583690987127e-06,
|
| 2489 |
+
"loss": 1.521,
|
| 2490 |
+
"step": 3830
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 8.24,
|
| 2494 |
+
"learning_rate": 1.138197424892704e-06,
|
| 2495 |
+
"loss": 1.5228,
|
| 2496 |
+
"step": 3840
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 8.24,
|
| 2500 |
+
"eval_loss": 1.4147212505340576,
|
| 2501 |
+
"eval_runtime": 78.136,
|
| 2502 |
+
"eval_samples_per_second": 511.928,
|
| 2503 |
+
"eval_steps_per_second": 7.999,
|
| 2504 |
+
"step": 3840
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 8.26,
|
| 2508 |
+
"learning_rate": 1.125536480686695e-06,
|
| 2509 |
+
"loss": 1.5228,
|
| 2510 |
+
"step": 3850
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 8.28,
|
| 2514 |
+
"learning_rate": 1.1128755364806867e-06,
|
| 2515 |
+
"loss": 1.5232,
|
| 2516 |
+
"step": 3860
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 8.3,
|
| 2520 |
+
"learning_rate": 1.100214592274678e-06,
|
| 2521 |
+
"loss": 1.5213,
|
| 2522 |
+
"step": 3870
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 8.33,
|
| 2526 |
+
"learning_rate": 1.0875536480686698e-06,
|
| 2527 |
+
"loss": 1.5173,
|
| 2528 |
+
"step": 3880
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 8.35,
|
| 2532 |
+
"learning_rate": 1.074892703862661e-06,
|
| 2533 |
+
"loss": 1.5206,
|
| 2534 |
+
"step": 3890
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 8.37,
|
| 2538 |
+
"learning_rate": 1.062231759656652e-06,
|
| 2539 |
+
"loss": 1.5224,
|
| 2540 |
+
"step": 3900
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 8.39,
|
| 2544 |
+
"learning_rate": 1.049570815450644e-06,
|
| 2545 |
+
"loss": 1.5215,
|
| 2546 |
+
"step": 3910
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 8.41,
|
| 2550 |
+
"learning_rate": 1.036909871244635e-06,
|
| 2551 |
+
"loss": 1.5211,
|
| 2552 |
+
"step": 3920
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 8.43,
|
| 2556 |
+
"learning_rate": 1.0242489270386264e-06,
|
| 2557 |
+
"loss": 1.5201,
|
| 2558 |
+
"step": 3930
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 8.45,
|
| 2562 |
+
"learning_rate": 1.0115879828326181e-06,
|
| 2563 |
+
"loss": 1.5248,
|
| 2564 |
+
"step": 3940
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 8.48,
|
| 2568 |
+
"learning_rate": 9.989270386266093e-07,
|
| 2569 |
+
"loss": 1.5194,
|
| 2570 |
+
"step": 3950
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 8.5,
|
| 2574 |
+
"learning_rate": 9.862660944206012e-07,
|
| 2575 |
+
"loss": 1.5202,
|
| 2576 |
+
"step": 3960
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 8.52,
|
| 2580 |
+
"learning_rate": 9.736051502145921e-07,
|
| 2581 |
+
"loss": 1.518,
|
| 2582 |
+
"step": 3970
|
| 2583 |
+
},
|
| 2584 |
+
{
|
| 2585 |
+
"epoch": 8.54,
|
| 2586 |
+
"learning_rate": 9.609442060085834e-07,
|
| 2587 |
+
"loss": 1.5199,
|
| 2588 |
+
"step": 3980
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 8.56,
|
| 2592 |
+
"learning_rate": 9.482832618025752e-07,
|
| 2593 |
+
"loss": 1.5221,
|
| 2594 |
+
"step": 3990
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 8.58,
|
| 2598 |
+
"learning_rate": 9.356223175965663e-07,
|
| 2599 |
+
"loss": 1.5222,
|
| 2600 |
+
"step": 4000
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 8.58,
|
| 2604 |
+
"eval_loss": 1.4144057035446167,
|
| 2605 |
+
"eval_runtime": 78.0968,
|
| 2606 |
+
"eval_samples_per_second": 512.185,
|
| 2607 |
+
"eval_steps_per_second": 8.003,
|
| 2608 |
+
"step": 4000
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 8.6,
|
| 2612 |
+
"learning_rate": 9.229613733905583e-07,
|
| 2613 |
+
"loss": 1.5217,
|
| 2614 |
+
"step": 4010
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 8.63,
|
| 2618 |
+
"learning_rate": 9.103004291845494e-07,
|
| 2619 |
+
"loss": 1.5225,
|
| 2620 |
+
"step": 4020
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 8.65,
|
| 2624 |
+
"learning_rate": 8.976394849785405e-07,
|
| 2625 |
+
"loss": 1.5232,
|
| 2626 |
+
"step": 4030
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 8.67,
|
| 2630 |
+
"learning_rate": 8.849785407725323e-07,
|
| 2631 |
+
"loss": 1.5248,
|
| 2632 |
+
"step": 4040
|
| 2633 |
+
},
|
| 2634 |
+
{
|
| 2635 |
+
"epoch": 8.69,
|
| 2636 |
+
"learning_rate": 8.723175965665236e-07,
|
| 2637 |
+
"loss": 1.5213,
|
| 2638 |
+
"step": 4050
|
| 2639 |
+
},
|
| 2640 |
+
{
|
| 2641 |
+
"epoch": 8.71,
|
| 2642 |
+
"learning_rate": 8.596566523605154e-07,
|
| 2643 |
+
"loss": 1.5182,
|
| 2644 |
+
"step": 4060
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 8.73,
|
| 2648 |
+
"learning_rate": 8.469957081545065e-07,
|
| 2649 |
+
"loss": 1.5203,
|
| 2650 |
+
"step": 4070
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 8.75,
|
| 2654 |
+
"learning_rate": 8.343347639484976e-07,
|
| 2655 |
+
"loss": 1.523,
|
| 2656 |
+
"step": 4080
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 8.78,
|
| 2660 |
+
"learning_rate": 8.216738197424896e-07,
|
| 2661 |
+
"loss": 1.5213,
|
| 2662 |
+
"step": 4090
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 8.8,
|
| 2666 |
+
"learning_rate": 8.090128755364807e-07,
|
| 2667 |
+
"loss": 1.5235,
|
| 2668 |
+
"step": 4100
|
| 2669 |
+
},
|
| 2670 |
+
{
|
| 2671 |
+
"epoch": 8.82,
|
| 2672 |
+
"learning_rate": 7.963519313304725e-07,
|
| 2673 |
+
"loss": 1.5236,
|
| 2674 |
+
"step": 4110
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 8.84,
|
| 2678 |
+
"learning_rate": 7.836909871244636e-07,
|
| 2679 |
+
"loss": 1.5238,
|
| 2680 |
+
"step": 4120
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 8.86,
|
| 2684 |
+
"learning_rate": 7.710300429184549e-07,
|
| 2685 |
+
"loss": 1.5228,
|
| 2686 |
+
"step": 4130
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 8.88,
|
| 2690 |
+
"learning_rate": 7.583690987124464e-07,
|
| 2691 |
+
"loss": 1.5228,
|
| 2692 |
+
"step": 4140
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 8.9,
|
| 2696 |
+
"learning_rate": 7.457081545064378e-07,
|
| 2697 |
+
"loss": 1.5223,
|
| 2698 |
+
"step": 4150
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 8.93,
|
| 2702 |
+
"learning_rate": 7.330472103004289e-07,
|
| 2703 |
+
"loss": 1.5201,
|
| 2704 |
+
"step": 4160
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 8.93,
|
| 2708 |
+
"eval_loss": 1.4172754287719727,
|
| 2709 |
+
"eval_runtime": 78.1438,
|
| 2710 |
+
"eval_samples_per_second": 511.877,
|
| 2711 |
+
"eval_steps_per_second": 7.998,
|
| 2712 |
+
"step": 4160
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 8.95,
|
| 2716 |
+
"learning_rate": 7.203862660944207e-07,
|
| 2717 |
+
"loss": 1.5214,
|
| 2718 |
+
"step": 4170
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 8.97,
|
| 2722 |
+
"learning_rate": 7.077253218884118e-07,
|
| 2723 |
+
"loss": 1.5222,
|
| 2724 |
+
"step": 4180
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 8.99,
|
| 2728 |
+
"learning_rate": 6.950643776824037e-07,
|
| 2729 |
+
"loss": 1.5215,
|
| 2730 |
+
"step": 4190
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 9.01,
|
| 2734 |
+
"learning_rate": 6.824034334763948e-07,
|
| 2735 |
+
"loss": 1.6641,
|
| 2736 |
+
"step": 4200
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 9.03,
|
| 2740 |
+
"learning_rate": 6.69742489270386e-07,
|
| 2741 |
+
"loss": 1.5223,
|
| 2742 |
+
"step": 4210
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 9.06,
|
| 2746 |
+
"learning_rate": 6.570815450643778e-07,
|
| 2747 |
+
"loss": 1.5223,
|
| 2748 |
+
"step": 4220
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 9.08,
|
| 2752 |
+
"learning_rate": 6.44420600858369e-07,
|
| 2753 |
+
"loss": 1.5227,
|
| 2754 |
+
"step": 4230
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 9.1,
|
| 2758 |
+
"learning_rate": 6.317596566523608e-07,
|
| 2759 |
+
"loss": 1.5217,
|
| 2760 |
+
"step": 4240
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 9.12,
|
| 2764 |
+
"learning_rate": 6.19098712446352e-07,
|
| 2765 |
+
"loss": 1.5217,
|
| 2766 |
+
"step": 4250
|
| 2767 |
+
},
|
| 2768 |
+
{
|
| 2769 |
+
"epoch": 9.14,
|
| 2770 |
+
"learning_rate": 6.064377682403431e-07,
|
| 2771 |
+
"loss": 1.5196,
|
| 2772 |
+
"step": 4260
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 9.16,
|
| 2776 |
+
"learning_rate": 5.937768240343349e-07,
|
| 2777 |
+
"loss": 1.5198,
|
| 2778 |
+
"step": 4270
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 9.18,
|
| 2782 |
+
"learning_rate": 5.81115879828326e-07,
|
| 2783 |
+
"loss": 1.5228,
|
| 2784 |
+
"step": 4280
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 9.21,
|
| 2788 |
+
"learning_rate": 5.684549356223179e-07,
|
| 2789 |
+
"loss": 1.5192,
|
| 2790 |
+
"step": 4290
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 9.23,
|
| 2794 |
+
"learning_rate": 5.55793991416309e-07,
|
| 2795 |
+
"loss": 1.5203,
|
| 2796 |
+
"step": 4300
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 9.25,
|
| 2800 |
+
"learning_rate": 5.431330472103002e-07,
|
| 2801 |
+
"loss": 1.5235,
|
| 2802 |
+
"step": 4310
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 9.27,
|
| 2806 |
+
"learning_rate": 5.304721030042921e-07,
|
| 2807 |
+
"loss": 1.523,
|
| 2808 |
+
"step": 4320
|
| 2809 |
+
},
|
| 2810 |
+
{
|
| 2811 |
+
"epoch": 9.27,
|
| 2812 |
+
"eval_loss": 1.4170712232589722,
|
| 2813 |
+
"eval_runtime": 78.1196,
|
| 2814 |
+
"eval_samples_per_second": 512.035,
|
| 2815 |
+
"eval_steps_per_second": 8.001,
|
| 2816 |
+
"step": 4320
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 9.29,
|
| 2820 |
+
"learning_rate": 5.178111587982832e-07,
|
| 2821 |
+
"loss": 1.5227,
|
| 2822 |
+
"step": 4330
|
| 2823 |
+
},
|
| 2824 |
+
{
|
| 2825 |
+
"epoch": 9.31,
|
| 2826 |
+
"learning_rate": 5.05150214592275e-07,
|
| 2827 |
+
"loss": 1.5211,
|
| 2828 |
+
"step": 4340
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 9.33,
|
| 2832 |
+
"learning_rate": 4.924892703862662e-07,
|
| 2833 |
+
"loss": 1.5211,
|
| 2834 |
+
"step": 4350
|
| 2835 |
+
},
|
| 2836 |
+
{
|
| 2837 |
+
"epoch": 9.36,
|
| 2838 |
+
"learning_rate": 4.798283261802574e-07,
|
| 2839 |
+
"loss": 1.5206,
|
| 2840 |
+
"step": 4360
|
| 2841 |
+
},
|
| 2842 |
+
{
|
| 2843 |
+
"epoch": 9.38,
|
| 2844 |
+
"learning_rate": 4.6716738197424914e-07,
|
| 2845 |
+
"loss": 1.5224,
|
| 2846 |
+
"step": 4370
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 9.4,
|
| 2850 |
+
"learning_rate": 4.545064377682403e-07,
|
| 2851 |
+
"loss": 1.5221,
|
| 2852 |
+
"step": 4380
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 9.42,
|
| 2856 |
+
"learning_rate": 4.4184549356223147e-07,
|
| 2857 |
+
"loss": 1.5203,
|
| 2858 |
+
"step": 4390
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 9.44,
|
| 2862 |
+
"learning_rate": 4.291845493562233e-07,
|
| 2863 |
+
"loss": 1.5229,
|
| 2864 |
+
"step": 4400
|
| 2865 |
+
},
|
| 2866 |
+
{
|
| 2867 |
+
"epoch": 9.46,
|
| 2868 |
+
"learning_rate": 4.1652360515021444e-07,
|
| 2869 |
+
"loss": 1.5219,
|
| 2870 |
+
"step": 4410
|
| 2871 |
+
},
|
| 2872 |
+
{
|
| 2873 |
+
"epoch": 9.48,
|
| 2874 |
+
"learning_rate": 4.038626609442062e-07,
|
| 2875 |
+
"loss": 1.5188,
|
| 2876 |
+
"step": 4420
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 9.51,
|
| 2880 |
+
"learning_rate": 3.912017167381974e-07,
|
| 2881 |
+
"loss": 1.52,
|
| 2882 |
+
"step": 4430
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"epoch": 9.53,
|
| 2886 |
+
"learning_rate": 3.785407725321886e-07,
|
| 2887 |
+
"loss": 1.5204,
|
| 2888 |
+
"step": 4440
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 9.55,
|
| 2892 |
+
"learning_rate": 3.6587982832618037e-07,
|
| 2893 |
+
"loss": 1.5176,
|
| 2894 |
+
"step": 4450
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 9.57,
|
| 2898 |
+
"learning_rate": 3.5321888412017154e-07,
|
| 2899 |
+
"loss": 1.522,
|
| 2900 |
+
"step": 4460
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 9.59,
|
| 2904 |
+
"learning_rate": 3.405579399141634e-07,
|
| 2905 |
+
"loss": 1.5204,
|
| 2906 |
+
"step": 4470
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"epoch": 9.61,
|
| 2910 |
+
"learning_rate": 3.2789699570815456e-07,
|
| 2911 |
+
"loss": 1.5212,
|
| 2912 |
+
"step": 4480
|
| 2913 |
+
},
|
| 2914 |
+
{
|
| 2915 |
+
"epoch": 9.61,
|
| 2916 |
+
"eval_loss": 1.4148975610733032,
|
| 2917 |
+
"eval_runtime": 78.114,
|
| 2918 |
+
"eval_samples_per_second": 512.072,
|
| 2919 |
+
"eval_steps_per_second": 8.001,
|
| 2920 |
+
"step": 4480
|
| 2921 |
+
},
|
| 2922 |
+
{
|
| 2923 |
+
"epoch": 9.63,
|
| 2924 |
+
"learning_rate": 3.1523605150214567e-07,
|
| 2925 |
+
"loss": 1.5188,
|
| 2926 |
+
"step": 4490
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 9.66,
|
| 2930 |
+
"learning_rate": 3.0257510729613753e-07,
|
| 2931 |
+
"loss": 1.5224,
|
| 2932 |
+
"step": 4500
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 9.68,
|
| 2936 |
+
"learning_rate": 2.8991416309012864e-07,
|
| 2937 |
+
"loss": 1.5266,
|
| 2938 |
+
"step": 4510
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 9.7,
|
| 2942 |
+
"learning_rate": 2.772532188841205e-07,
|
| 2943 |
+
"loss": 1.5218,
|
| 2944 |
+
"step": 4520
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 9.72,
|
| 2948 |
+
"learning_rate": 2.645922746781116e-07,
|
| 2949 |
+
"loss": 1.5205,
|
| 2950 |
+
"step": 4530
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 9.74,
|
| 2954 |
+
"learning_rate": 2.5193133047210283e-07,
|
| 2955 |
+
"loss": 1.5228,
|
| 2956 |
+
"step": 4540
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 9.76,
|
| 2960 |
+
"learning_rate": 2.3927038626609463e-07,
|
| 2961 |
+
"loss": 1.5214,
|
| 2962 |
+
"step": 4550
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 9.78,
|
| 2966 |
+
"learning_rate": 2.266094420600858e-07,
|
| 2967 |
+
"loss": 1.5233,
|
| 2968 |
+
"step": 4560
|
| 2969 |
+
},
|
| 2970 |
+
{
|
| 2971 |
+
"epoch": 9.81,
|
| 2972 |
+
"learning_rate": 2.1394849785407693e-07,
|
| 2973 |
+
"loss": 1.5209,
|
| 2974 |
+
"step": 4570
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 9.83,
|
| 2978 |
+
"learning_rate": 2.0128755364806876e-07,
|
| 2979 |
+
"loss": 1.5241,
|
| 2980 |
+
"step": 4580
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 9.85,
|
| 2984 |
+
"learning_rate": 1.8862660944205993e-07,
|
| 2985 |
+
"loss": 1.521,
|
| 2986 |
+
"step": 4590
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 9.87,
|
| 2990 |
+
"learning_rate": 1.7596566523605173e-07,
|
| 2991 |
+
"loss": 1.5219,
|
| 2992 |
+
"step": 4600
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 9.89,
|
| 2996 |
+
"learning_rate": 1.633047210300429e-07,
|
| 2997 |
+
"loss": 1.5209,
|
| 2998 |
+
"step": 4610
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 9.91,
|
| 3002 |
+
"learning_rate": 1.5064377682403406e-07,
|
| 3003 |
+
"loss": 1.5249,
|
| 3004 |
+
"step": 4620
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 9.93,
|
| 3008 |
+
"learning_rate": 1.3798283261802586e-07,
|
| 3009 |
+
"loss": 1.5231,
|
| 3010 |
+
"step": 4630
|
| 3011 |
+
},
|
| 3012 |
+
{
|
| 3013 |
+
"epoch": 9.96,
|
| 3014 |
+
"learning_rate": 1.2532188841201703e-07,
|
| 3015 |
+
"loss": 1.522,
|
| 3016 |
+
"step": 4640
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 9.96,
|
| 3020 |
+
"eval_loss": 1.4186652898788452,
|
| 3021 |
+
"eval_runtime": 78.1323,
|
| 3022 |
+
"eval_samples_per_second": 511.952,
|
| 3023 |
+
"eval_steps_per_second": 7.999,
|
| 3024 |
+
"step": 4640
|
| 3025 |
+
},
|
| 3026 |
+
{
|
| 3027 |
+
"epoch": 9.98,
|
| 3028 |
+
"learning_rate": 1.1266094420600883e-07,
|
| 3029 |
+
"loss": 1.5239,
|
| 3030 |
+
"step": 4650
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 10.0,
|
| 3034 |
+
"learning_rate": 1e-07,
|
| 3035 |
+
"loss": 1.5179,
|
| 3036 |
+
"step": 4660
|
| 3037 |
+
},
|
| 3038 |
+
{
|
| 3039 |
+
"before_init_mem_cpu": 933711872,
|
| 3040 |
+
"before_init_mem_gpu": 0,
|
| 3041 |
+
"epoch": 10.0,
|
| 3042 |
+
"init_mem_cpu_alloc_delta": 2791841792,
|
| 3043 |
+
"init_mem_cpu_peaked_delta": 413900800,
|
| 3044 |
+
"init_mem_gpu_alloc_delta": 497994240,
|
| 3045 |
+
"init_mem_gpu_peaked_delta": 512,
|
| 3046 |
+
"step": 4660,
|
| 3047 |
+
"total_flos": 1.0068024259071967e+19,
|
| 3048 |
+
"train_loss": 1.525319737733178,
|
| 3049 |
+
"train_mem_cpu_alloc_delta": 8646197248,
|
| 3050 |
+
"train_mem_cpu_peaked_delta": 402849792,
|
| 3051 |
+
"train_mem_gpu_alloc_delta": 2010885120,
|
| 3052 |
+
"train_mem_gpu_peaked_delta": 15468135424,
|
| 3053 |
+
"train_runtime": 218631.5085,
|
| 3054 |
+
"train_samples_per_second": 174.954,
|
| 3055 |
+
"train_steps_per_second": 0.021
|
| 3056 |
+
}
|
| 3057 |
+
],
|
| 3058 |
+
"max_steps": 4660,
|
| 3059 |
+
"num_train_epochs": 10,
|
| 3060 |
+
"total_flos": 1.0068024259071967e+19,
|
| 3061 |
+
"trial_name": null,
|
| 3062 |
+
"trial_params": null
|
| 3063 |
+
}
|